The Latvian Gambit Lives!

Tony Kosten

B.T. Batsford Ltd, London

First published in 200 1 e Tony Kosten 2001

ISBN 0 7 134 8629 S

British Library Cataloguing-in-Publication Data. A catalogue record for this book is available from the British Library.

All rights reserved. No part of this book may be reproduced, by any means, without prior permission of the publisher.

Printed in Great Britain by Creative Print and Design (Wales), Ebbw Vale for the publishers, B.T. Batsford Ltd, 9 Blenheim Court, Brewery Road, London N7 9NT

A member of the � Group plc

A BATSFORD CHESS BOOK

Contents

Page

Introduction S

Main line---3ltJxeS "f�Introduction and 7 f3 7

2 Main line---other seventh moves 30

3 Nirnzowitsch's variation, 6ll)e3 47

4 Bronstein's variation, 6 �e2 SS

S Leonhardt's variation, 4 ll)c4 68

6 3 ... lL\c6 and other third move alternatives for Black 100

7 3 �c4 fxe4 4ll)xeS "gS 117

8 Svedenborg's variation 4 ... dS 140

9 Other replies to 3 �c4 16S

10 3 exfS 17S

11 .3 d4 199

12 3 lL\c3 Mlotkowski' s variation 210

13 Unusual third moves for White 217

Acknowledgements

For this new work I am indebted, once again, to John Elburg, editor of the Latvian Newsletter, and Georgio Ruggeri Laderchi who sent me a copy of their CD Latvian Gambit into the next Millennium, plus nu­merous updates, and other material, and who were happy to discuss variations with me.

I would also like to thank these last two, plus Hagen Tiemann, and Mauro Voliano for consenting to play a training Latvian Gambit tour­nament with me, so that I could run through some of my analysis, and, finally to Stefan Bucker for sending me copies of his excellent publica­tion, and unpublished analysis.

Bibliography

Books The Latvian Gambit, by Tony Kosten Developments in The Latvian Gambit, by K.Grivainis & J.Elburg The Latvian Gambit, by K.Grivainis Latvian Gambit made easy, by K.Grivainis & J.Elburg Le Gambit Letton, by F.Destrebecq Encyclopaedia of Chess Openings 'C' Sahovski Informators #55 and #56 Frank Marshal/ 's Best Games of Chess, Frank J. Marshall N90, J.Nunn, G.Burgess, J.Emms & J.Gallagher

MagulDes Kaissiber, ed. Stefan Bucker Latvian Correspondence Chess & Latvian Gambit

Electronic Latvian Gambit into the next Millennium, by John Elburg & Georgio Ruggeri Laderchi The Latvian Newsletter, edited by J.Elburg

I was also assisted by ChessBase 7.0 and Fritz 6

Introduction

The Latvian Gambit can be thought of as a sort of King's Gam­bit played by Black:

I e4 e5 2 1013 f5

The move 2 ... f5 was originally quoted by Damiano in 1512 (and not in the Gottingen Manuscript, which only mentions 1 e4 eS 2 lL)f3 d6 3 .*.c4 f5 4 d3 f4), and for centuries the opening was known as the Greco Counter Gambit after the Italian Gioachino Greco (1600-1634). Then, from around 1900, the Latvian Karlis Betin! started to play and analyse the opening, winning many exciting games. He was joined by other Latvian masters, and from 1934 to 1936 the Seniors Club of Riga, under Betin!, employed the opening to gain a famous victory in the correspondence match against Stockholm. Recognizing the contribution of the Latvian chess players to the renewal of interest in the opening, it was renamed the

Latvian Gambit at the 1937 FIDE Congress.

When I agreed to revise my book The Latvian Gambit for Batsford, I told them that I did not want to spend more than a week or two over it. Fortunately for the reader, and unfortunately for my free time. that was four months ago! This revision has become a labour of love for me. I have been through all 9000 or so Latvian Gambits ever published, with the help of the ChessBase 'Tree' function to catch any trans­positions, and have deeply analysed all the critical lines.

Although I have kept the same basic structure (after all, it seemed logical to me in 1994, and it still does) the contents have been thoroughly reworked. and updated. There is no subjective bias, this is no 'Winning with ...• book-when a line is good for White I say so.

Interestingly, when I wrote the original tome, I based it around games of the strongest over-the­board players (in particular lonny Hector), because the standard of many correspondence games was not that high. However, seven years on, it is the games by the correspon­dence players that are the real main­stay of the book. Not only are these players now armed with databases containing all the relevant games and can use published literature to recite the most theoretical lines, but they also have access to powerful

6 Introduction

analytical engines. This means that many correspondence games are now tactically flawless, and that the theory has been pushed to new limits.

say which lines are definitely good, which bad. I hope the reader gets much enjoyment out of reading this work, and scores many points!

The Latvian Gambit is now a mature opening, and it is possible to

Tony Kosten, Chamalieres, March 200 I

Symbols wed in this book

+ check +/- winning advantage for White ± large advantage for White 1; slight advantage for White -/+ winning advantage for Black + large advantage for Black � slight advantage for Black

level position good move

!! outstanding move !? interesting move ?! dubious move ? bad move ?? blunder 1-0 the game ends in a win for White 0-1 the game ends in a win for Black Jh-Ih the game ends in a draw corr. correspondence game ch championship m match

1 Main line: 3 It)xe5 1i'f6 Introduction and 7 13

I e4 e5 2 lL)f3 f5 3 lL)�e5 This is White's most natural

continuation and also traditionally considered to be his strongest. capturing the important eS pawn and simultaneously threatening -..5+.

3 •••• f6

At first sight a strange reply. im­mediately throwing the black queen into the fray before developing any of the other pieces. The mundane tactical point is that Black parries the -..5+ menace whilst attacking the white knight on eS. thus guaran­teeing the recapture of the pawn on e4. Less obviously. although Black suffers temporarily from a lack of mobilization. in the long tenn he hopes to be able to profit from the aggressive placement of his queen on the kingside which. when com­bined with the rook on the open f-fiIe. opens possibilities of an eventual attack on the white king. Other third moves for Black. more

in keeping with real gambit play. are considered separately, in Chap­ter 6.

4 d4 Apart from this, most obvious of

moves, White can also play the 4 lZ)c4 of Chapter 5, but should cer­

tainly avoid: a) 4 f4?! d6 5 lZ)c4 fxe4 (6 lZ)c3

can be met by either 6 ...• g6, or by 6 ...• xf4 7 d3 .g4! with a very �Ieasant endgame in store) 6 d4 .g6 7 lZ)c3 lOft) 8 lZ)e3 c6 9 �e2 dS 10 0-0, lzuel-Cordoba, corr. 1975176, when the simple 1O .•. �d6 grants Black an edge.

b) 4 lL)O?! fxe4 5 .e2 dS (S .. ,.e7! 6 lZ)d4 lL)ft) which is quite equal after 7 d3 [7lZ)c3 dS 8 d3?! cS 9 lL)b3 .i.g4! and White is losing, Vaemik-Dreibergs, corr., USA 1968] 7 . . . exd3 8 .xe7+ �xe7 9 .i.xd3 0-0 and Black has an extra central pawn) 6 d3 �f5 7 dxe4 dxe4 8 lL)fd2?! (8 lL)bd2) 8 ...• e6 9 lZ)c4 lZ)c6 10 .i.e3 0-0-0 11 lL)bd2 lL)ft) 12 h3 .i.b4 13 g4!? lbxg4?, Alden­Sawyer, USA 1990, and now both players overlooked 14 hxg4! �xg4 IS .xg4! .xg4 16 �h3 when' Black has only two pawns for the piece.

4 ••• d6 Chasing the knight away,

4 . . . lZ)c6?! 5 lbxc6 dxc6 6 eS .17 transposes to Chapter 6, whilst 4 . . . fxe4? 5 �c4, 4 ... dS? 5 exdS �d6 6 �bS+, and 4 . . . �? 5 exf5 are all terrible.

8 Main line: 3 ltll-e5 "/6 Introduction and 7 f3

S liX4 Otherwise: a) There is no reason to give

Black free tempi by 5 lDf3 (Fine's recommendation, although first played by Staunton in 1841 ), S . .. fxe4 6 lDgS (6 "e2?! "e7! 7 lDgS [7 lDfd2 is no better, 7 ... lDf6 8 c4 .i.f3 9 1Dc3 Landgren-Zalitis, Sweden 1978, when 9 . ../0:6! would have been difficult to meet, e.g. 10 lDb3 lDb4-d3+] 7 . . . lDf6 8 1Dc3 dS 9 .i.f4 h6, when White is forced to play the ugly move 10 lDh3, Polland-Pupols, USA 1968. and 6 .i.gS?? loses a piece, 6 . . ... g6 [6 ..... 13 amounts to the same] 7 .e2 .i.e7 8 .i.xe7 exf3 0-1, Deronne-Gedult, Paris 1975) when 6 . . . dS with the threat of . . . h6 is best, e.g. 7 c4 h6 (probably Black should withhold the move ... h6 until it is most effective: 7 ... c6! 8 1Dc3 .i.b4 9 "'3 [9 cxdS cxdS 10 a3 .i.xc3+ 11 bxc3 lDe7 12 c4 0-0 with an edge, as ... h6 is coming, Martinez­Krustkains, Atars 1979] 9 ... .i.xc3+ 18' bxc3 h6 1I lDh3 lDe7 12 .i.e3 0-0, and the h3-knight is offside, Destrebecq-senechaud, corr. 1997) 8 lDh3 .i.xh3 9 �S+! (9 gxh3 Steiner-Apsenieks, Kemeri 1937, 9 . . . .i.b4+!? [or 9 . . . dxc4 10 .i.xc4 lDc6---Destrebecq] 10 1Dc3 lDe7 I1 cxdS 0-0 equal) 9 ... g6 10 .xh3

1Dc6 1I .i.e3 lDb4 12 1Dc3! lDe7 (if 12 . . . 1Dc2+ 13 �d2 lDxal 14 lDxdS .0 then IS .g4! intending 16 .xe4+, followed by 17 .i.d3 and 18 l:lxal-Destrebecq) 13 l:lcl c6 14 a3 with some advantage, Destrebecq-Zerbib, corr. 1997.

b) Similarly, 5 lDd3 fxe4 6 lDf4 has little to recommend it, 6 ... c6 (6 ...• 0 looks like a good response, planning ... dS and ... .i.d6) 7 c4! .0 8 1Dc3 /l)f6 9 .i.e2 .i.e7 100-0 0-0 11 dS is slightly better for White, Franco-Hector.

S ••• fse4 Black sometimes tries S ... 1Dc6?

here, but White has at least two good replies: 6 dS lDes 7 1Dc3 lDe7 8 lDxeS .xeS 9 f4 .f6 10 .i.d3 g6 11 lObs �d8 12 .i.e3 cS 13 .i.d2 .i.d7? 14 .i.c3 1-0 Schilling-Scott, corr. 1988, and 6 eS dxeS 7 dxeS with advantage: not then 7 ... lDxeS?? Vitols-Vildavs, Latvia, when 8 .e2 .i.d6 9 f4 appears to win a whole piece.

6lOc3 Immediately attacking the e4

pawn. The drawback of the text compared with the moves 6 .i.e2 and 6 lDe3, which are covered in Chapters 3 and 4, respectively, is that c4 is no longer possible. In­stead, Bronstein mentions the possi­ble pawn sacrifice 6 �S+!? g6 7 .e2 (7 .dS is met by 7 ...• 13 8 lDe3 lDf6! 9 "xf3 gxf3 10 c3 dS with extra space, Cabanas-Aziz, Madrid 1999) 7 ..... xd4?! (too risky for practical play, 7 . . . dS is best, 8 lDes lDd7! 9 f4 c6 and Black has nothing to fear, Collins-Hage, corr. 1999) 8 1Dc3 (but 8 .i.d2!, threaten­ing .i.c3, looks unpleasant) with a useful lead in development: a likely continuation is: 8 . . . dS 9 lDbS .g7 10 .i.d2, unclear.

Main line: 3'fu-e5 ./6 Introduction and 7 f3 9

6 ...• '6 A key move for Black-he will

decide on the placement of his mi­nor pieces according to his oppo­nent's play. In the meantime White will have to find time to defend his g2 pawn. Other, less well docu­mented, possibilities:

a) 6 . . . .i.f5?! with, a I) 7 It)e3 .i.g6 (conceding the

important light-squared bishop is dangerous for Black: 7 . . . c6?! 8 �f5 .xf5 9 .i.e2 /l)ffi to 0-0 J..e7 11 0 exO 12 :xO .d7 13 dS 0-0 14 J..gS .c7 IS J..d3 /l)bd7 16 :h3 :ae8 17 J..f5 J..d8 18 .d3 hS, Grivainis-Diepstraaten, corr. 1977/80, and now 19 J..e6+ is a kil­ler, and 7 . . . lt)e7!? is probably best met with 8 dS, rather than by 8 �dS /l)xdS 9 �dS .d8 to c4 llXl7 with no particular problems for Black. Haluska-Stepanski, USA 1987) 8 It)edS!? .t7 9 J..c4 c6 10 /l)b6 dS 11 /l)xa8 dxc4 12 dS ( 12 .g4! .d7 13 .g3 might just allow the knight to escape) 12 . . . J..cS 13 �e3 J..xe3 14 fxe3 /l)ffi 1 S .d4 0-0, Grivainis-Pupols, corr. 1971172, and now 16 d6 gets the knight out.

a2) White can try for a refutation with 7 g4!? J..g6, and now 8 J..g2 (8 h4!?) 8 . . . c6? (8 . . . �6 is Black's

best chance: 9 dS [9 J..e3?! dS to It)es 0-0-0 is fme for Black, Zjajo-Steinhart, Unter8!Ombach 1999J 9 . . . lt)eS to �eS "xeS 11

.e2 0-0-0 12 J..e3 a6 13 0-0-0 with a small advantage, Hemandez­Bedevia, corr. 1990) 9 J..xe4 J..xe4 to �e4.e6 11 .e2 dS 12 /l)ffi+!! ( 12 /l)cd6+ is also emphatic, 12 . . . J..xd6 13 �d6+ �d7 14 .xe6+ �e6 IS �b7, GUrd­Svendsen, corr. 1994198) 12 . . . � 13lt)eS+! �ffi 14 J..gS+! �gS IS .0 J..b4+ 16 �fl! (this wins the black queen, not 16 c3? g6! 17 cxb4 llXl7 18 .e3+ � 19 �d7+ .xd7 20 .eS+ �f1 21 .xh8 :e8+ 22 �d2 /l)ffi winning! JacOon­Svendsen, corr. 1991) 16 . . . g6 17 /l)f1+ .xt7 18 h4+ �h6 19 .xt7 with mate to follow, Kozlov­Svendsen, corr. 1994/98.

b) 6 . . .. t7?! and bl) Note that 7 J..e2! is very

strong here, forcing a transposition into Chapter 4, variation (d).

b2) Also, it is not clear whether the e4 pawn is really immune from capture in this position after all, 7 �e4!? .e7?? 8 J..d3! /l)ffi (there are no better alternatives for Black on move 8 either, i.e. 8 . . . dS 9 Whs+ .f1 to .xt7+ �t7 11 It)eS+, or 8 . . . J..f5 9 .e2 dS to /l)cd6+ cxd6 11 �d6+ �d7 12 �f5, both crushing for White.) 9 0-0 J..e6 to :e 1 �d7 11 /l)gS 1-0, Gaard­Lemay, corr. 1989.

The question remains: what would White have played �gainst the somewhat superior 7 . . .. e6 ? Per­haps he intended something like 8 J..d3 (my suggestion 8 0 dS 9 It)es dxe4 to J..c4 .d6 11 0-0 gives White an attack for the piece) 8 . . . dS 9 It)es dxe4 10 J..c4 � (to . . .. f5

JO Main line: 3 �e5 ./6 Introduction and 7 j3

11 lLlf7) 1 1 1rhS+ g6 12 .*.f7+ ,*,e7 13 .*.gS+ lLlffi 14 .*.xg6 .xd4 (14 ... .*.g7 IS .*.xh7! :fS 16 .g6 .*.h8 17 .xe4 .*.e6 18 0-0-0 1-0, Dorer-Stader, corr. 1982) IS .*.e8!

.*.e6 16 lLlg4 .*.&7 17 c3 .c4?! [ 17 ...• dS!] 18 lfuffi .*.xffi 19

.*.xffi+ '*'xffi 20 1rh4+ '*'g7 21 .gS+ '*'fS 22 .*.bS! .dS 23 .ffi+ '*'g8 24 :d I winning, Novak­Bonte, corr. 1995.

b3) 7 ltle3 lLlf6 (7 ... c6?! loses a eawn 8 lLlxe4 dS 9 lLlgS .ffi 10 Illg4 '.*.b4+ 11 c3 .e7+ 12 ltles -*.d6 13 1rhS+ g6 14 .e2 lLlffi IS lLlgf7 0-0 16 lLlxd6 .xd6 17 .*.h6 with advantage, Rittenhouse-Krantz, corr. 1993) 8 .*.c4 .*.e6 9 dS (9 -*.xe6 .xe6 10 dS .e7 11 0-0 with some advantage, Grivainis­Morgado, corr. 1977179) 9 ... .*.c8 (9 ... .i.d7?! obstructs the b8-knight 10 0 exO 1 1 .xO g6 12 0-0 .*.g7 13 .*.d2 0-0 14 :ae I and Black has problems developing his queenside, Jelling-Bruun, Denmark 1980) 10 0-0.*.e7 1 1 0 exO 12 LO (if 12 .xO lLlbd7 intends ... lLleS) 12 ... 0-0 (now 12 ... lLlbd7!? also with the idea ... lLleS (Melchor) can be met by 13 lLlbS -*.d8 14 lLlfS, but 12 ... a6 may be better, 13 lLlfS .*.xfS 14 :xfS lLlbd7 is reasonable, Diepstraten) 13 lLlbS lLla6 14 a3 with the knight aiming for d4, thence e6, Ruisdonk­Peet, corr. 1977.

c) 6 ... c6? 7 lLlxe4! .e6 8.e2 dS, von Bilguer-von der Lasa, Berlin 1839, when 9 lLled6+ '*'d8 10 �b7+ ,*,c7 1 1 .xe6 .*.xe6 12 lLlcaS would have been convincing.

d) 6 ... lLlc6? 7 dS lLles 8 lLlxe4 .g6 9 lLlgS with advantage, Downey-<>ren, corr. 1994198.

e) 6 ... lLle7!? 7 dS .g6 8 .d4 .*.fS 9 lLle3 ltid7 10.*.d2 lLlffi =i= Mileika­Benins, Latvia 1977 , is more interesting.

And so we come to the first major crossroads, how should White pro­ceed? If Black can successfully complete his development, he will be fine. His position is solid and r0-bust. On the negative side, he is temporarily behind in development and his e4 pawn is exposed. It is perhaps not surprising, there�ore, that White is most successful With 7 0, attacking the e4 pawn head on and immediately opening up.the p0-sition. More positional methods fail to trouble Black, partly because White has already slightly compro­mised his position by blocking his c2 pawn.

7 f3! Chronologically one of the first

moves employed against the Latvian Gambit. The alternatives are consid­ered in Chapter 2.

7 ••• edJ Normal, although recently

7 ... lLlffi?! has been attracting more attention, but the doubtful assess­ment remains (7 ... .*.e7?! 8 fxe4 [but 8 lLlxe4 is possible, for if 8 ... dS 9

ltleS] 8 ... lLlffi transposes). 8 fxe4 (8 ltle3 is sometimes preferred, and transposes into a sub-variation of B), Chapter 2: 8 ... .*.e7 9 .*.c4 etc.) 8 ... .i.e7 (Black is in no position to recapture the e-pawn, 8 ... lLlxe4? loses after the obvious 9 .*.d3 when 9 ... dS! is the only move to avoid losing a piece [9 ... -*.fS?! 10 .e2,

Main line: 3 1ili-e5 "/6 Introduction and 7 f3 / /

viz 1O ... .*.e7 11 lDxe4 dS 12 IOes � 13 lDcs with advantage Schaller-Zernack. corr. 1992, and 10 ... dS 1 1 lDxdS .*.d6 12 lDxd6+ cxd6 13 lDf4 1-0, Marin-Harju, corr. 197 1] 10 lDxdS [ 10 O-O! is a good, and simple alternative] 1O ...• xg2 but I would not fancy Black's chances after 1 1 �S+ WdS 12 lIfl lDa; 13 lDxa; gxa; 141Oes!. Neither are S ... .*.g4?! 9 .*.e2 a6 10 0-0 lDbd7 11 eS .*.h3 12 .*.0, Rehfeld­Rebber, corr. 1992, and S ... 1Dc6 9 .*.e3? [91Oe3! leaves White a pawn to the good] 9 ... lDxe4!? 10 .*.d3 .*.f5 11 .0 dS? [II...lDxc3 12 .t.xfS .0 is _play able ] 12 lDxdS 0-0-0 13 lDf4 WeS 14 g4 lDxd4 15 .i.xd4lhd4 160-0-0 .t.d7 17 lIhel .t.c6 IS c3 with advantage Svendsen-Jackson, corr. 1990, any improvement) 9 eS! is the theoreti­cal 'refutation' (although a simple continuation like 9 "d3 is possible, 9 ... 0-0 10 1Oe3 [10 .*.e3 can't be bad either, of course, Black has no great lead in development to speak of] 1O ... 1Dc6 11 .t.d2 Rajlich­Lapshun, New York 1999 [the fur­ther pawn grab, 1 1 "c4+?! WhS 12 dS �S 13 "xc7 lIt7 is far too risky], when the evident II...lDg4! 12 liSxg4 [ 12 lDcdS?! .*.h4+ 13 g3 lDf2] 12 ... .*.xg4 traps the white king in the centre; Black will continue ... lIaeS, and obtain good play).

Now Black has three alternatives: a) 9 ... lDhS 10 .t.d3!? ( 10 .t.e3 0-0

1 1 g3! lDc6?! [ 1 1....t.g4 12 .t.e2 .*.xe2 13 "xe2 is a better chance, but still excellent for White] 12 .t.e2, Hempel-Grava, corr. 1971172, is probably even stronger, showing up the awkward placing of the hS-knight, and on 12 . ..t&g3 13 lIg I should win) 1O ..... xg2 1 1 "xhS+ g6 12 .t.xg6+ "xg6 13 "xg6+ hxg6 14 0-0 .t.e6 (l4 ... .t.f5 15 exd6 .t.xd6 16 lDxd6+ cxd6 17 lLldS Wd7 Is IOe3 .t.e4?! 19 dS lieS 20 lDg4 1-0, Grivainis-caUinan, corr. 1971n2, or 14 ... dxeS 15 IO:ts .t.h3?! 16 lIel .t.h4 17 lIe3! .t.e6 IS lDxc7+ Wd7 19 lDxe6 winning quickly, Grivainis-Zemitis, corr . 1970) 15 dS .*.h3 16 lIel with ad­vantage Grivainis-Gabrans, corr. I 969nO.

b) 9 ... �g4!? First played by lonescu, but popularised by Krantz, this move is the reason for the re­newed interest in this pawn sac line. Black plans tactics involving ... 0-0, and ... .i.h4+.

10 .t.d3, as usual, is the sharpest ( 10 exd6!? is greedy, but might just be possible, 1O ... .t.h4+!? [ 1O ... cxd6 II .t.d3 �S 12.0 is clearly good for White, as in Ghinda-Ionescu, Odorheiu Sec::uiesc 1993, the inau­gural game in this line] II g3 0-0 12 dxc7?! hardly necessary, [ 12 gxh4! is strong, as after 12 ... lDt2 White saves material with the tactic 13 IOeS! and the queen has no satisfac­tory square, e.g. 13 . . .• xd6 14 lObs .f6 15 .t.gS] 12 ... lDxh2!! 13 cxbS= • .t.xg3+ 14 "xg3? [ 14 Wd2 lhbS 15 1tle2 is a better defence] 14 ..... xg3+ 15 Wd2 lDo+ and White should have given his queen to avoid mate, Svendsen-Krantz, corr. 199419S), 10 ...... 5 with a choice:

12 Main line: 3 ltlxe5 "/6 Introduction and 7 f3

bl ) 11 .i.f4!, the kingside needs defending, 11...0-0 (11...llX6? 12 exd6 .i.h4+ 13 g3 0-0 14 .d2!

lOxd4 15 O-O-O! led to a quick loss for Black in Fressinet-�nechaud, CrCon 1996) 12 .f'3!. Heap has been very successful with this move, White defends the bishop, and prepares to castle (once again 12lDdS?! tempts fate, 12 ... .i.h4+ 13 g3 c6!? [ 13 ... llxf4 141Oxf4 .i.xg3+ 15 hxg3 .xh I + 16 Wd2 is unclear] 14 1:&.7 Ilxf4 15 gxh4? dxeS 16

lOxeS? "xh4+ 17 Wd2 lOxeS 18 dxeS "gS 19 lOxa8 1le4+! 0-1,­Gllrd-Krantz, corr. 1994/98, 20 Wc3 .xeS+ 21 Wb3 1lb4+!! 22 Wxb4 "xb2+ leads to mate)' 12 ... llX6 (I2 ... dxeS 13 lOxeS -*.d6 14 0-0 Wh8 15 h3 lOf6 16 .xhS

lOxhS 17 lOg6+! 1-0. Heap-Sireta. corr. 1994/98, as 17 ... hxg6 18 .i.xd6 Ilxfl + 19 Ilxfl cxd6 20 1lf8+ Wh7 21 lDe4! is soon mate) 13 "dS+! Wh8 14 exd6 (the point. White of­fers the exchange of queens, and kills the black attack) 14 ... -*.h4+ (14 ..... e8 15 O-O! cxd6 161Oxd6 is crushing for White, Heal!:'Kozlov. corr. 1994/97, and 14 .... xdS 15

lOxdS .i.xd6 16 lOxd6 cxd6 17 .i.xd6 :d8 18 0-0 is also hopeless for Black. Heal!:'Senechaud, corr. 1994/96) 15 g3 .e8+ 16 Wd2 -*.e6 17 "e4 lOf6 18 "g2 lOhS 19 -*.e3 and Black's initiative has ground to a halt. Heap-Krantz, corr. 1994/96.

b2) 11 lDds courts danger, 11....i.h4+ 12 g3 0-0 ( I2 ... -*.xg3+? Il-Wd2) 13 -*.e4 ( 131Oxc7?! allows Black good play: 13 ... .t.xg3+ 14 Wd2 .t.f4+ 15 Wc3 dxeS 16 -*.xf4 exd4+ [17 Wb3!? lDa6! {17 ... llxf4 18 Ilel !) 18 a41Oxc7 19 -*.xc7 bS is unclear] 17 Wd2 17 ... llxf4 18 Wcl [18 �a8 �6!] 18 ... bS 19

lOxa8 �6 20 .d2 1Of2 21 llX1?

[21 Ilel ] 2 I...lDxh I and Black should win, Villat-Radovic, 3rd Lat­vian Gambit WCh 1999; 13 _ph4 .xh4+ 14 Wd2lOf2 15 .el .xd4 is also wild) 13 ... llX6!? (develop­ment!) 14 lDxc7 (141Of4) 14 ... dxeS ISlOxa8 exd4 16.e2 Ilf2 17.d3

1Ob4 Black starts throwing the kitchen sink at White! 18 �3 d3! 19lOc3+! (19lDd6+ Wh8 20 -*.xd3 -*.xg3 21 lDe4 1lxh2+ 22 lOxg3 .eS+ is good for Black) 19 ... Wh8 20 cxd3 1lxh2 21 O-O! .i.e7 22 lDxg4 -*.xg4 23 d4 lle2 24 .t.g2 (it seems that White is consolidating, but Black throws some fresh wood on the fire) 24 ... llX6 25 .dS 1lxg2+! 26 .xg2 lDxd4 27 .xb7

lOc2+ 28 Wf2 .i.d6 29 We I �xg3+ 30 Wd2 .gS+ 31 Wd3 .d8+ 32 Wc2 lDd4+ 33 Wc3 .i.eS 34 .i.gS

1Of'3+ 35 Wb3 .t.e6+ 36 Wa4 -*.d7+ 37 Wb4 -*.d6+ 38 Wb3 -*.e6+ 39 Wa4 and now, rather than 39 ... lDxgS!? playing for the win, but going on to lose, Downey-Krantz, corr. 1994/98, Black should proba­bly take the perpetual check by 39 ... .i.d7+.

c) 9 ... lDe4 1O .f3 -*.h4+ 11 g3 lDxg3 12 bxg3 .t.xg3+ 13 Wd2 1lf8 14 WdS, Grivainis-Strautins, corr . 1971/72, is also insufficient.

I.KO

Main line: 3 �e5 ./6 Introduction and 7 f3 13

The key opening position, White enjoys a lead in development and will use the open e and f files for his rooks. Black will often need to try to exchange queens on g4 to blunt the white attack, but the ensuing endgame (assuming White cannot advantageously avoid it) favours White who can often play lile3-f5, gaining the bishop pair.

Now we have:

A 8 ••• 1t1f6, B 8 ••• .*.e7!? C 8 •.. ltlc6?!

Others: a) After 8 ...• g4?! 9.12, keeping

queens on the board, probably Black should try 9 . . . 1t1a; (as, unfortu­nately, 9 ... 1Oc6 fails to force a trans­position to the main line, White can answer either to ltlb5 or 10 ltld5, and 9 ...• f5?! wastes more time with the queen, 10 .*.f4 ltla; II 0-0-0 �e7 12 lile3 .a5 13 �c4 and Black was in trouble, Morgado­Pupols, corr. 1970/72) 10 �e2, the problem, to ...• e6 II lile3 (menac­ing .*.c4) II.. .d5?! ( l l . . .c6 was nec­essary) 12 lOcxd5! ltlxd5 13 .*.g4 �b4+ 14 c3 ltlxc3 15 O-O! l:tf8?! (this loses a piece, 15 ...• c6 16 .n+ Wd8 17 bxc3 is not particu­larly appetising, though) 16 �xe6 l:txf2 17 �xc8 and the threats to both c3, and b7, are decisive, Downcy-Cattennole, corr. 1999.

b) 8 . . . c6? is a luxury Black can ill afford: 9 �d3 .a; 10 .e3+ (10 .*.f4!? d5 [ to ...• xd4? II 0-0-0 must win quickly J II O-O-O! with advantage, Clarke-Kozlov, corr. 1987) 10 ...• e6, Endzelis-Strelis, corr. 1970/72, when II lile4! d5 12 ltlcd6+ would have been most unpleasant.

A 8 ••• 1t1f6

This quiet move aims to complete kingside development, and is rec­ommended by the German theoreti­cian Bucker.

9�d3 Otherwise, 9 lile3!? is an interest­

ing idea, as Black often seeks the exchange of queens on g4 in this variation, White decides to make this square unavailable, even if it means sacrificing a pawn. 9 . . . 1Oc6 (9 . . . c6?! is passive, to .t.d3 .n [the endpme after to . . . Wh5 II 1rxh5+ fi)xh5 12 0-0 Jl..e7 is also unpleasant for Black after the stan­dard 13 ltlf5, Clarke-Elburg, corr. 1994195J 11 0-0 �e7 12 Ite I! [even stronger than 12 ltlf5J 12 . . . 0-0 13 �c4 �e6 [forced, 13 . . . d5? J4 lOcxd5! cxd5 15ltlxd5 winsJ 14 d5! Clarke-Svendsen, corr. 1987, and now only 14 . . . cxd5 15ltlexd5 Wh8 avoided immediate loss) to �d3!? (a pawn sac, 10 ltlb5!? also causes a certain amount of inconvenience, and may be superior, to . . . Wd8 11 �d3 .e8 12 c3 a6 13 I.Oa3 although 13 . . . �_g4 14.g3 d5 15 0-0.t.d6 16 Wh4 Wh5 kept White's advantage to a minimum, Pihlajasalo­Hamalainen, Helsinki 1996) 10 ...• n (l0 .. . �5? loses to II 1rxh5+ ltlxh5 12 �, Tzermiadianos-Fragiadakis, Hania

14 Main line: 3 llh-e5 ./6 Introduction and 7 f3

1995) 11 0-0 lOxd4 12 .dl �e6 13 �c4! 1Oc6 (Bucker's 13 . . . cS is also possible) 14 � and now, rather than 14 ... Wd7?! IS .lgS with some pressure, Svendsen-Melchor, corr. 1990, Bucker points out that 14 . .. 0-0-0! is simple, IS lOxR) (IS �gS �e7 16 lOxffi gxR) 17 �xc6 bxc6 18 .d4 cS is also fine for Black) IS ... gxR) 16 �xc6 bxc6 17 .d4 I had previously intimated that this was good for White, but now Bucker suggests the line 17 .. . cS 18 .84 Wb7 when White can force a draw by 19 �S+ Was 20.c6+ but no more.

9 •••• g4

IO .e3+ The endgame following 10 .xg4

offers White a small plus, as Black no longer has the resource .. . lOc6-b4, as in the main line C: 10 .. . �xg4 110-0 1Oc6 12 �gS �e7 13 :f4! (13 Ilae l 0-0-0, Hempel­Strautins, corr. 1970, and now 14 �3�fS is best) 13 . . . �hS!? (1�L.�7 14 �xR) �xR) IS 1Ods, Downey-Elburg, corr. 1994/96, when Bucker suggests IS .. . �gS 16 :el+ Wd8 17 :t7 g6 but I don't like the look of the rook on the sev­enth rank. myselt) 14 �xR) (critical, 14 :e l 0-0-0 IS :fS �g6 16 dS lbes 17 lOxeS dxeS 18 :fxeS �cS+ 19 �e3 �xe3+ 20 :Sxe3 lOxdS is

equal, and 14 dS!? lbes IS lOxeS dxeS 16l1h4 �f1 17 :e l � 18 �xR) gxR) 19 Zlh6 �e7 20 1hh7 1hh7 21 �xh7 lId8 22 lId l c6 grants Black sufficient play thanks to his bishop pair) 14 .. .... xR) IS lOds .bd4+ 16 hd4 lOxd4 17 lOxc7+ Wd7 18 lOxa8 has 19 :el (if 19 �xh7 then Black obtains 8 useful initiative by 19 ... dS! 20 lbe3 lbe2+ 21 Wh I d4 22 lOc4 :f8 23 �d3 :t2 24 :e I lLlf4) 19 .. . �f7 20 c3 1Oc6 21 a4 h6 22 �f5+ Ih-Ih Elburg-Kosten, emaiI2001.

However, to .t2 appears to make more sense here than in the main line, as White does not lose his d-pawn: 10 .. . �e7 11 0-0 0-0 the disadvantage is the white queen's placing on the f-fiIe, opposite the black rook, 12 h3 Whs 13 �f4 lOc6 14 Ilae I �d7 White has more space, but Black is solid, Bergsma­Wijnands, Rotterdam 1939.

IO .. ie7 1 1 0-0 Wb5!?

Bucker adorns this move with two exclamation marks, the point being that dS is further controlled, and White's dS-push stopped.

Instead, l l...lOc6 forces a trans­position to the main line, and offers Black the dual advantage of avoid­ing the line with 9 dS, and 9lLlbs.

Bucker's analysis continues: Il �dl

Main line: 3 �e5 .}is Introduction and 7 f3 15

1 2 ll)bs can be met by 12 . . . lOa6 now, e.g. 1 3 �d2 when 13 .. 0-0 is possible, as 14 .xe7!? lIe8 traps the queen I S .xc7 �c7 16 �c7 although White may be better here, anyway.

1 2 lIel is easily met by 1 2 . . . liX6 now, as 1 3 dS?! ( 1 3 .i.e2 �g4) 13 .. . �dS 14 �dS .xdS I S �e4 Wf7 is fine for Black.

12 ••• lbc6 13 lObS 13 Jlael 0-0 14 dS!? �dS I S

Jlxf8+ �xf8 16 .i.xh7+ [ 16 �dS .xdS 17 .e8? �h3] 16 . . . �h7 17 .d3+ �fS 1 8 1IxdS �g6 seems OK for Black, who can rely on his bishop pair.

13 ... Wd8 The black king's position is awk­

ward but, as Bucker notes, "the white pieces aren't ideally placed', either.

14 lDas If 14 h3 .xbS!? (instead, Bucker

indicates 14 . . . a6 1 S liX3 ll)b4) I S IOes .xb2 16 1Ot7+ ( 16 lIfbl .xb 1 + 1 7 Jlxb 1 dxeS 1 8 dxeS /Od7 should favour Black) 16 ... Wd7 1 7 IOxh8 ( 1 7 �c3 1Ia3) 17 . . .• xd4 ap­pears very promising for Black.

And a "sensible' move like 14 liX3?! is met by 14 . . . lOg4 e.g. I S • f4 �d4!? (or I S . . . lIf8 16 1Ig3 �f6) 16 h3 1Ics 17 Wh l 1If8.

14 ... lOus IS �uS b6 16 �d1 lIe8 17 .gJ Wg4 18 .xg4 .i.xg4 19 �r4

Else Black plays .. . c6. 19 .... 6 10 � rill Jlad We7 . . . when White's advantage is tiny.

This line is certainly worth testing.

B 8 .. �e7!?

This move has the advantage that Black can offer the exchange of queens from f6, as well as from g4.

9 �d3 9 /Ods �h4+ 10 g3 �d8 I 1 �d3

1Ig4 12 .e3+ .e6 is not too fright­ening, Downey-MalmstrOm, COIT. 1994/96.

9 .... g4 All the same, Black should prefer

this square here, when White can

play an endgame with an edge, but no more, as after 9 . . .• f6?! lO �f4! is a problem lO . . . Wh4+ ( l O . . .• xd4? loses after 1 1 ll)bs .cS 12 �e3) 1 1 �g3 11.84 ( 1 1 . . ..xd4? again loses to 1 2 �S i.e. 12 . . .• cS 13 �f2 Wb4+ 14 c3, soon winning, Elburg­Yus, COIT. 2000) 12 0-0 ( 12 .12 looks more aggressive) 12 . . . 1Of6 1 3 Jlae I and Black suffers consider­ably, Grivainis-<:astelli, WIT. 1970n2 .

10 .xg4 10 o-O!? .xO 1 1 JlxO 1Of6 12

1Oe3 0-0 13 IOedS, White has some pressure, but the black position is solid, Kosten-Voliani, email 2001 .

1 0 ... �xg4 nlOe3 �e6 I I . . .�d7 is similar, 12 0-0 c6 13

IOrs �xrs 14 �xfS with the bishops, and some advantage, Downey-GrivainislHayward, COIT. 1997.

11 0-0 12 �e4!? c6 1 3 IOf5 �xrs 14

�xf5-White has the bishop pair, Heap-Melchor, COIT. 1994/96.

11 ... 10(6 13 �f5 �d7 1410eds

16 Main line: 3 �e5 "/6 Introduction and 7 f3

IS �gS?! IOxdS! 16 lOxdS �xgS 17 1Ou7+ Wd8 18 1Ou8i0c6

But the knight never escaped from a8. and Black won. Grivainisl Hayward-Strautins. corr. 1 999.

C 8 ••• i0c6?!

By far the most popular. Black hopes to gain time counterattacking tlfe d-pawn. but the disadvantage is that White's pawn push to dS will gain a tempo at some time.

White has four dangerous replies:

Cl 9iOds, Cl 9 lObS! Cl 9d5 C4 9 �d3

Otherwise. 9 �e3 is harmless. 9 . . . 1016 (but not 9 . . . 1Ob4? when. rather than to lOa3?! c6 1 1 dS. Webster-Crimp. Australia 1994. af­ter which 1 1 . . .1016 is fine. 10 lDbS! is strong. i.e. 1O . . . lDxc2+ 1 1 Wd2 lDxal 12 lDxc7+ Wd8 1 3 �d3 .16 14 lDxa8 which should win) to �d3 .g4 1 1 0-0 Sprod-Krustkains. corr. 1979. and now 1 1 . . . IOb4 is best.

9 �e2? lDxd4 10 .e3+ lOe6 1 1 �d2 lOf6 12 0-0-0 �e7 13 h3 0-0 ; Stockholm-Riga. corr. 1934/36.

Cl 9iOds

9 •. �g4! There is no other sensible way to

defend c7. IO �d3!? 10 .e3+ also has its points.

10 . . . Wd7 (or 1O . . . Wd8 1 1 �d3 "e6 12 0-0 lDge7 i) 1 1 �d3 .. n (but not 1 1 . . .�f5 12 IOf4 .16 13 �xf5+ .xf5 14 0-0 with a dangerous initiative. or 1 1 .. ..e6?! 12 IOf4!) 1 2 .gS and now. instead of 12 . . .• e6+? when either knight to e3 wins a piece. i.e. 13 lDde3 ( 1 3 lOce3 iDh6 14 0-0 also wins) 1 3 . . . dS 14 lDeS+ lDxeS IS dxeS iDh6 16 iDxg4 and the knight cannot be recaptured without allowing a fork with �f5+. Schrader-Mlotkowski. St.Louis 1 904, lilY old suggestion 1 2 . . . �e6 13 :0 We8 still looks unclear.

Main line: 3 �e5 ./6 Introduction and 7 f3 1 7

10 .f2?! allows the king to escape, though, 10 ... 0-0-0 1 1 .1d3 Whs 1 2 lOf4 .e8+ 1 3 .1e3 dS 14 lLxI2 and now, rather than 14 . . . lOxd4?! I S lrO 1Oc6 16 .1xa7 d4?! 17 .i.bS gS Endzelis-Tomson, corr. 1974, when 1 8 :ael is strong, 14 . . . 1Ob4 or a sensible developing move like 14 . . . 1Oft) is fine.

10 ••• .*. .. 3 The simplest, 1O ...• e6+?! is best

answered by 1 1 lOce3! .i.xf3 12 lOxc7+ �d7 1 3 lOxe6 �e6 14 gxf3 lOxd4 (analysis of Malmstrom) when I S �f2 is better for White, due to his bishops ( I S lrO was also reasonable, Stevens-Poland, corr. 1993, but why not keep the king close to the centre?).

1O . . . WhS!? is ambitious, but rather than 1 1 .e4+ �d7 1 2 lrO lIe8 1 3 .f4 .i.e6 14 lOde3 ( 14 lOce3!? .*.xdS I S g4 is possible) 14 . . . gS I S .12 lOxd4 which won a pawn, although the position remains messy, Charlick-Mann, corr. 1 88 1 , 1 1 We3+ �d7 12 1Of4 .fi 1 3 .g3 is promising.

1 1 .1l:g6+ hl:g6 Il lO:s:c7+? A typical computer mistake, 12

gxf3 lOxd4!? (12 . . . �d7 is safer) 13 lDxc7+ �d7 14 lOxa8 lOxc2+ should be tried.

Il ••• �d7 13 1O:s:a8 .1l:gl 14 IIgl .i.e4 15 cl .1e7!

The bishop goes to h4 to keep the e7-square free, I S . . . lOf6 16 b4! bS 1 7 lOe3 .i.e7 1 8 a4! is quite unclear, Kharitonov-Vujadinovic, corr. 1998.

16 b4 .i.h4+ 17 �dl lOf6 18 bS lOe7 19 .i.a3 1Of5!

There is no rush! lO �dl ll:s:a8 Black has a winning material ad­

vantage, Fritz 3 - Tasc RJO Ver.2, Bauermeister I99S.

Cl 9 1ObS!

This rare line looks very similar to the previous variation, is even more dangerous, and comes close to being a refutation of 8 ... 1Oc6.

9 • • ..*.g4 9 ...• e6+ 10 .i.e2 .d7 is the only

sensible alternative, but it doesn't look especially tempting.

10 Wcl! The latest idea, White prepares

threats to c7. Others: a) 10 .e3+?! is less effective,

1O . . . �d7 1 I .i.d3 .f7! to stop White from castling short. 12 11ft ( 1 2 .gS .i.e6 13 IIft lO(6) 12 . . . lOf6 1 3 .f4 WhS! (even stronger than 1 3 . . . lIe8+ 14 .*.e3 .1e6 as, although Black obtained the advantage after I S lDeS+!? dxeS 16 dxeS lOds?! [ 1 6 ... a6!] 17 .xf7+ .1xf7 1 8 11xfi+ �c8 19 .i.f4? g6 20 .1g3 lOxeS and went on to win in Macht-Betinl, corr. 1933/34, 19 lOxa7+ lOxa7 20 .1xa 7 IIxeS+ 2 1 �d2 looks play­able, and good, as 2 l . . .b6 can be met by 22 .*.&6+) 14 lDe3 1Ob4! I S .i.f5+ .1xf5 16 .xfS+ .xfS 17 IIxfS lIe8 1 8 �d I c6 19 1Oc3 dS and after some forcing play Black has a plus, GrivainislHayward­Downey, corr. 1997.

b) 10 .1d3?! .1xO ( 1O . . .• e6+?! I l lDe3! .1xf3 12 lOxc7+ transposes to 9 lOdS) 1 1 .1xg6+ bxg6 1 2 gxf3

18 Main line: 3 lOre5 ./6 Introduction and 7 f3

( 1 2 �c7+? also transposes to the above) 12 . . . �d7 1 3 dS ( 1 3 c3?! lIeS+ 14 lOe3 dS menaces ... a6, and favours Black) 13 ... 1Ob4 14 lOe3 lOf6 is very pleasant for Black, whose kingside pawns are no weaker than White's.

10 •• .Ilc8 This is not completely satisfac­

tory, but what else? a) IO . . . ()-(H)?! loses: 1 1 dS a6

(obviously the c6-knight cannot move without allowing mate) 1 2 dxc6 ( 1 2 lOxc7 is also convincing: 12 . . . �c7 1 3 dxc6 lIeS+ 14 .i.e3 [ 1 4 lOe3!?] 14 . . . dS I S .as+; this looks horrific for Black although in Malmstrom-Hage, corr. 1999, Black discovered considerable resources: IS . . . �bS 16 c7+ �cS 1 7 lOb6+ �c7 I S �dS+ �bS 19 .c7+ �aS 20 h3? [20 .i.bS! axbS 2 1 0-0 wins] 20 ...• gS! 2 1 eb6llxe3+ 22 .xe3 .xdS 23 hxg4 [23 .eS+ �a7 24 .xiS .eS+ forces a perpet­U!lI] 23 .. . lOffi 24 .i.e2 .i.cs 2S Wd3 .e6 26 .f5 .d6 27 IIh3 lieS with a powerful attack on the dark squares surrounding the uncastled white king) 12 . . . axbS 1 3 lOe3 �S? ( 1 3 ... dS 14 .i.d3 .eS is better) 14 a4! bxc6 I S .xc6 lOe7 16 .a6+ �d7 17 .i.xbS+ c6 I S �7+ and Black is quite lost, MalmstrOm-Van Gameren, corr. 1999.

b) and 1O •.•• f7?! 1 1 dS leaves the knight without a square, e.g. l l . . .lCxiS 12 lOe3.

c) 10 . . . �dS?! I I dS a6 1 2 �c7 �c7 1 3 dxc6 is also very strong.

1 1 d5 lOce7 Il M 1Of6 As Ruggeri Laderchi pointed out

to me after our game, 1 2 . . . a6?! is worse, 13 lOa7! ( 1 3 �c7+ �d7 14 .i.d3 �S I S .d4 �c7 16 �g4 .xdS and at least Black is not mate­rial down) 1 3 . . . IlaS 14 .xc7! llxa7 I S .i.d3 �S 16 �S+ lOcs 1 7 .i.e2! and White regains his piece, with interest.

13 .i.d3 "'S 14 1OJ:c7+! �d8 IS 1OJ:14 11s:c7 16 1OJ:f6 "'4+ 17 13 lIJ:u

Following 1 7 ...• xffi I S .xf6 gxf6 19 c4 the endgame is hopeless for Black, Gentinetta-Rosso, corr. 1995.

18 pM 1IJ:d3 19 cJ:d3 Id6 20 1I11ll}J:dS 21 .i.d2

Black has no compensation for the exchange, Kosten - Ruggeri Laderchi, email 200 I .

C3 9 dS

9 ••• 1Ob4 9 . . . lOeS?! 10 �eS dxeS I I .i.d3

.ffi 12 .g3 .i.d6 13 lOe4 .n 14 IIfl lOffi I S .i.h6!, undermining the black knight, with a crushing

Main line: 3 �e5 .,6Inlroduclion and 7 J3 19

position, Noomen-Simmelink, email 1999.

IOlDeJ 10 lDbS? is not as ridiculous as it

appears, 10 . . . lDxc2+ 1 1 Wf2 lDxal? meets the rejoinder 1 2 .i.d3! .ffi 13 lDxc7+ Wd8? 14 .i.gS! .xgS I S .xf8+ Wxc7 16 .xd6 mate! But Black has the interposition I I . . ..i.g4! and should win.

IO •• lllf6 Black has tried finding another

move here to avoid problems, but each move has its drawbacks, i.e. 10 ... .i.e7 1 1 .i.bS+ Wd8 1 2 0-0 lDffi 1 3 .e2! (making a3 a serious men­ace, the b4-knight is very exposed) 13 . . . c6?! 14 dxc6 bxc6 I S a3 ! lDxc2 16 .i.xc6 lDxal 1 7 �S 1 -0, Downey-Sireta, corr. 1994/9S.

1 183 a) 1 1 .i.bS+!? is unusual, but

good: I I ....i.d7 1 2 .i.xd7+ lDxd7 ( 1 2 . . . Wxd7?! has been played, but apart from 13 �+ Wd8 14 0-0 with some advantage, Svendsen­Spiegel, corr. 1991 , 1 3 .e2, threat­ening �S+, 1 3 . . . lDa6 14 �S+ Wc8 I S 0-0 is simple, and strong) 13 .i.d2?! ( 1 3 O-O! is better, 1 3 . . . 0-0-0 14 .12 and I S a3 with a significant plus) when, in the game Dutreeuw-Hector, Geneva 1990, the Black position became very uncom­fortable following 1 3 . . . .i.e7? 14

.c4! .xc4 IS lDxc4 lDa6 16 0-0 lDffi 1 7 lDgS Wd7 1 8 Ilae I c6 19 lDf5 .i.f8 20 dxc6+ bxc6 2 1 lOe6 lDds 22 c4 with advantage. I think that Black should take the bull by the horns: 13 . . . lDxc2+! 14 lDxc2 .xc2 I S llcl .xb2 16 Ilbl .c2 17 .e2+ .i.e7 1 8 0-0 (not 1 8 Ilxb7? when 1 8 . . . 0-0! wins) with compen­sation for the two pawns.

b) 1 1 .e2!? is untried, and yet threatens both �S+ and a discov­ered attack with the knight; 1 1 . . . .i.d7 seems best, 1 2 a3 lDa6 13 .0 with some advantage.

1 I ••• lDa6 Forced, I l . . .lDxc2+? is now

known to lose: 12 lDxc2 .xc2 13 .i.d3 �3 14 lilc4! (not 14 O-O?! which gives Black time to recover, 14 . . . .i.g4 I S .e3+ Wd8 16 .gS h6 17 Wh4 .i.e7, Gaard-Malmstrom, corr. 1991 , White's compensation is not quite sufficient) 14 . . . Wc7 (this loses, but so does everything else, 14 . . .• a4 I S lDxffi+ gxffi 16 .xffi .as+ [if 16 . . . llg8 17 0-0 wins on the spot, i.e. 1 7 . . .• d7 1 8 .i.gS 1-0, Clarke-Logunov, corr. 1994/9S] 1 7 .i.d2 .xdS 1 8 O-O! .xd3 19 .t7+ Wd8 20 .i.gS+ .i.e7 2 1 .xe7 mate, Heap-Jackson, corr. I 994I9S, 14 . . . .i.e7?? is quickest of all, I S lDxffi+ 1-0, Lane-Valverde Lopez, corr. I99S, as I S . . . .i.xffi 16 .i.g6+

20 Main line: 3 ltlKe5 W/6 Introduction and 7 J3

wins the exposed black queen, and 14 . . . �c4?? I S Wxc4+ is also hope­less I S .. .'�t7 16 ()-()+ ci'gS 17 WO) I S .i.gS Wxb2 ( I S ...• xdS 16 .i.xf6+ gxf6 17 Wxf6+ ci'd7 [ 1 7 . . . ci'eS I S 0-0-0 .eS 19 �d6+ .i.xd6 20 Ilhel 1 -0, Trani-Rebaudo, COrT. 1999] I S .xhS Wxd3 19 llks+ dxcS 20 lid I winning the queen, Downey-Rosso, COrT. 1999) 16 0-0 Wd4+ ( l6 .. . h6 1 7 lbxf6 hxgS I S :ae I + 1 -0 Svendsen-Browning COrT. 1991) 1 7 ci'hl .xdS I S .i.xf6+ 1 -0, Ruggeri Laderchi­Elburg, COrT. training game 1995.

12 .i.d3

• �.tJIi'OO bE '<Ii'1tV� � r�t :� ?� � f� ... ��./.: f�} �.IL� ... � r.'�, :.- 1iiiJ � .� \w� � �, �.� i%J, �..!J.� .. �; w.(.�?: W�� �?�:t. ?!:@/; � �J r� � fit1 � � ��� ..... .' ;.' � ... %"@.�

'/�;; � ""'>,, WiW � ;� �ni Jl w.rI *� l1. �g. ,�" ( ; � �,£ � cS ? ,\y; >#%; e

12 •••• h5 White maintains the advantage in

the endgame after this, the most popular choice at the present time. But 1 2 . ..• t7?! is worse, 1 3 b4! (to keep the black knight offside, al­though 13 0-0 will probably amount to the same, as 1 3 . . . llks is coun­tered by 14 .i.bS+ .i.d7 I S b4! a6 16 bxcS axbS 1 7 c6 bxc6 I S dxc6 .i.e6 19 lbxbS ci'dS 20 lbd4 which is very unpleasant for Black, Kruijer­Schoesser, COrT. 1995) 1 3 . . . .i.e7· ( 1 3 ... .i.d7 14 0-0 0-0-0 I S .i.xa6 bxa6 16 .e2 .eS 1 7 .xa6+ ci'bS I S lbc4 with advantage, Gaard­Nyman, COrT. 1991) 14 0-0 0-0 IS lbf5 (currently favoured, although I S .i.b2 would be my preference,

I S . . . .i.d7 16 :ael [ 16 lbc4!?] 16 . . . Whs 1 7 Wg3 with a very aggressive set-up, Svendsen-Heap, COrT. 1991) I S . . . .i.dS!? (to conserve both bishops, I S . . . .i.xf5 16 .i.xf5 ci'hS 17 .i.gS llaeS I S :ae I .i.dS 19 IlxeS llxeS 20 Wh3 g6 21 .i.e6 with considerable advantage, Clarke­Gaard, COrT. 1990) 16 .i.h6!? (aim­ing for a direct refutation, 16 .i.gS is also good, with the point that 16 .. . h6? loses to 1 7 .i.xh6 1-0, Clarke-Borrmann, COrT. 1992, and 16 .i.b2 likewise) 16 . . . Whs 17 J.xg7! WxO I S IlxO .i.xf5 19 .i.xfS .i.xd3 20 cxd3 �fS 2 1 lbc4 ci'g7 22 :an lbxdS?! (22 . . . �c4 23 dxc4 .i.gS is a better chance) 23 1117+ ci'g6 24 IlfS IlbS? (24 ... ci'h6 had to be tried) 2S IlgS+ ci'h6 26 110 and the threat of mate forces the gain of material, Svendsen­Stummer, COrT. 1994/9S.

13 Wxh5+ lbxh5 14 0-0 14 b4, immediately confining the

a6-knight to the edge, 14 . . . .i.e7 (14 . . . g6!? is an alternative, I S 0-0 .i.g7 16 .i.d2 lbf6 1 7 lbbS 0-0 offers Black a playable position, senechaud-Budovskis, COrT. 19941 9S) I S 0-0 transposes.

14 ... .i.e7 Trying to bring the knight back

into the game by 14 .. . lbcS?! is most simply answered by I S lbbS, forcing the retreat IS ... lOa6 when Black's position has only worsened, Heap-Elburg, COrT. 1994197, as I S . . . ci'dS loses a pawn to 16 .i.xh7. The simple 14 . . . ltlf6?! is also possi­ble, I S b4 .i.e7 16 J.b2 0-0 1 7 Ilae I lbd7 ( 1 7 . . . .i.d7?! I S lbf5 .i.dS?! 19 �g7! winning, Ruggeri Laderchi­Trani, COrT. 1999, as 19 . . . �g7 20 lbc4 threatens the almost unstoppa­ble 2 1 �f6 .i.xf6 22 lle7+) I S IlxfS+ .i.xfS 19 lbc4 h6 20 c4 and White enjoys a very consequent

Main line: 3 ltixe5 ./6 Introduction and 7 f3 2 J

space advantage, Simmelink-Bartsch, COlT. 1999.

IS b4 Once again the key to White's

play, the a6-knight is shut out. IS •. �f6 I S . . . �a;?! transposes into the pre­

vious note. 16 .i.d2

i� �. � }B <,.;;c.'.

t {� "�." ." < 7::. '� t 1.It �.aj � �.aj .... ''%.0W' 'r ..... '".:eN '� '� .. �f% r£:� � � � � � fj::j: 'i2%' ...... � � .Ll. � � ... r� � .. t�.·� � r.!W � tff.0 �

� ';R�/f!,.;� � .. ' �$.b" � :�� �{ � � ;� � .Ll. � �J: .Ll. fJ],;.

§ � �§W

16 •.• .i.d4 The bishop seeks greener pas­

tures, and anyway, 16 . . . 0-o? 1 7 g4 forces Black to exchange this one active piece, 1 7 . . . .i.xc3 ( 1 7 . . . lleS IS gxhS llxe3 [ unfortunately, IS . . . .i.xc3 19 .i.xc3 Ilxe3 20 Ilae I Ilxel 21 Ilxel , intending 22 lle7, is also hopeless for Black] 19 .i.xe3 .i.xc3 20 Ilab I gives Black precisely zero compensation for the exchange, Ruggeri Laderchi­Malmstrom, COlT. 1997) IS IlxfS+ (IS .i.xc3? �f4 is less convincing, Svendsen-SCnechaud, COlT. 1990) IS . . . Wxffi 19 .i.xc3 �a; 20 Ilfl �e7 2 1 gS winning material.

17 �hl ' 17 Ilael �a; IS �bS (IS Ilf4

.i.b6 19 .i.bS+?! .i.d7 20 .i.xd7+ Wxd7 2 1 g4 Ilhffi 22 �g2 offers White a slight edge, Svendsen­Niemand, COlT. 1994/96, and IS �hl transposes) I S . . . .i.b6 19 a4 0-0 20 as .i.xe3+ 2 1 llxe3 .i.d7 22 c4 �g4! heading for eS, Black is only

slightly worse, Stockholm Vasa Chess Club - Riga Seniors Chess Club, COlT. 1934/36.

17 ... �ffi 18 Ilael 0-0 19 �f5 .i.xf5 20 .i.xf5 �b8 21 .i.e6

White -has a slight advantage, Sakellarakis-GHrd, email I99S.

C4 9 .i.d3

The most logical, not being afraid to give up the d-pawn, if need be.

9 •••• 14 Black attempts to cover his devel­

opment problems by entering the endgame.

9 . . .• e6+!? is possibly not as bad as I originally lil,ought, 10 .i.e3 ( 10 �f2?! leads noWhere, 10 .. . �d4 I 1 lMs �xO 12 �c7+ �t7 1 3 �e6 .i.xe6 14 WxO IlcS I 5 �3 level, Riga-Trani, COlT. 1997) 1O .. . �d4 1 I .f4 lOc6 1 2 0-0 �a; ( l2 . . . �ge7!? is interesting: 13 1lael .g4 14 .t7+ �dS IS Ilf4 .i.e6! 16 �xd6! cxd6 17 .xffi+ Ilxffi IS Ilxffi+ �d7 19 IlxaS �S!? which is messy, E1bUIJ-Sveinsson, COlT. 2(00) 13 Ilael .i.e7 ( l 3 . . . IOdS?? is a blunder, 14 �dS .xdS IS .i.e4! �dS (pointless, but 1 5 ...... 5 allows 16 .i.xc6+ bxc6 1 7 .xf8+ ! and mate in 6] 16.i.xdS 1 -0, Dzervenis­Zagata, COlT. 1993) 14 .i.d4 .t7 ( l4 . . .• 8S? IS �d6+! cxd6 16 .xd6 Wt7 17 ha; gxa; IS IOdS

22 Main line: 3 ltlxe5 "/6 Introduction and 7 f3

I'roved decisive in Meyers-Crowl, /\ustralia 1930) 1 5 .i.xf6 gxf6 16 • xf6 "xf6 17 lhf6 �5! 1 8 lOxe5 .i.xf6 19 lbg6+ Wd8 20 lOxh8 .i.xh8 2 1 .i.xh7 and Black's power­ful dark-squared bishop is not enough compensation for the two passed pawns, Melchor-Crimp, corr. 1995.

Now White has a choice between the endgame, and two more compli­cated lines:

C41 10"xg4 C42 10"1'2 C43 10"e3+

C41 10"xg4 White can gain a small edge after

the exchange of queens: 10 ••• .i.xg4

nlbe3 Best, other moves led to a small

advantage for Black in both Alberts­Strautins, corr. 1969, after I I .i.e3?! lbb4! 1 2 .i.e4 lbf6 13 a3 lbxe4 14 axb4 lbxc3 15 bxc3, and in Schmidt-Hempel, Germany 1969, which went I I O-O?! lbb4! (White's lead in development means little in this ending, whereas possession of the two bishops is important) 1 2 .i.e4 (White could have tried to muddy the waters a little with 1 2

lbb5!?) 1 2 . . . lbf6 13 .i.xb7? ):tb8 14 .i.f3 .i.xf3 15 lhf3 lOxc2 16 ):tb I lbxd4 + .

1 1 •• id7 12 lbeds 0-0-0 13 lObs lbf6 14lbbxc7 l:te8+!? IS Wfl

15 lbxe8? lOxd5 traps the e8-knight.

Is ••• lOxdS 16 lilxdS lbxd4 17 .i.e3lbc6 18 ):the I

. .. with an edge, because of the isolated d6-pawn, Scarani-Rosso, corr. I998.

C42 10"1'2

10 ..... xd4 Taking the d4 pawn whilst at the

same time continuing to offer the exchange of queens. Apart from its basic numerical value, the gain of this pawn is also strategically desir­able for Black as he no longer has to worry about White pushing dS, and he can now use the e5 square. Nev­ertheless, if the note to White's next move proves convincing Black may have to resort to 1O . . . lbf6 which is tried occasionally: I I 0-0 (1 1 h3?! .e6+ 12 .i.e3 d5?! [ 1 2 ... lbb4!] 1 3 lbb5 .i.b4+ 14 c3 .e7 1 5 lbd2? [ 1 5 �5! is clearly better] 1 5 . . . a6 16 lOxc7+ .xc7 17 cxb4 lOxb4 18 .i.bl [ 1 8 .e2 lOxd3+ 19 .xd3 .g3+ is also terrible] 1 8 ...• c l+ 19 We2 .xb2 20 a3 0-0 + Yudovich­Elzov, Moscow 194 1 ) 1 l . . ..i.e7

Main line: 3 lOre5 ./6 Introduction and 7 J3 23

(1 I . . .lLlb4 is more combative) 1 2 lLlbS ( 12 h3?! �S 1 3 .i.e3 0-0 14 .g3 is level, GHrd-Muller, COlT. 1988) 1 2 ... 0-0 13 lLle3 ( 1 3 lLlxc7?! lLldS) 1 3 ...• d7 14 lLlfS lLlb4 (14 . . . .i.d8 I S .g3 lLlb4 might be more accurate) I S .g3 lObS? ( I S . . . .i.d8 16 �c4+ �h8 1 7 �3 lLlc6 is fine for Black) and now, as well as the 16 lLlxe7+ .xe7 17 lIxfB+ .xiS 18 WM, with advantage, of Grivainis-Eglitis, COlT. I 977n9, White also had 16 .el ! lbfS?! 1 7 .*.xfS .xbS 1 8 .i.xc8 winning. Incidentally, at the end of this game White makes the most awful blunder, that I can only explain in one of two ways: a) it was a slip of the pen or b) he had the wrong position set up on his board!

Otherwise 1O . . . .*.e7?! allows 1 I lLlds, Medina-Delara, Caracas 1954.

I I .i.e3 The most common, by far, but not

the best. 1 1 .e2+! creates more problems,

I l . . .lLleS ( 1 1 .. ..i.e7? 12 lLlbS) 12 lLlbS! (if 12 lLlxeS then going for the endgame by 12 ...• xeS is simplest, e.g. 1 3 lLlds .xe2+ 14 .i.xe2 �d8 I S 0-0 lLlf6 [ I S . . . i.e7!? 16 lLlxe7 lLlxe7 17 1117 .i.e6 18 lIxg7 lLlf5 is also feasible] 16 .i.gS h6! , as 12 . . . dxeS? is too risky, 1 3 1Whs+ �e7 14 .i.gS+ [ 14 .gS+! lLlf6 I S .i.e3 .d6 16 0-0-0 is really nasty] 14 . . . lLlffi I S o-o-o!? [ I S lIf1 ] I S . . . .i.g4 16 lLlds+ �d6 1 7 lLlxffi?! [whilst amusing, this is hardly cor­rect! 17 .f7! lLlxdS 18 .i.fS must be better for White] 1 7 . . . .i.xhS 1 8 lLlxhS �c6 19 lIhel .a4! , good for Black, Frolik-Griinfeld, Torremol­inos 1986) 1 2 . . . �4+ 1 3 g3 .e7 14 lLlxeS dxeS IS i.gS lLlffi ( I S ...• xgS!? 16 lLlxc7+ �f7 17 lLlxa8 It)ffi might be a better

practical try) 16 0-0-0 (threatening 17 .i.xffi gxffi 18 �S+) 16 ... .i.g4 (going into a difficult endgame, but there is nothing better) 1 7 hffi .i.xe2 1 8 .i.xe7 .i.xdl 19 .i.xiS .i.xc2 20 .i.xc2 WxiS 2 1 lie I c6 22 lLld6 b6 23 lbeS, White is clearly better, Zaniratti-Tatlow, COlT. 1999.

1 I •••• f6

a� _ �t�� � �i;1t ���� . '�

� � � � ��� � � � � �

�� � � � � � ��� � .ft �ft1.ft � � .ft �:ffi § � 0'" �§

Il .gl The other possibilities: a) 1 2 .e2 also avoids the ending

but 12 . . . .i.e6 gives Black good play: a l ) 1 3 lLle4?! WM+! (better than

13 . . .• e7 14 .i.gS .d7 [ I4 . . . lLlffi? I S 0-0 .i.xc4 16 .i.xc4? .xe4 1 7 .xe4+ lLlxe4 18 :ael with enough play for a draw, Grivainis-Hempel, COlT. 1970n3, �t White missed 16 lbf6! .i.xd3 17 cxd3 which wins material] I S 0-0 .i.e7 16 .i.xe7 .xe7?! [ 16 . . . lLlgxe7!? 17 Whs+ �d8 keeps the game alive] 17 :ael .i.xc4 1 8 �S+! g6 19 lLlffi+ �d8? [ 1 9 . . . lLlxffi 20 lbe7+ lLlxe7 offers some drawill8 chances] 20 lbe7 gxhS? 2 1 lIe8 mate, Grivainis­Gunderam, COlT. 1970) 14 .i.f2 (White wants to keep the queens on, alternatively: 14 lLlg3 .g4 I S .xg4 [ I S .f2? .i.xc4] I S .. . .i.xg4 16 0-0 �ffi [ 16 . . . dS! is even better] 1 7 h3 .i.d7 1 8 .i.gS .i.e7, White doesn't have enough for a pawn, Grivainis­Budovskis, COlT. 1977n9, or 14 g3

24 Main line: 3 llhe5 "/6 Introduction and 7 j3

"g4 IS "f2? [ I S li)�S "xe2+ 16 *xe2 �g4+ 17 �f2 l£lffi does not really give White much compensa­tion for the pawn. Zabaznov­Tatlow. email l99S] I S . . . .i.xc4! also Grivainis-Budovskis. when White tried a good bluff 16 h3 !? as he saw that 16 �xc4? "xe4 1 7 "n+ �dS IS "xfS+ ci>d7 loses after both 19 "xaS "xe3+! and 19 "n+ li)ge7 20 �d2 IlafS. In the game Black could have won by 16 . . ... e6. but instead he played 16 . . ... g6? and White saved himself by 17 .i.xc4 "xe4 I S "n+ �dS 19 "xfS+ �d7 20 "xg7+ li)ge7 2 1 "g4+!. revealing the importance of the h3 move.) 14 ...• g4 IS "e3 !? dS ( I S . . . O-O-O!? 16 h3 ! "xg2 1 7 lOexd6+ .i.xd6 IS "xe6+ �bS 19 :g l is unclear) 16 h3 d4 1 7 "d2!? ( 1 7 hxg4 dxe3 IS li)xe3 is about equal) 17 . . ... xg2 IS li)g3? ( I S 0-0-0 is stronger) Knudsen­Hayward. COIT. 19S6, I S . . . li)ffi with two extra pawns.

a2) 1 3 :fI "e7 14 0-0-0 ( 14 lOe4? .i.xc4 + Swafield-Dreibergs. COIT. I 96S) 14 . . . 0-0-0 ( 14 . . . dS? meets the rejoinder I S li)xdS ! .i.xdS 16 �cS) I S :de l dS 16 �cS! "gS+ ( l 6 . . ... xcS?! 17 "xe6+ �bS IS li)xdS!) 1 7 .i.e3 = . Zschom­Tiemann, COIT. 19S9/90.

a3) 1 3 0-0-0 0-0-0 14 lOe4?! ( 14 :hfl "e7 transposes into the previous note) 14 ...• e7 I S li)gS dS ! ( I S . . . li)ffi is not bad, either. 16 li)xe6 "xe6 17 :hfl [17 :he I?! and now, instead of the catastrophic 1 7 . . . �bS?? IS .i.xa7+ 1 -0. Hailey­Kennedy. email 1993. 1 7 . . . dS I S "f3 lOe4 is perfectly reasonable] 1 7 . . . �bS?? [the same blunder. better is 1 7 . . ... 17] I S .i.xa7+ 1-0. Niemand-Tiemann. COIT. 19S3 I S . . . .i.d7?! 16 J:thfl ?! [ 1 6 li)xh7] 16 . . . li)ffi 17 h3 h6 IS li)f3 dS Black

is back on top. Niemand-Tiemann. COIT. 19S3) 16 lDd2 d4 17 .i.f4 .i.xa2 with a clear Black advantage.

a4) 1 3 li)bS?! .n 14 .i.f2 �d7! (1ixe6+ was the threat) IS 0-0-0 :eS 16 :hfl li)ffi. threatening .. . .i.g4. and consolidating. Behrmann-Parzefall. COIT. 1996.

b) 12 lDdS?! "xf2+ 13 .i.xf2 �dS 14 .i.h4+ li)ge7 ( 14 . . . .i.e7 I S 0-0) I S 0-0 .i.e6 16 li)f4 .i.n 1 7 li)h3 .i.xc4 IS .i.xc4 lOeS?! ( l S . . . h6! is more accurate. when Black remains a sound pawn up) 19 li)gS ! c6 20 li)n+ li)xn 2 1 :xf7 :Cs 22 :el (22 .i.e6 looks interesting; if then 22 . . . :c7? 23 J:td l dS 24 .i.g3) 22 . . . :c7 Grivainis-Kozlov. COIT. 1975/9. when 23 .i.e6 keeps White on to�.

c) 1 2 O-O!? .xf2+ 13 :xf2 l£lffi ( 1 3 . . . .i.e6!? 14 :e2 �d7 IS li)bS �xc4 16 �xc4 lOes Diepstraten) 14 .i.gS .i.e7 IS .i.xf6 .i.xffi 16 lDds Snayer-Goedhart. COIT. 19S I . when 16 . . . 0-0 17 c3 ( 1 7 It)xc7?! .i.d4) 17 . . . :n is level.

12 •••• n! The most common. and clearly

best. although 1 2 . . . .i.e6? has also been seen. but without success: 13 lOe4 "e7 14 .i.gS .d7 ( l4 .. . li)ffi IS 0-0 li)xe4 16 .i.xe4 .d7 17 Ilae I [ 1 7 'fIh4 is also strong. 17 . . . .i.e7 { 1 7 . . . .i.xc4? IS .i.xc6! wiMing} I S .lg6+! �dS 19 .xh7! .eS

Main line: 3 �e5 "/6 Introduction and 7 j3 25

{ 19 . . . lle8 20 .i.xe8 "xe8 2 1 "xg7 with advantage} 20 .i.xe7+ 1 -0, Winkelmann-Schlenker, COrT. 1989, as 20 . . . �xe7 2 1 .i.xe8 llxh7 22 1lf8 c6 23 Ilel .i.xc4 24 .i.xc6+ Wc7 25 llxe7+ Wxc6 26 Ilxa8 is crushing] 17 . . . .i.e7 [ 1 7 . . . �7 1 8 .i.xb7 Ilb8 19 �S is no improvement, GUrd­Hayward, COrT. 1990) 18 .i.xc6 bxc6 19 .i.xe7 "xe7 20 2:lxe6 "xe6 21 lie I wins, Pape-Jackson, COrT. 1994/95) I S 0-0 dS (this loses, but the others are even worse: IS . . . �ge7? 16 �xd6+! cxd6 17 �d6+ Wd8 1 8 Ilad I is over­whelming, I S .. . .i.xc4? 16 .i.xc4 �S 17 Ilae I wins, Tiemann­Fiorito, COrT. 1980, but misses the elegant 1 7 �S! "c6 1 8 "xeS+! dxeS 19 .i.f1 mate, IS . . . .i.e7? 16 .i.xe7 "xe7 1 7 Ilael and the threat of �gS is decisive, and finally, I S . . . h6? Becker-Gonschior, COrT. 1979, allows a couple of immediate wins, for example: 16 �S dxcS 1 7 .i.g6+) 16 IlxfS+! WxfS 1 7 Ilfl + .i.f1 1 8 �S "e8 19 "xc7 �f6 20 .i.xf6 decimating the black position, Tiemann-Grivainis, COrT. 1985.

IJ � The knight covers the gS-square

to deter Black from playing . . . .i.e6, 13 0-0-0 .i.e6 gives White a prob­lem with his c4-knight, 14 �! ( 14 �?! i s too acquiescent, 14 . . . 0-0-0 [ I4 . . . �f6!) IS Ilhfl �f6 16 �e4 ICJxe4 17 li:)xe4 �S White doesn't have quite enough for the pawn, Terblanche-Doyle, COrT. 1994) 14 . . . 0-0-0?! ( 14 . . . .i.xc4 should be tried, I S �gS .dS 16 .i.e2 [ 16 .i.g6+? is too generous, 16 . . . hxg6 1 7 IlxdS .i.xdS) 16 . . ... eS 1 7 .i.xc4 "xg3 1 8 hxg3 �S, about level) I S Ilhfl �f6?? (blundering a piece, I S . . ... d7 is OK) 16 �gS "d7 17 �xe6 winning immediately, Pape-

Evans, COrT. 1994/95, as 1 7 .. ... xe6? drops the queen to 1 8 .i.f5.

1J ••• .i.e7 1 3 . . . �b4!? is a good plan, the

white light-squared bishop is an im­portant piece, so Black liquidates it, 14 �8� ( 14 Ilfl !?) 14 .. .ti�xd3+ I S cxd3 .g6 16 0-0 h6 17 � ( 17 Ilae I .i.e7 also forces the exchange of queens) 1 7 . . ... xg3 1 8 �g3 Reinke-Stummer, COrT. 1994/95, and Black is better.

14 .i.gS

14 ••• dS The sharpest, although 14 ... li:)f6

turned out well in Svendsen-Elburg, COrT. 1990: I S 0-0 .i.d7 ( l S . . . �e4? 16 .i.xe4 "xc4 IY.i.xc6+ bxc6 1 8 Ilael .i.e6 19 .i.xe7 is unclear) 16 .i.xf6 gxf6 and now, 1 7 1lxf6 .dS 1 8 Ilafl 0-0-0 should be equal.

IS lln And certainly not I S "xc7??

.i.b4+ 0- 1 , Leisebein-Gaard, COrT. 1992, or I S o-o-o? dxe4 0-1 Jensen­Melchor, COrT. 1989190.

IS ••• �r6! This seems to win 'by force', so

there is little need to bother with the old line I S . . ... e6 16 0-0-0 (rather than 16 �3?! dxe4 1 7 .i.c4 .d6 1 8 .i.f7+ Wf8 19 .i.b3+ �f6 defend­ing, Krantz-Downey, COrT. 1990) 16 ... dxe4 17 .i.xe7 .xe7 1 8 .i.xe4

26 Main line: 3 lfu-e5 "/6 Introduction and 7 J3

.i.e6 19 .i.xc6+ bxc6 20 :de I ! 0-0-0 2 1 ltles which is tricky for Black, Weber-Doyle, corr. 1993.

16 .i.d6 pf6 17 lbf6 .C8 18 .xc7 dxe4 19 Zlxc6 exd3 20 fDd6+ .i.xd6 21 Zlxd6 .g5!

White can't be allowed to castle. 22 Zlxd3 Zl18 23 Jladl .i.C� 24

.13 Wf7! 25 Zle3 ZlaeS 26 Zldd3 Zlxe3+ 27 Zlxe3 Zld8 0-1 Melchor­Downey, corr. 1994196.

C43 10"e3+

I am not so sure that this move is White's best any more, it rather de­pends on the outcome of I S . . .• xc8, in the main line.

10 •• ..i.e7

1 1 0-0 As it appears that Black cannot

take the d4 pawn, this would cer­tainly seem to be the best choice here, but White does have other options:

a) 1 1 d5 1Ob4 ( l 1 . . . .. xg2!? 1 2 .i.e4 .i.h4+ 1 3 Wdl .i.g4+ 14 .i.f3+ IOge7) 12 0-0 lOxd3?! (it is prob­able that 1 2 . . . 1Oft) is the best, trans­posing to the main game) 13 .xd3 .i.d7 (to castle queenside, but it is quite risky; the unplayed 1 3 . . . IOft) is almost certainly better) 14 lOaS! is tricky: ( 1 4 .i.d2 O-O-O?! 1 5 Zlf4

�S? [ I S . . .• g6!J 16 .d4 b6 Heap­Sawyer, corr. 1988, and now 17 .xg7 .g6 1 8 .xg6 bx.g6 with ad­vantage to White) 14 . . . 0-0-0?! ( 14 . . . .i.ft)?! is met by 1 5 lObs with advantage, Selart-Menta, corr. 1972174, but Black can try 14 .. . 1Oft) answering 1 5 lObS?! with I S . . . O-O! 16 lOxc7 Zlac8, and I S lOxb1? with I S .. . � 16 .a6 � when a white knight is trapped; unfortu­nately, I S :f4 �S 1 6 lOxb7 wins a pawn without fuss) I S .i.e3 Wb8? 16 Zlf4 �S 1 7 Zlb4 1-0, Morgado­Priede, corr. 1970173.

b) 1 1 lObS? .xg2! 1 2 Zlgl .xh2 1 3 IOxc7+ (White might as well take this, neither 1 3 c3 Wd8 14 Zlxg7 hS I S .12 .xt2+ 16 Wxf2 a6, lIyin Genevsky - Betin!, corr. 192 1 /23, nor 1 3 .i.d2 �+ 14 Zlg3 Wd8 I S .f3 IOh6, Neff-Dreibergs, corr. 1964, hold out much hope for White) 1 3 .. . Wd8 14 lOxa8 .i.h4+ I S Wd I 10ft) 16 Zlxg7 Zle8 (but 16 . . . .i.g4+! 1 7 Zlxg4 1llx.g4 is crush­ing) 1 7 "gl Zlel+ 1 8 .xel .i.xel 19 Wxe I 'tIb4+ with a powerful attack, Stebahne-Brusila, 1986.

c) 1 1 .i.e4? 10ft) 12 .i.xc6+ bxc6 1 3 0-0 .i.a6 14 Zlf4 ( 14 Zlel 0-0 I S b3 Zlae8 was also much to Black'!$> likin� in Purins-Tomson, corr.) 14 . . .• d7 I S b3 0-0 16 .i.d2 lOd5 17 ZlxfS+ ZlxfS 18 .e4 lOb6 19 IOxb6 axb6 20 dS? .i.ft) 2 1 .a4 (2 1 dxc6 .17, threatening some nasty bishop moves) 2 1 .. ..f5 0- 1 . This looks a little premature, but 22 .xa6 .i.d4+ (or 22 . . .• xc2) 23 Whl bS 24 h3 .xc2 should certainly win for Black, Bemsdorfer-Hazenfuss, corr. 1937138 .

d) 1 1 lOdS? blunders a pawn, I I . . ..xg2 12 .e4 .xe4+ 1 3 .i.xe4 .i.h4+ 14 Wd2 Wd8, Nobbe-Hage, corr. 1982.

Main line: 3 ltlxe5 ./6 Introduction and 7 J3 27

11 ••• lOf6 1 1 . . . 1Ob4?! is a new try, but will

soon be discarded, no doubt, 1 2 Ad2 lOxd3 1 3 lOds! ( 1 3 .xd3 Ad7 14 :ael 0+0 I S lOds Af6 16 1lc4 .g6 17 lOaS! c6 1 8 �3 is not bad either, Villar-Profeta, corr. 1999) 1 3 . . . �d8 14 .xd3 1Of6 I S lOxe7 Wxe7 16 :ael+ �d8 1 7 lOeS! dxeS 1 8 dxeS+ Ad7 19 exf6 gxf6 20 1lxf6, winning quickly, Downey­Destrebecq, corr. 1997.

The ending that follows I I . . ..xd4?! 1 2 lObS! .xe3+ 1 3 lOxe3 is not favourable to Black, by any means, 1 3 . . . �d8 14 lOds 1Of6 I S lObxc7 lOxdS! 16 lOxdS GrivainislHayward-Budovskis, corr. 1997.

Il d5 1 2 lOb5 �d8 1 3 c3 IlfS 14 h3

1ih5 is fine for Black, Doplmayr­Bonte, corr. I99S

1 2 J:tr4!? .e6 13 lOb5 ( 1 3 .a can be met by 1 3 . . . 1Ob4) 1 3 . . . lOds ( I 3 ...• xe3+ 14 Axe3!? lOds I S Ae4! lOxf4 16 lOxc7+ �d8 1 7 lOxa8 lOe2+ 1 8 �f2 lOexd4 19 Axd4 lOxd4 20 lid I IlfS+ 2 1 �e3 lOfS+ led to equality in Pape­Reinke, corr. 1 994/95) 14 .xe6 Axe6 1 5 1lc4 Afl is very comfort­able for Black, Decker-Beutel, corr. 1997.

Il ••• lOb4 13 Ilf4

1l ...• d7 At the present moment in time,

the text seems adequate, however the alternative 1 3 . . . 1ObxdS!? has now received some attention, 14 Ilxg4! ( 14 lOxd5 lOxdS 1 5 Ilxg4 lOxe3 1 6 lOxe3 Axg4 1 7 lOxg4 0-0, and an endgame is reached where White has bishop and knight for rook and pawn. Black should strive to exchange a pair of rooks and the dark squared bishops, bring his king to the centre, and mobilize his queenside pawns. I would evaluate this as ;t, but not without practical prospects for Black, 1 8 -*.d2 d5 19 Ilfl?! Ac5+ 20 Ae3 Axe3+ 2 1 lOxe3 Ilxfl + 22 Wxfl c6 and Black eventually managaUo draw, Holt­Major, corr. 1995) 14 . . . lOxe3 1 5 Ilxg7 (+1- NCO), but 1 5 . . . lOxc4! ( I S . . . lOeg4?! 16 AgS Wd8? 1 7 h3 [ 1 7 llxe7! Wxe7 1 8 lOds+ �fl 19 Ilfl i s even simpler] 1 7 . . . h6 1 8 Ah4 with a decisive advantage, Downey­Melchor, corr. 1 994/96) 16 Axc4 lOhS, curiously nearly trapping the rook! 17 Ah6! (if 1 7 Ilfl c6 and . . . dS) 1 7 . . . lOxg7 1 8 Axg7 IlfS 19 AxfS AxfS (after the forcing play, Black is behind in development, but enjoys a useful pair of bishops) 20 lOds (otherwise, 20 llfl c6! 2 1 Ad3 Ag7 22 Axh7 Ae6, or 20 lOb5 �d8 2 1 Ilfl Ae7 22 Ilfl dS 23 Ad3

28 Main line: 3 �e5 .,6 Introduction and 7 J3

.*.cS+ 24 ci'ft c6 and Black has chances, 25 1Oc3 h6 26 Ilh7 .*.e3 etc.) 20 .. . ci'dS 2 1 Ilft Ag7 22 1lf7 Ad4+ 23 ci'ft c6 24 liJc7 IlbS 25 Ae6 Axb2 26 AxcS IlxcS 27 1Oe6+ ci'eS 2S Ilxb7 AeS 29 llxa7 IlbS and Black eventually managed to grovel a draw, Pepe-Elburg, corr. 1999.

14 Af5 After 14 lDb6 axb6 I S Ilxb4 0-0

( l S .. . lDg4! looks interesting) 16 .*.d2, Castelli-Hempel, corr. 19701 7 1 , Black is a little worse, but no more.

Otherwise. the tactical continua­tion 14 lDxd6+?! .xd6 I S lDbS is not quite correct, I S ...• cS 16 d6 lDxd3 1 7 lDxc7+ ci'dS I S lDxaS .*.xd6 19 cxd3 ltXIs 20 .xcS .*.xcS+ 2 1 ci'h I .*.e6! 22 Ilc4 .*.d6 23 1la4 AbS!?, definitively trapping the white knight, 24 b3 Ad7 25 AgS+ ci'cs 26 1[e4 Ilm 27 Ae7 IlI7 2S Ilc l + Ac6 and, inexorably, Black rounded up the knight, and won, Pape-Krantz, corr. 1994/95.

14 •••• d8 The queen sacrifice 14 ...• xf5? is

no longer considered Black's best chance, I S 1lxf5 Axf5 16 .f4! ( 16 lDbS?! is a sUP, 16 . . . 0-0! 1 7 lDxc7 [and not 1 7 "xe?? llaeS I S .xc7 Ile 1+ 19 ci'f2 lDxc2 20 lDbxd6 Ag6 handing Black a powerful attack.]

and instead of 1 7 . . . lDbxdS? Rublevsky-Malyutin, Russia 1992, Black missed 17 . . . lDxc2! I S .xe7 llaeS! 19 lDxeS llxeS with equality) 16 . . . Ag6 17 lDe3 lDa6 I S lDf5 Axf5 19 .xf5 0-0 20 Ae3 main­tains a large advantage, Trim­Vaughan, CSA Ch 1993.

IS Axc8

IS •••• xc8! The best, otherwise, we have the

following possibilities: a) I S . . . lOxc2? 16 .d3 lDxal 1 7

Axb7 0-0 ( l 7 . . . llbS I S .*.c6+ ltXI7 19 1ib I with advantage) I S AxaS .xaS 19 lDe3 c6 20 dxc6 .xc6 2 1 1ib I , winning the stranded knight, Gumhill-Monciunskas, Brighouse 1966.

b) I S . . . llxcS 16 .xa7! ( 16 .e6!? lDxc2 17 1lxf6! gxf6 I S lDe4 1lf8 19 lOcxd6+ cxd6 20 lDxd6+ .xd6 21 .xcS+ ci'f7 22 .xc2 .cS+ only leads to a draw, Downey-Reinke, corr. 1994196) 16 . . . lDbxdS 17 lDxdS lDxdS I S 1lf5! lDb4 ( l S . . . lDb6 19 lDxb6 with advantage) 19 Ad2 .d7 20 Ilaft is better for White, Foulds­Ottenbreit, corr. 1997.

c) I S ... lDbxdS?? 16 lDxdS lDxdS fails to the shot 1 7 Axb7! 1 -0, both Melchor-Krantz, corr. 1994/95, and Downey-Vitols, corr. 1994/95 as 17 . . . lDxe3 I S .*.c6+ recovers the queen with interest.

Main line: 3 /t)xe5 ./6 Introduction and 7 f3 29

16 lD'Id6+ cxd6 17 :'1b4 This position was considered

(even by myself) to give White the advantage, however . . .

17 ••• 0-0!

Black moves over to the attack! 18 .i.d2 Alternatives: a) 1 8 .e6+ .xe6 19 dxe6 d5! 20

Lb7 (20 :b5 a6 2 1 lOxd5 might pose more problems) 20 . . . .i.c5+ 2 1 Cit>hl (2 1 WfI !?) now, instead of 2 1 . . .lDe4!? 22 ,3 :f1 + 23 Cit>g2 :gl + 24 Cit>h3 1llf2+ 25 Cit>h4 :b I 26 g4 .i.d6 27 .i.f4! .i.xf4 28 :xh I lDxh I 29 lOxd5 which is wild, Hunwick-Tatlow, COff. 1994, 2 1 . . .lDg4 is simpler, 22 :f7 lDf2+ 23 Cit>gl lDe4+ 24 Cit>f1 lDxc3 25 bxc3 :xf7+ 26 exf7+ �f7 with equality.

b) 1 8 h3?! .i.d8! the bishop finds a perfect diagonal ! 19 .e6+ .xe6 20 dxe6 .i.b6+ 2 1 Cit>h2 :ae8 22 -*.g5 Le6 23 -*.xf6 :fxf6 24 lDd5 :f2 25 lDxb6 :ee2 Black has no problems, Turley-Tatlow, COff. 1996.

c) 18 .xe7? .c5+ forks rook and king.

18 •• id8! 19 .e6+ .'1e6 20 d'le6 -*.b6+ 1I Cit>hl

2 1 Lb6!? axb6 22 :el might be a better try.

1I •• ':'e8 II :el

1l •• ':'1e6! 13 :'1e6 lDdS Simultaneously hitting the

b4-rook, and threatening a back rank mate.

24 :e8 lD'Ib4 2S :'118+ �18 26 lDbS lD'Icl 27 lD'Id6 -*.d4 28 b3 b6 29 lDbS 86 30 1Oc7 Cit>e7 31 lD'I86 Cit>d6 32 lDb4 -*.e3 Ih-Ih

Arias-Miguel, COff. 1995, the active black tang holds the balance.

To sum up, the main line with 8 . . . 1Dc6 seems fine for Black if White plays 10 .e3+, and also after 10 .f2 Wxd4 1 I -*.e3, but in this last line 10 .e2+ is rather more worrying. Furthermore, both 9 lDb5! and also 9 dS are also disturbing for Black.

Perhaps Black should prefer 8 . . . lDf6, currently bolstered by Bucker's ideas, or even 8 .. . -*.e7.

2 Main line: other seventh moves

In this chapter we look at White's other possibilities following I e4 e5 1 lOo f5 3 lOxes .f6 4 d4 d6 S ltX4 fxe4 6 lLlc3 .g6, which are:

A 7 i.f4 B 7 lLle3 C 7 i.e3 D 7 lLlds E 7 .el F 7 dS G 7 g3 H 7 b4

A 7 .i.f4

t:t �. : v{;JJ# �t:t:

56 ;" ;:i� �a

·�· t r� � 1� t :�.� t�:· �� �� �� �� �.�

� � � � �4J�fti t � �

� &J. � � .a. /�1 .a. � /� .a. 9'$ .ua �.a.i .ua � �� J.!. �� ;�?,. � .. '.M,� 1'1:1 �. � 'cS � g ,'GJ�� e

As White's king's knight almost invariably finds itself hopping back from c4 to e3, it is evidently very logical for White to develop his bishop to f4 beforehand, thereby avoiding the possibility of its incar­ceration on c I . This move can also be a precursor to a quick queenside castling. The most obvious drawback, however, is the likely

exposure of the bishop on f4. If Black can play . . . lObs when the bishop is bereft of retreat, he will be doing well.

7 ••• lOf6 Now the alternatives are: AI 8 lLle3 A1 8 .dl Lesser moves: a) 8 h3 .i.e7 9 dS 0-0 10 .<12

.i.d7!? (1O . . . lObd7 1 1 0-0-0 [ 1 1 lObS?! .i.d8 12 0-0-0 a6] I I .. .a6 will effect a transposition to the main game) I I lLle3 �6 12 0-0-0 lLlcs 13 Wbl bS? (this is a little im­patient, 13 . . . a6 is more sensible) 14 .i.xbS :ab8 1 S .i.xd7?! lOfxd7 16 i.g3 .i.f6 17 b3 ( 17 Wa1? :Xb2 18 Wxb2 �4+) 1 7 . . . :b4 1 8 :del :tb8 with a strong attack. Borochow-Zemitis, USA 196 1 . White's fifteenth move was very obliging, and allowed the black pieces to take up very aggressive positions without loss of time. Black would have had a harder time justi­fying his sacrifice after 1 S .i.c4.

b) 8 lObS?! is pointless now, 8 . . . �6 9 dS .i.e7 10 g3?!, very ugly, Ribes-Melchor, Barcelona 1997, as now 1O . . . .i.g4 1 1 .d4 .i.O with a sizeable advantage, or 1 1 .i.e2? .i.xe2 and no matter how White recaptures, the d-pawn will be lost, i.e. 12 Wxe2 �S+. Typical ofOrn play!

8 lLle3 .i.e7

9 .i.c4?! Very popular, but not very good.

Other moves are: a) 9 .i.e2! 0-0 10 .i.g3 ( 10 .d2

would be inaccurate, see (b) 1O . . . c6 ( l0 . . . a6!? 1 1 a4 lOc:6!? [profiting from the hole on b4 created by I I a4] 12 as .i.d7 1 3 lOc:ds .i.d8?! 14 li)f4 "'6 IS .tc:4+ Wh8 16 .i.e6 with an edge, Small-Hector, Royan 1988) I 1 dS a6 12 0-0 cS, setting up a Benoni-style position 1 3 a4 ( 13 It)c:4 is more awkward) 13 . . . :d8 ( 13 . . . li)bd7?! 14 It)c:4 IId8 IS .i.xd6!? .txd6 16 li)xd6 li)b6 17 It)xc8, Lorand-Destrebec:q, Belfort 1983, 17 . . . llaxc8 regaining the �awn, with near equality) 14 as illbd7 I S It)c:4 li)� 16 lie I .i.f5 17 .d2 h6 1 8 Ilab I Ilac8 19 b4 cxb4 20 1lxb4 (White's play is logical, but the black position is solid, and not without counter chances) 20 . . . lIc7 2 1 lIebl .tc8 22 .d4 li)8h7 23 It)e3 (surrounding the e­pawn, but in return Black will _Bet the as-pawn) 23 . . . lIcS 24 Ila l Wf7 25 li)xe4 It)xe4 26 .xe4 .i.gS 27 c:4 .i.f6 28 1la3 .c7 29 lib I li)� 30 .i.g4 .i.xg4 3 1 .xg4 IIxaS, unclear, Nevanlinna-Hector, Jyvaskyla 1993.

b) 9 .d2 c6 10 .te2 (10 dS?! transposes to 9 It)e3 in the main line) 10 . . . 0-0 1 1 0-0 ( 1 1 dS It)hs 12 .i.xhS .xhS 13 dxc6? Ilxf4 14 .dS+ .xdS I S lOc:xdS 1If7 16 lOc:7

Main line_· other seventh moves 31

bxc6 17 It)xa8 It)a6 traps, and wins, the errant knight, Chemer-Clarke, corr. 1998) and rather than 1 1 . . . dS 12 0 exO (this is more sensible than 1 2 . . . .i.e6?! as occurred in Rehfeld-Sc:hmidt, corr. 1992, when, instead of the dreadful 13 li)g4?? It)xg4 14 fxg4 .i.xg4 IS .i.xg4 .xg4 =+ White should certainly have played 13 fxe4 dxe4 with some advantage; and 12 . . . bS?!, intending the typical tactic 1 3 fxe4 b4, is met, unexpectedly, by 14 lOc:xdS cxdS IS eS l&4 16 It)xdS with advantage) 13 .i.xO � Flohr-V ec:sey, Prague 1930, I I . . .lt)hS! is again possible.

9 ••• c6 10 d5 This is the logical follow up, but

10 .i.e2 might be better, despite los­ing a tempo on note a), just above. The game Alonso-Desoto, Spain 1942, went instead: 10 O-O? dS?! (of course, instead of all this, IO . . . li)hS 1 I .i.g3 li)xg3 12 hxg3 dS 13 .te2 0-0 would have been wonderful for Black) I I .i.e2, reaching a similar position, and continued 1 1 . . . 0-0 12 o li)hS? when 13 .i.xb8 Ilxb8 14 fxe4 would have been strong, but Black should ptefer 1 2 . . . exO.

IO ••• li)bS! Black exchanges the f4 bishop,

when his own dark-squared bishop has the potential to become very powerful.

32 Main line: other seventh moves

10 . . . bS?! is worse: I I .i.e2 ( 1 1 .i.b3 !? b4?! [ 1 L..lDhS] 12 /De2 cS 13 .i.g3 �g4 14 llX:4 Olafsson­Diepstraten, simul, Hilversum 1976, when 14 ... 0-0 limits White's edge) I L. .b4 12 �4 0-0 13 a3 �g4! (a big improvement on the game Sir G. Thomas-Tartakower, England 1926, which continued: 13 . . . bxa3?! 14 lba3 .i.d7?! IS �b6 axb6 16 lbaS �dS 17 �dS cxdS I S .i.g3 with advan�.8e) 14 .i.g3? ( 14 .i.xg4 .i.xg4 IS "xg4 .xg4 16 �xg4 lbf4 17 /De3 should keep some ad­vantage) 14 . . . �xe3 I S fxe3 .gS 16 axb4 .i.b7 when White has prob­lems finding a safe haven for his king, Dagenais-Polland, COrT. 1984.

1 1 .i.g3 �:s:g3 Il h:s:g3 I don't believe that 1 2 fxg3 !? can

be good. Perhaps 12 . . . .i.gS is the sharpest rep_�y, with the two possi­bilities: 1 3 "d4 cS 14 .i.bS+ WdS! I S .d2 �6 16 �I Ilffl, and 13 .e2 bS!? 14 .i.b3 b4 IS �I cS 16 0-0 Ilffi.

12 ••• li)d7

l3 .d4 Otherwise: a) 13 .d2, in Littrell-Dreibergs,

COrT. 1964, Black was successful with 13 . . . .i.ft)! ( I 3 . . . /DeS?! 14 .d4 wins a pawn by 14 .. . �c4 I S .xc4 .i.gS 16 .xe4+ .xe4 17 �e4 .i.xe3 I S fxe3, rather than 14 O-O-O?!, Bullockus-Schild, COrT.

1979, when 14 . . . �xc4 I S �c4 .i.gS 16 /De3 0-0 is clearly good for Black, t2 is weak and Black has the two bishops) l l4 0-0-0 .i.eS!? ( 14 . . . .i.xc3 I S .xc3 cS may be sim­I?ler) I S Ilh4 ( I S dxc6 bxc6 16 1lh4 fl)ft) transposes) IS . . . �ft) 16 dxc6 ( 16 .i.b3! intending llX:4) 16 . . . bxc6 17 .e2?! ( 1 7 �e4? �e4 IS .i.d3 .i.xb2+ 0-1 , Bullockus-Scott, COrT. 197 1 ) 17 . . . .i.xc3 I S bxc3 dS.

b) 13 .i.e2, with a threat! 13 ... 0-0 (simple and good, 13 ... m? 14 .i.hS �hS I S .xhS .xhS 16 lbhS g6 17 1lh2?! 0-0 IS �e4, when White has won a pawn, al­though Black's bishops offer some compensation, La Mar - Portillo, COrT. 197 1 ) 14 dxc6!? bxc6 IS lDeds .i.gS 16 �7 IlbS 17 .i.c4+ dS I S �xdS WhS! ( I S . . . cxdS 19 .xdS+ 1lf7 is interesting, but not 19 . . . WhS?! 20 �) 19 fl)f4 .i.xf4 20 gxf4 �b6 2 1 .i.n .i.g4 and the active black pieces are well on top, Sowden-Saunders, COrT. 1996.

c) 1 3 /De2?! /Des (it seems to me that 1 3 . . . .i.gS! is a better choice here; . . . .i.xe3 is the threat, and 14 �f4 appears to lose a pawn to 14 . . . .i.xf4 I S gxf4 .ft» 14 �f4 .ft) IS .i.b3 .i.d7? 16 dxc6 bxc6, Toro-Saaveira, COrT. 196716S, when 17 .e2 is unclear, and 17 lDedS!? cxdS IS �xdS .gS 19 �7+ WdS 20 �aS e3! likewise.

13 ••• �f6 1 3 . . . �S! 14 0-0-0 .i.ft) IS .d2

.i.eS is worth serious consideration, Black seems to be better.

14 0-0-0 The 14 /De2 cS I S .i.bS+ Wf7 16

.d2 a6 17 �f4 .gS I S .i.e2 bS 19 0-0-0 .eS of Veiss-Dreibergs, COrT. 19S2, is similar .

14 ••• cS!? 14 . . . .i.d7 avoids the displacement

of the king, I S /De2 cS 16 .d2,

Wang-Dreibergs. corr. 1965. when 16 . ..IohS would stop the e2-knight coming to f4.

IS .i.bS+ � 16 .d1 .6 17 -*.e1 bS 18 :M illS

1 8 . . . b4 19 lOa4 Ilb8 of Alexander-Hasenfuss. Folkestone 1933. weakens the c4-square.

19 11dbl �g8 Black has a promising Benoni­

style position. McKenna-Saunders. corr. 199O.

A1 8 .d1

White intends to castle long. but Black is well placed to meet this.

8 ••• .i.e7 9 � 9 lOe3 c6 10 dS?! ( 10 -*.e2! trans­

poses to the note 9 .d2 of 8 lOe3) 1O . . . �hS! 1 1 -*.g3 0-0 12 -*.c4 ( 12 .i.e2) 12 . . . cS 13 0-0 a6 14 a4 �7 I S Ilael It)xg3 16 (xg3 (unfortu­nately for White. 16 hxg3? is im­p'ossible: 16 . . . lOes 17 -*.a2 -*.gS 18 • d I -*.xe3 19 lIxe3 �g4 20 1Ixe4 �S winning the exchange) 16 . . . lIxfl+ 17 IIxfl lOes 18 lOe2 -*.d7 19 �f4 �6 20 as IlfS 2 1 lie I ? .i.gS 0-1 . Schabanel-Kosten. French League 1994. since 22 .i.a2?! say. is answered by 22 . . . lIxf4 23 gxf4 -*.xf4 24 h3 .gS 25 �hl .i.xh3 26 gxh3 .g3.

9 ••• 0-0 ID b3 Black was actually threatening a

tactical operation involving . . . �g4.

Main line: other seventh moves 33

followed by .. . lIxf4 and .. . -*.gS. and this move is played to control g4. and prepare a retreat square for the bishop. Other possibilities:

a) I 0 �b I takes the king off the exposed diagonal. 1O ••• a6 1 1 -*.e2 bS ( I 1 . . ..xg2!? is possible as is 1 1 . . . �7 like the main game) 12 lOe3 c6 13 -*.g3 ( 1 3 g4!? Badii­Lopez de Castro. Paris 1994. is more aggressive. but 1 3 . . .• f7 looks like a good reply. or 1 3 . . . dS fol­lowed by brinsing a piece to c4) 13 . . . dS 14 h4 .i.e6 I S hS .f7 16 f4 -*.d6 17 -*.h2 .c7 with chances for both sides. Siguljonsson-Gundersen. Graz 1972.

b) 10 h4? controls gS. but it is un­wise to weaken g4. i.e. 1O . • . -*.g4 1 1 f3?! exf3 12 .i.d3 -*.13 13 �xf3 .xf3 14 IIbfl ? ( 14 libel was only l!) 14 . . . bS ! (winning!) I S Ildel 1lf7 0-1 . Lynberg-Hector. Malmo 1990.

c) 10 -*.e2? instructively ignores the threat: 1O . . . �g4 1 1 -*.e3. Vutov­Dimov. Teteven 1991 . when I can see nothing wrong with the evident I I . . .�xf2 12 -*.xf2 llxf2.

10 •••• 6! In Ciocaltea-Destrebecq. Val

Thorens 1980. Black tried 1O •.• �7 1 1 �bl �b6?!. but this left the queenside ,pawn formation somewhat inflexible after 12 It)xb6 axb6 13 dS! .

1 1 dS �bd7 11 �bl A necessary preparation for g4 •

which. if played immediately. loses at least a pawn to 12 g4? IOXg4! 13 hxg4 11xf4 14 .xf4? -*.gS.

ll ... bS Not the prelude to a queenside

onslaught. as one might expect. but part of a strong plan to artificially isolate White's dS pawn. and re-arrange Black's minor pieces. The knights are aiming for b6 and cS. respectively. and the dark squared bishop for £6.

34 Main line: other seventh moves

I3 lOa5 lilb6 14 a3 There is nothing to be gained by

14 1Oc6 as 14 . . .• n followed by .. . i.b7 puts great pressure on the d-pawn.

14 ••• lilfd7 15 13 Of course, with the knight on as

instead of e3, I S .i.e2? would leave the g-pawn en prise.

15 ••• liX5 16 :i.ll?

Allowing a small combination, but 16 i.e3 i.f6 is also better for Black.

16 •• Jbf4! 17 W:d4 Or 1 7 gxf4 Wxg2 1 8 :dgl Wo

19 Wd4 �f8 � as now 20 IIxg7+?? fails to 20 . . . i.xg7 2 1 IIgl .i.g4 22 Ilxg4 Wxg4! .

17 ••• -*.g5 18 -*.xe4 lilxe4 19 Wxe4 i.f5 10 Wd4

White cannot defend the c-pawn with 20 We2 because of 20 . . . lIe8.

10 ••• -*.nl+ 1I �.l .i.f6 II Wdl i.xdl 13 l1xdl

The net result of the combination is a vulnerable d-pawn for White, and a strong bishop for Black. Con­siderable advantages in the hands of Sokolov.

ll .. .Ile8 14 14 Wn 15 �.I .i.e5 16 Wd3 .xfl 17 liD WIl 18 .f5 .i.f6 19 IIfl WxhJ 30 liX6 "'6 31 �.l WI5 31 WXl5 -*.X15 33 lilb4 i.f6 J4 lilu6 .i.n3 35 bxcJ lilxd5

0-1, Apicella-I.Sokolov, European Teams ch, Debrecen 1992.

B 7 lLleJ

Although apparently impeding the movement of White's Queen's bishop, the knight is well placed here, and should the f-file become open the knight can often hQp advantageously to f5.

7 ••• lilf6 Natural, and best. If 7 . . . c6!? 8 0 is a good choice,

(8 dS lilf6 9 -*.e2 transposes to the main line) 8 . . . dS 9 fxe4 dxe4 10 .i.c4 and, although Black has some space, the fact that he cannot castle kingside, because of the open a2-g8 diagonal, is going to be a problem, Pedersen-Taksrud, corr. 1979. And 7 . . . lOc6?! 8 lLleds (8 .i.e2 lilf6 9 lLlcds, French-Zemitis, corr. 1968, is less effective) 8 . . . �d8 9 lLlbs .i.g4 10 .d2 is awkward.

8 .i.el! Not the sharpest move, but the

most sensible, and strongest. The others are:

a) 8 -*.c4 c6 9 dS (White does not want Black to play . . . dS, 9 O-O?! dS 10 -*.b3 i.d6 1 1 0 [it is difficult to arrange to play c4, as 1 1 lile2?! meets I l . . .lll.s4! 12 lilg3 hS 13 lie I 1Oxit2?! 14 i.&dS! Galje-De Vries,

corr. 1991 , when the natural 14 . . . cxd5 I S .i.xd5 �dS 16 fue4 leaves the situation somewhat con­fused, but 1 3 . . . h4 14 ltlgfl h3 I S g3 0-0 is exceedingly dangerous for White] I l . . .exO 12 .xO 0-0 13 h3 .i.OO, the white queen is awkwardly placed on the Hile, Black is better, Lonsdale-Svendsen, corr. 19S7) 9 . . . .i.e7 (the immediate 9 . . . ltlbd7! is even better, 10 0-0 �5 1 1 .i.e2 .i.e7 12 ltlc4? ltlxc4 13 .i.xc4 .i.h3 14 g3 .i.xfl IS .i.xfl 0-0 0-1 , Mazuchowsky-Dreibergs, Michigan 1963) 10 a4!? (played by the great Bobby Fischer although he was only a lad at the time, and it was a long time before he was to become World Champion! 1 0 0-0 ltlbd7 1 1 o exO 1 2 .xO?! �5 = Gabrans­Tomson, corr. 1974) 1O . . . ltlbd7 1 1 as �5 1 2 .i.e2 0-0 13 0-0 .i.d7 14 �h I �hS IS 1Oc4 lilfg4 with plenty of activity, Fischer-Pupols, US Junior ch, Lincoln 1955.

b) S ltlcd5 (S lLled5 ltlxd5 trans­p.o5eS, of course) S . . . ltlxd5 9 ltlxd5 'fIf7 10 .i.c4 c6

b I ) White can take the exchange here, but he is well advised not to: 1 1 lilb6?! d5 12 fuaS (White can gain a pawn, at the cost of a tempo, by 12 .i.xd5!? cxd5 13 fuaS .i.d6 14 0-0 0-0 I S .i.e3 1Oc6 16 c4 [ 16 o appears more to the point, but

Main line: other seventh moves 35

after 16 . . . -.,5 17 h3 Black wins by 1 7 . . . .i.xh3 ! I S gxh3 .xh3 19 11f2 exO, without even having to capture the as-knight! ] 16 . . . dxc4 17 d5 �5 I S .a4 ltlO+! forcing mate, Spaans-Diepstraten, Hilversum 19S5. 1 2 lOxcS?? is just a blunder, though, 12 . . . dxc4 13 .g4 lLld7 14 .xe4+ �dS trapping the knight, I S lilxa7 .i.b4+ 16 Wdl llxa7 and White can resign, Ravel-Sireta, Auvergne 1993) 12 . . . dxc4. Clearly White will be unable to extract his knight from as and when Black eventually takes it he will have two pieces for a rook. What can White do in the meantime? The evidence is: not very much; there is only one file for the white rooks, further­more, Black has the use of the d5 square. Some examples:

b l l ) 13 d5 .i.d6! 14 .i.e3 ( 14 dxc6 0-0 I S 0-0 lidS 16 .e2 ltlxc6 also favours Black) 14 . . . c5 ( I4 . . . cxd5!?) I S .e2 0-0 16 .xc4 ltla6, Hohnes-Dreibergs, corr. 1954, when 1 7 .xe4 .i.fS I S .c4 llxaS favours Black.

b 12) 1 3 O!?, Sloan-Elburg, corr. training game 1999, is most simply answered by 13 . . . e3 !? keeping the f-fiIe closed, e.g. 14 .i.xe3 .i.d6 I S 0-0 b5 16 a4 b4.

b l 3) 13 .e2 .i.d6! 14 .xe4+ ( 14 o e3 IS .xe3+ .e7!? [ 1 5 . . . .i.oo resembles the main line] 16 0-0 .xe3+ 17 .i.xe3 ltla6 IS c3 .i.fS 19 .i.f2 �d7 and . . . llxaS ; Buter­Dreibergs, corr. 195 1 ) 14 . . . .i.oo ( 14 ...• 00 is also reasonable, I S .xe6+ .i.xOO 16 .i.e3 .i.d5 17 b3 c3 I S 0-0 lLld7 19 life I 0-0 with the usual two piece for rook advantage, Waiter-Purser, corr. 1977) I S 0-0 ( IS .i.e3 0-0 16 0-0-0 .i.d5 ; Gabnns-Purins, corr. 1974176) 1 5 . . . 0-0 Farooqui-Smit, Teesside 1973. Black will move his knight

36 Main line: other seventh moves

and capture its white counterpart; the I S . . . Wet7?! of Harris-Kampars, corr. 1969, looks too contrived.

b2) 1 1 tOe3 dS 12 .t.b3 .t.e6 ( l 2 . . . .t.d6 is also good, not fearing c4: 13 0-0 0-0 14 f4 [ 14 f3 �S IS f4 GrlvainislHayward-Malmstrom, corr. 1997, and 1 would be hap'py to play the endgame here, l S . . .• xdl 16 lDxdl .t.e6 etc.] 14 . ..Iod7 I S c4 �f6 = Doelling-Dreibergs, corr. 19S2) 13 0-0 llXI7 14 f3 (or 14 f4) 14 . . . exf3 I S :Xf3 ( I S .xf3 .xf3 16 :Xf3 0-0-0 = Fletch�Dreibergs, corr. 1966) I S . . . �f6 16 .e2 0-0-0 = Venesaar-Kampars, corr. 1967/68.

c) 8 f3 .t.e7!? (not completely necessary, 8 . . . exf3 9 .xO �6 leads to a set up examined in the �revious chapter, after 8 . . . �f6 9 �3)

c l ) 9 .i.c4 OO?! (I feel that Black should take on f3 now: 9 . . . exf3 10 .xf3 OO!?) 10 fxe4 ( 10 dS?! �bd7 I I .e2 exf3 12 .xf3 lOeS Jackson­Hayward, corr. 199 1 ) 1O . . . lDxe4 I 1 o-O!? (not I I .t.d3?? .t.h4+ winning on the spot, but I I lDxe4! .xe4 12 0-0 dS 13 .t.d3 also gives a clear edge, the knight will come to fS in due course) 1 l . . .�f6 12 dS �bd7? 13 dxc6 bxc6 14 �fS is awful for Black, Ciric-Strobel, Imperia 1966.

c2) 9 �S lDxdS 10 lDxdS Wd8!? ( 1O . . . .t.d8! is superior, as the

attempt to win the e4 pawn back­fires: I I .e2 0-0 12 lOf4 [ 1 2 fxe4 lIe8 1 3 �3 .t.fS] 1 2 . . . exf3! 13 lDxg6 fxe2 14 lDxfB exfl =.+ I S IIxfl .t.e7 and the knight is lost!) I I lDxe7 Wxe7 12 fxe4 ( 12 .e2, Logunov-Elburg, corr. 1994/96, is risky, for Black can try 12 . . . lIe8! 1 3 fxe4 Wd8 and the white king is also not ideally placed) 1 2 . . .• xe4+ 1 3 .e2 and White has an edge, Bakkel'"'Diepstraten, corr. 1975.

c3) 9 .t.e2 0-0 10 lDxe4? ( 10 0-0 exf3 I I .t.xf3 00 transposes to a later line) 1O . . . lDxe4 1 1 fxe4 .xe4 12 .t.f3?! .t.h4+ 1 3 g3? :Xf3 14 gxh4 :Xe3+ is a catastrophe. Weening-Backhuijs. corr. 1982.

c4) 9 fxe4 lDxe4 10 .t.d3?? 0-1 , Walthel'"'Traut, corr. 1986. but in Heisterhagen-Huehn, Hessen 1992. White actually played on seventeen moves! 1O . . . .t.h4+ 1 1 We2 .t.g4+ 12 lDxg4 .xg4+ 13 Wfl 0-0+ 14 WgI .t.f2+ etc.

8 ... c6 There was a threat of �S to

contend with. so 8 . . . .t.e7?! is worse, 9 �S! (9 0-0 0-0 10 f3 exf3 1 1 .t.xf3 c6 12 dS transposes to the main line, once again) 9 . . . .t.d8 (9 . . . lDxdS?? 10 .t.hS) presents White with a choice of two good lines:

a) 10 lOf4 finds the black .9ueen embarrassed for squares. 1O . . .• gS!? 1 1 g3 ( 1 1 g4!? succeeded admirably in Kvist-Bondick. corr. 1977, I I . . .c6 [ 1 1 . . .• xf4? 12 lDg2 wins the queen for two pieces, but 1 1 . . . 0-0 is clearly best, when White may regret his kingside weaknesses] 1 2 lill5! .t.xf5 13 h4 and Black resigned as his queen is once again lost. How­ever, he should have continued, as after 13 . . .• xg4 14 .i.xg4 .i.xg4 I S .d2 0-0 Black has a certain amount of compensation!) 1 1 . . .• aS+ 12 .t.d2 -.b6 1 3 .i.c3 c6 14 dS .c7 I S .d2 0-0 16 h4 with advantage to White, Rocha-Cruz, Lisbon 1997.

b) 10 lDxffi+ .xft) ( lO . . . .i.xft)?? 1 1 .i.hS) 1 1 0-0 �6 12 c3 0-0 13 o .g6? ( 13 . . . exO 14 .i.xO .e7 is not very nice, but better than that pJayed) 14 fxe4 :xfl+ I S .xfl Wxe4? 16 .i.c4+ .i.e6 17 .i.xe6+ .xe6 1 8 dS 1 -0, Miiller-Gorla, Switzerland 1979, by transposition.

9 0-0 a) Play can transpose into the

main line on 9 dS .i.e7 10 0-0, but Finnish Grandmaster Westerinen tried 9 . . . 1Obd7!? against Pri� at Fourmies, 1988, and although the game did not turn out too well, this was not because of the move itself: 10 b3 �S 1 1 .i.a3 cS ( 1 I ...lDt7) 12

Main line: other seventh moves 37

lDbS .t7 13 .i.b2 a6 14 .i.xeS dxeS I S �3 bS 16 0-0 .i.b7? ( 16 ... .i.d6 is fine) 17 .i.g4! :d8 18 .i.e6 � 19 �e4 and White was already winning.

b) Otherwise, the immediate 9 0 is possible 9 .. . exO 10 .t.xO dS!? (the most ambitious, Black wants to develop his dark-squared bishop straight to d6; 1O .. . .i.e7 is safer, 1 1 .i.e2!? 0-0 12 0-0 dS 13 .i.d3 .e8?! [ I 3 . . .• t7 looks more active] 1 4 lDf5 .i.xf5 I S .i.xf5 .i.d6 16 .0, Rittenhouse-Svendsen, corr. 1992, when Black missed the sil¥le 16 . . . g6! 17 .i.d3 lOe4 18 .e2 :xfl + 19 .xfl �c3 20 bxc3 �7 fairly level) 1 1 0-0 (Although I can­not find any games featuring 1 1 .e2!? it seems interesting, 1 1 . . . Wd8 [ I 1 . . . .i.e7? 12 �xdS cxdS 13 �dS] 12 0-0 .i.d6 brings about a similar position to that after 1 1 0-0, but with the black king in the centre. I 1 �xdS!? is optimistic, 1 1 . . . cxdS 12 ltixdS Wd8!? [ 1 2 . . . ltixdS?? 13 .i.hS, Zschom-Melchor, corr. 1987, 1 2 ... .i.d6 1 3 0-0 transposes into a later note] 1 3 lDxf6 gxft) White's compensation should not be suffi­cient) I I . . ..i.d6 1 2 li)exdS!? (this is quite popular, although 12 .i.e2 is probably stronger, 12 . . . 0-0 13 .i.d3, fairly level, Albisetti-Faraoni, Zurich Open-B 1990) 12 . . . cxdS 13 .e2+ ( 13 �dS?! 0-0 14 .t.f4 lDxdS I S .i.xdS+ Wh8 16 .i.xd6 .xd6 17 �S �7 1 8 :XfB+ .xfB 19 :fI lDft), Black is consolidating, Zschom-Strelis, corr. 1987; 1 3 �S Wd7! 14 �d6 Wxd6 I S .i.f4+ Wd7 16 c4 Wd8 17 cxdS .i.g4 18 'eh3 .i.xO 19 :XO .17 unclear, Szabo-Eberth, Miskolc 1998) 1 3 ... Wd8 14 �dS iOxdS I S .i.xdS � 16 c3?! :e8 ( 16 . . . .i.xh2+?! looks tempting, but 17 wxh2 .d6+

38 Main line: other seventh moves

1 8 �h I .xdS 19 c4 is actually very messy) 17 1rbS a6 18 1rb6+ �c7 19 .cS .i.d6 Ih-Ih Niermann­Leisebein, East Gennany 1983.

9 ••• .i.e7 Again, Black can consider playing

9 . . . dS and putting his king's bishop on d6: 10 0 ( 1 0 lOcxdS? is worse here than after 9 0, or 1 0 0, as White has no open Hile to use; on 1O . . . cxdS 1 1 �dS �d6 1 2 �f4 .i.xf4 [ 1 2 . . . lOxdS 13 .i.hS �xf4 14 .i.xg6+ bxg6 IS h3 is unclear] 1 3 lOxf4 .gS 14 g3 0-0 the white compensation is quite insufficient, Zschorn-Kozlov, COIT. 1987) 1O . . . exO 1 1 .i.d3 ! (this idea was not possible previously; the automatic 1 1 .i.xO is considered in the note to 9 0) I I . . .�S 12 .xO .xO 13 :xO -*.d6 14 .i.d2 0-0 I S :aft ;t, Maric-Smit, Strasbourg 1973.

10 f3 10 dS 0-0 1 1 0 exO 12 .i.xO

transposes. 10 ..• exf3 Black's position after 1O . . . dS !? 1 1

fxe4 dxe4 12 .i.c4 :fS, Svendsen­Downey, COIT. 1990, is compro­mised, but as the main line is also favourable to White, this is no worse. 1 3 �2 lObs (else 14 IOf4 with advantage) 14 dS?! (I don't like this, why open the gl-a7 diago­nal? 14 J:lxf8+ �xf8 I S .i.d2) 14 . . . J:lxft+ I S .xft .i.cs 16 �d2 .gS!? 1 7 .f2 bS 1 8 .i.b3 is close to equality.

And 10 . . . 0-0 1 1 fxe4 IOxe4 12 J:lxf8+ .i.xfS 13 IOxe4 ( 1 3 �c4+!? �h8 14 .0 IOffi I S �d2) 13 ...• xe4 14 .i.d3 �4, Christoph­Cherubim, 19S9, is also a little bet­ter for White after IS c3.

I I .i.xf3 White has two other good

alternatives:

a) 1 1 .i.d3 �S 1 2 .xO (ex­changing the queens makes Black's defence more difficult, 12 J:lx0 trans� to 1 1 :xO, below) 1 2 . . .• xO 13 :xO dS ( 1 3 . . . 0-0 14 1Of5 .i.xfS IS J:lxf5 [IS �xf5! is better as the bishop is more active on the h3-c8 diagonal] I S . . . lObci7 16 .i.gS :ae8 Black enjoys near equality, Boudre-Destrebecq, �I­fort 1983) 14 1Of5 ( 14 .i.d2 is also 800d, 14 . . . 0-0 IS J:le l .i.d6 16 10fS .i.xf5 17 .i.xf5 with a plus, Nilsson­Niemand, COIT. 199 1 ) 14 . . . �xf5 IS .i.xf5 0-0 16 �gS 10&6 17 :e I �d6?! ( 1 7 . . . :ae8 1 8 :fe3 .i.d6 limits White's advantage) 1 8 J:le6 J:lad8?! 19 :h3 winning a pawn, Bengtsson-Ekstroem, Sweden ch 1971.

b) 1 1 J:lx0 0-0 12 .i.d3 1rhs 1 3 1Of5 ( 13 .i.d2 dS 14 .e2 J:le8 I S :aft 10&6 16 h3 �6 is quite reasonable for Black, Ladisic­Destrebecq, corr. 1984) 13 . . . .i.xf5 14 .i.xf5 �h8!? I S dS? (opening the position prematurely, I S .el main­tains a plus) I S . . . cxdS! 16 lOxdS IOxdS 17 .xdS lOc6 1 8 .i.e4 .xdS 19 �xd5 lOd4 and White is in trou­ble, Svendsen-Koser, corr. 1992.

1 1 ••• 0-0 12 d5 White can also consider 12

.i.e2-d3, as per the Rittenhouse-

Svcndsen game in the note to White's ninth move.

1l .•. c5 13 .e1 It is a matter of taste how best to

tackle this Benoni-style position. Other moves:

a) 1 3 :Cl is also good, 13 . . . lDbd7? (in light of the problems Black had in this game, 13 . . . lDa6 would seem to be an improvement, keeping an �e on the c7 square, e.g. 14 lDc4 1leS I S ,j.f4 lDb4!?) 14 1Dc4 lieS 1 S lDbS -*.dS 16 lDbxd6 Ilxe 1 + 1 7 .xe 1 lDf8? I S .e2 b6? 19 iDes 1 -0, Gaard-Jackson, corr. 1993.

b) 13 -*.e2 lDa6 14 ,j.d3 .eS I S lDfS ,j.xfS 16 -*.xfS lDc7 17 .d3 �S IS ,j.d2 lDg4 and Black has enough play, Merenyi-Spielmann, Budapest 1925.

13 ••• b6!? 13 ... lDbd7?, Piris-Loureda, Spain

I99S, whilst tempting (Black hopes to establish this knight on eS, when all his troubles will be behind him), is again a mistake, 14 lDc4 .17 I S -*.f4, there is no way to defend d6, i.e. I S . . . iDeS 16 ,j.hS g6 17 ,j.h6.

14 a4 14 lDc4 ,j.a6 I S .xe7 ,j.xc4 16

1Ire6+ is also possible. 14 •• .Ild8 Black embarks on a lot of unnec­

essary prophylaxis, 14 . . . ,j.a6 IS lDbS lDtXi7 16 lDc4 llaeS is more active, although White has a certain advantage.

15 b3 lDa6 16 lDb5 .e8?! 17 ,j.b1 .d7 18 Ilael 1le8 19 lDc4 ,j.b7?!

19 . . . ,j.f8. 10 .e6+! .u6 II Ilxe6 lDxd5

ll lDcxd6 lDac7 13 lDxb7 lDxe6 14 -*.xd5 l-o

Bauer-SCnechaud, Metz 1997, a model White victory.

Main line: other seventh moves 39

C 7 ,j.e3

Quite a popular move in OTB games.

7 ••• lDf6 8 h3 In order to prevent the possibility

of the .. . lDg4 sortie, but in Kraus­Dreibergs, corr. 19S4, White contin­ued S Wd2 ,j.e7 9 0-0-0 anyway, 9 . . . 0-0 (as after 9 . . . lDg4 White can reply 10 lLldS! ,j.dS [ 1 O .. . lDxe3? 1 1 .xe3 ,j.gS meets 1 2 f4 as the c4-pawn is pinned] 1 1 0 exO 12 gxO lDxe3 13 .xe3+ gives White a certain initiative) 10 0 ( 10 h3 trans­poses to the main line) 1O . . . exO 1 1 gxO with some play down the g­file. S dS is common, and transposes into F.

8 ••• -*.e7 9 1Ird1 In order to castle; 9 g4!? is

sharper, but weakening, 9 . . . 0-0 10 ,j.e2 c6 (intending the standard .. . bS) 1 1 gS lLldS! 12 lDxdS cxdS, forki�g a piece, and a pawn, 13 ,j.hS WfS 14 -*.g4 Wn I S lOd2? ( I S ,j.xcS IlxcS 16 lOd2 is OK) I S . . . ,j.xgS! , winning an i!"portant pawn, Lillo-Cueto, Santiago de Chile 1997.

9 ••• 0-0 So far both players attempt to

complete their mobilization in �e most straightforward manner, but ID Oren-Krantz. COrT. 1990, Black

40 Main line: other seventh moves

went in for a bit of restraint: 9 . . . hS?! 10 0-0-0 J.d7 1 1 J.e2 as 12 a4?! Itlc6 1 3 lObS?! 1Ols! 14 �bl 0-0 with complicated play.

10 0-0-0 c6

Unanimously preferred here, . . . bS becomes an unpleasant threat.

1 1 g4?! White continues his kingside plan,

but this is dubious. 1 1 J.f41 stops . . . bS, but allows

something even nastier: 1 1 . . .lDdS! when 12 J.e3 is virtually forced, leaving Black on top, Meyer­Haenisch, Hessen 1990; not 12 lOxdS1 cxdS 1 3 lDas lhf4 when the white king and queen prove to be on an embarrassing diagonal.

1 1 dS is probably the best move, as then 1 1 . . .bS 12 lDas b4 13 lile2 frees the c4-square for the knight, 1 3 . . . lDxdS? losing a piece to 14 lDxc6 J.e6 IS .xdS! .

1 I ••• bS In the game Oren-Melchor, corr.

1990, Black played the awful move 1 1 . . .aS1 and was severely punished following 1 2 lDb6 Jb7 1 3 dS! cS 14 J.xcS! (although 14 lOxc8 lhc8 IS gS lOfd7 16 J.g2 can't be bad either) 14 . . . dxcS I S d6 J.xd6 16 .xd6 lOfd7 17 J.c4+. Curiously this strange idea ( . . . as) has some-

thing of a history, because in the game Roach-Gabrans, corr. 1968/69, they reached the same p0-sition except that the white queen was on e2 and the white g-pawn on g3: after l 1 . . .aS? 1 2 1Ob6 Jb7 1 3 J.g2? dS, this time it was Black who quic�e out on top.

11 With the bishop on e3, this knight

has nowhere else to go. 12 •• .J.d8! The point! 13 &2 lDds 14 lDg3 lOse3 IS

fse3 :0 16 lObS J.g5 I7 lOf4 Oren-Svendsen, corr. 1990, when

in place of 1 7 . . . J.xf4? 1 8 exf4 .n 19 dS e3 20 .d3 .xf4 2 1 J.e2 :f2 22 dxc6 J.e6 23 .xd6, with advan­tage, 1 7 . . .• f7! would have been difficult to meet.

D 7 lDds

This whole idea looks suspi­ciously like a waste of time to me. These knight sorties in the Latvian tend to be playable if associated with a specific tactical idea, but here Black's simple reply defends c7 and attacks the knight, gaining time.

7 •••• " 8 lOde3 This suffers from the disadvan­

tage of occupying the c4-knight's

logical retreat square, but other moves also have drawbacks:

a) 8 �3?! is actually the most common in practice! 8 . . . Ae7! (Of course, 8 . . .• g6 is possible, offering a repetition, but why not exploit White's loss of tempi? Meanwhile, 8 .. .Illft;?! is less accurate, for after 9 .*.gS White will capture on ffi, e.g. 9 ... .*.00 Carlsson-Stalne, Vaxjo 1992, 10 .*.xffi gxffi [ 10 . . .• xffi I I �e4] 1 1 dS .*.fS 12 .e2) 9 .*.e2 (White is unlikely to get enough compensation for the piece on 9 �e4!? OO 10 .*.d3 dS) 9 . . . lOffi 10 (H) 0-0 I I .*.e3 �6, equal, Voigt­Mlotkowski, Philadelphia 192 1 .

b) The 8 lOf4 of Heilmann-MolI, Berlin 1907, is slightly better, 8 . . . lOffi 9 .*.e2 (White can also con­sider playing for control of 00 by an immediate 9 dS .*.e7 10 lOe3 0-0 1 1 lOe6 [White has no time to increase his hold on 00 by 1 1 .*.c4 as l l . . .lOfd7 gives him severe prob­lems on the Hile, 1 2 lOg4 AgS 1 3 lOe6 .*.xc l 14 :Xc i lOeS! winning] 1 l . . ..*.xOO 1 2 dxOO .g6 1 3 f4 exO Van Nouhuys-De Raadt Offerhaus, Zuidhom 1 883, 14 .xO lOe4 I S .g4 .xg4 16 �g4 dS favours Black) 9 . . . .*.e7 10 0-0 0-0 1 1 lOe3 c6 12 dS gS!? 1 3 .*.c4? ( 1 3 .*.hS) 13 ... bS 14 lOe6 bxc4 which should have been winning.

c) 8 lOce3?! allows Black to ex­pand in the centre with gain of time, 8 . . . c6 9 �3 dS (good, now that White cannot play c4, or �-eS) 10 .*.e2 lOffi 1 1 0-0 .*.d6 12 0 exO 1 3 .*.xO (H) 14 .d3! lOa6 I S a3 with preference for Black, if any­one, Baudry-Destrebecq, Belfort 1983, by transposition from 4 � (misplayed by White), and with a move more for each side.

Main line: other seventh moves 41

8 ••• lOf6 Neither 8 . . . dS? 9 lOes .00 10 c4

c6, Humpiueys-Simons, England 1992, 1 1 cxdS! cxdS 12 AbS+ with advantage, nor 8 . . . c6?! 9 dS cS, Castelli-Morgado, corr. 1970, 10 a4! with a good Benoni-style posi­tion, are as good, but 8 . . . lOcI7!? has more to it, 9 Ae2 dS 10 lOd2 Ad6 with a good position.

9 .*.e1 9 g4!? has to be taken seriously,

as it was played by a FIDE KO World Champion, albeit a rather roung one, 9 . . . h6?! (9 ... .i.e7 10 gS l&iS appears simpler) 10 c3 Ae7 1 1 h4 AOO 12 Ilgl ?! %7 1 3 %lbl c6 14 .c2 (H) I S Ilh2 lOcI7 16 .*.d2? .*.xh4 and Black went on to win, Anand-Destrebecq, simul, Lyon 1988.

9 •• .'*.e7 This is more straightforward than

9 .. . c6 10 (H) bS? 1 1 lOd2 with advantage, Goncalves-Grivainis, Munich 1958.

10 0-0 0-0 11 f3 1 1 dS as !? 1 2 b3 !? bS 1 3 lOa3 b4

14 lObS .i.d8 I S c4 lObd7 16 a3 ! favours White, Nieber-Diepstraten, Hilversum 1977.

1 I • • • ellf3 11 .*.llf3 c6 13 d5 cS 14 .4

. . . with a slight edge ID this Benoni-style position.

42 Ma;n line: other seventh moves

7 ••• lOf6 8 f3 The point behind White's strange

7th move is this attempt to win the e4 pawn, but the weakness of White's d4 pawn gives Black his counterplay.

White can also try to pick up the e-pawn by 8 dS �e7 9 lDd2!? but he risks falling behind in development, ��e3 0-0 10 h3 a6 1 1 g4?! bS 1 2

? ! [ 12 gS] 1 2 . . . b4 1 3 ltia4 �d7 14 b3 .i.bS I S lOc4 lOxdS is crush­ing, Schulz-Menningmann, Duis­burg 1991 , or 9 h3 0-0 10 �f4 lObd7 1 1 0-0-0 lOb6 Leibsen­Alberts, corr. 1978/80, although I would prefer 1 1 . . .lOcS) 9 . . . 0-0 10 lDdxe4 ( 10 g3?! weakens the light squares, 1O . . . c6! 1 1 dxc6 lOxc6 1 2 lOdxe4 .A.g4 13 lOxffi+ .A.xffi 14 .c4+ �h8 I S �g2 Zlae8+ 16 .A.e3 lOes 17 .a4?! lDd3+! winning, Burke-Stobbe, Upton, USA 1989, as 1 8 cxdJ �xc3+ 19 bxc3 .xdJ 20 .xg4 Jbe3+ 2 1 fxe3 .xc3+ 22 �e2 .c2+ 23 �el .xg2 is very good for Black) 10 . . . lOxe4 1 1 lOxe4 ( 1 1 .xe4?! �f5 12 � lDd7 13 1Wxb7 Zlae8 14 .xa7? [sinjle­minded! 14 .A.e3 �gS I S .xa7 l&S is also unpleasant, but does offer chances of defending] 14 . . . �h4+ 0-1 , Pavilons-Grabans, corr. 1968, as after I S lOe2, say, I S ... .A.g4 16

�e3 .xc2 there is no real defence against the black knight hopl'ing in to dJ) 1 1 . . .c6 12 .c4? ( 1 2 WdJ re­tains the possibility of defendinL the dS-pawn with c4) 12 . . .• f5! 13 Wd4 cxdS 14 lOgl .f7 IS .t.e3 lOc6 16 .d2 d4! 1 7 .t.xd4 lOxd4 1 8 .xd4 .t.ffi 19 � (trying to improve on 19 .d2? .t.xb2 20 Zlb 1 .02 =1= Trull-Dreibergs, corr. 1953) 19 . . . aS 20 .a3 �gS! 2 1 f3 Zle8+ 22 �e2 �f4 23 lOC4 dS 24 1Oc3 1Whs, the white king is uncomfortably stuck in the centre, Knorr-Dravnieks, corr. 1992 .

•••• lDc6!

9 .t.e3 White has no better than to play

this and take the pressure off e4. as the other moves offer little:

a) 9 lOxe4?! lOxe4 10 fxe4 (10 .xe4+ .xe4+ 1 1 fxe4 lOxd4 12 .t.d3 �e6 13 lOe3 �e7 14 c3 1Oc6 is nice for Black, Jurgenson­Tiemann, corr. 1987, 10 c3 looks tempting, after all the e4-knight is still pinned, but following 1O . . . dS 1 1 Illd2 .t.e7! 12 fxe4 0-0, . . . .t.h4+ is a threat, 1 3 gl �g4 and Black has a strong initiative) 1O . . . lOxd4 1 1 .dJ 1Oc6 12 .t.d2 �e7 13 0-0-0 0-0 is, once again, comfortable for Black, Fabian-Pupols, Philadelphia 1974.

b) 9 dS!? is very wild, but I don't really believe it. . .

9 . . . lCJd4 10 .dl ( 1 O .d2 exO 1 1 .xd4 merely transposes) 1 0 . . . exO 1 1 .xd4 fxg2 12 .*.xg2 .xg2 1 3 :gl .xh2!? (it is probably best not to take the c-pawn, although this is also played: 1 3 ...• xc2 14 lObS! [ 1 4 .*.f4 .*.f5 I S :g3? O-O-O?! 16 .xa7 ICIg4 17 :dl?! :eS+ I S � , Destrebecq-Gaard, corr. 1993, when I S . . . :e4!, Downey, wins] 1 4 . . . WdS I S .*.gS .*.e7 16 :c 1 .xh2 1 7 ICIxc7! Wxc7 I S ICIxd6+ Roscher­Pollaod, corr. 1992, and I S . . . Wd7! is best, 19 .a4+ bS 20 .xbS+ Wxd6 2 1 .c6+! WeS 22 .i.xf6+ gxf6 23 .c3+ forces a draw) 14 .*.f4 ( 14 ICIbS? meets the rejoinder 14 . . . fudS!) 14 . . . �S ( 14 . . . �4+ I S Wd2 h6 16 :h l .g4 1 7 :agl • f5 is not too clear, Destrebecq­Downey, corr. 1993, White will play his king to safety on c l , and then attack) I S Wd2! (again the right idea, 15 ICIbS?! WdS 16 .*.gS .*.e7 fails to solve White's king problem, 0-1 , Reinke-Clarke, corr. 1994I9S) I S . . . .i.d7 16 ltael+ WdS 17 Wc I b6 I S .i.gS WcS 19 a4 Wb7 and. having also brought his king to relative safety, Black could start ex­changing some pieces, Oren-Krantz, corr. 1994/97.

c) 9 ICIbS?! seems to confuse two plans, Borochow-Mlotkowski, Los Angeles 19 19, continuing 9 . . . WdS 10 .i.e3 a6 1 1 lOc3 exf3 1 2 gxfJ bS

Main line: other seventh moves 43

1 3 lCJd2 .xc2 14 :bl .i.f5 with a clear black advantage.

9 •• .,i,e7 10 0-0-0 0-0

1 1 dS Gaining space and displacing the

c6 knight. 1 1 .d2?! is worse: I I . . .exO 1 2 gxO WhS?! 13 :g I .f7 14 Wbl a6 I S .g2 bS 16 lCJd2 .i.f5= Zschorn-Hayward. corr. 199 1 .

1 l ••• 1CIb4 11 a3 .S!? A sharp piece sacrifice. opening

up the a-file, but 12 . . . 1C!a6 is quite adequate: 1 3 :g I .i.d7 14 lCJd2 1.f5 I S 1IfbS lOcs 16 b4?! c6. unclear, Hempel-Castelli, corr. I 970n I .

l3 ub4 ub4 14 ICIbl :.1 IS IC!cd1 ed3 16 pfJ lClxdS I7 IC!e4

Black also has compensation after 1 7 1C1b3 lCIxe3 I S lCIxa I .*.g5 .

17 ••• lCIxe3 18 .xe3 .i.e6 19 :11 .f7?!

19 ...• f5 would have been better. 10 1C!ed1? Missing 20 ICIgS .*.xgS 21 .xgS

with an edge: 2 1 . . ..*.a2 (or 2 1 . . .:faS?! 22 .i.d3 : laS 23 ..... g6 24 :gS, Hempel-Gunderam, corr. 1970, but 2 1 . . ..f4+ is better) 22 Wd2 .f4+ (note that 22 . . . Lb I?? 23 Lb I .*.xb I loses to 24 .i.c4) 23 .xf4 :xf4, Hempel­Grivainis, corr. I 970n2.

10 • • • .*.f6 II .i.d3 :al 11 %[de I .*.xb1+ 13 Wdl .*.dS 14 :gS .i.eS 15 :eal :'1 16 :nS dxeS 17

44 Main line: other seventh moves

IVIeS JIa5 18 Wc1 b6 19 .cM .e7 30 .14 Wb8 31 b4 :ea 31 � b3 33 ab3 -*.xe4 34 j,xe4 .eS+ 35 -*.cl .e3+ 36 Wbl .d4+ 37 Wc1 :'1 38 j,e4 �l+ 39 Wdl :"8+ 0-1 Muratov-Spassky, USSR 19S9.

F 7 d5

Another attempt to isolate and surround the black e-pawn, but again Black finds himself with a lead in development which more than compensates.

7 ••• ll)r6 8 J.e3 White would transpose into

Burke-Stobbe, E above, with 8 .e2?! .

8 •• .-*.e7

9 .d4 9 lObs cannot be good, 9 . . . 1&6

10 .d2 (In the correspondence

game, Diepstraaten-Hundley, White went in for the "esoteric I 0 �?! 0-0 I I .d2 when, rather than wast­ing time with I I . ..Wh8?! , Black should have played 1 1 . ../l)g4! ;. 12 0-0-0 would then be impossible: 12 .. .:xt2! 13 .hf2 .i.gS l=. Pawn­grabbing by 10 lOxa7?! presents Black with a whopping initiative af­ter 1O .. . j,d7 I 1 l&3 0-0 12 /l)7bS /l)g4) 10 . . . 0-0 1 1 h3 Wh8 12 a4?! ('positional', but sometimes tactics are more important!) 12 ... lOxdS!? (Good, but so is 12 . . . b6 which in­tends rounding up the dS.:J)8wn by . . . -*.b7) 1 3 .xdS c6 14 Wd2 cxbS I S axbS /l)c7 16 g3 -*.e6 17 ll)aJ ll)dS and Black is in charge, Semenova-Varga, Zalakaros 1993.

9 ... 0-0 10 IMl 10 0-0-0 is treated similarly

10 . . . cS! 1 1 dxc6 lOxc6 12 OO? lOxdS 13 .xdS+ -*.e6 14 .d2? ( 14 �S dS IS .xb7 is the only hope) 14 . . . dS I S ll)aJ ( I S /l)aS -*.b4) IS . . . d4! 0-1 , Pirk-Kulaots, corr. 1991 , White loses a piece: 16 -*.xd4 lOxd4 as 1 7 .xd4? j,gS+ 18 Wbl :ad8 wins the queen.

10 ••• cS! 1 1 du6 If 1 I .c4 then I I .. .a6 12 /l)cxe4

bS 1 3 .d3 j,f5. 1 1 • • • /l)u6 Il .c4+ Wb8 13 0-0-0 1 3 /l)dxe4? dS wins a piece. 13 •• .-*.14 14 f3 d5!

Opening up White's 'position. IS IOxd5 IllxdS 16 .xdS exO 17

pfJ :'c8? A bad slip, 17 ...• xc2+ 1 8 �c2

1llb4+ was possible, when Black has the tiniest of edges in the endgame, but, instead, 17 . . . IOb4! 1 8 -"3 :'c8 19 c3 ,j.e6 was a real crusher.

18 ,j.d3 White is back in the game,

Spielmann-Nimzowitsch, Semmer­ing 1926.

G 7 g3

Another rare idea, but it does look quite sensible, attacking the e4 pawn with the g2 bishop. The nega­tive side of the move is the weaken­ing of the kingside light-squares.

7 ••• lOf6 8 ,j.gl ,j.g4! This is the right time for this

move. 8 . . . dS? 9 1Oe3?! (9 Illes Wf5 10 h4 will win the black queen by j,h3!) 9 . . . c6 10 J.d2 ,j.e7 1 1 .e2 1lla6 12 0-0-0 1llc7 = Lejarza­Purins, corr. 1968/69.

8 . . . j,e7 9 0-0 0-0 (now, 9 . . . ,j.g4 is more simply answered by 10 We I J.O, Mumy-<>Wens, Electro game 1995, after which 1 1 J.xO exO 12 _e6 renders Black's further devel­opment difficult) 10 :Cl .fS ( 10 . . . ,j.g4 1 1 .d2) 1 1 1lle3 .e6 1 2

Main line: other seventh moves 45

dS _n 13 We2 and Black has problems defending his e4-pawn, Dupont-Lescot, corr. 1994.

9 .dl 9 0 exO 10 ,j.xO J.xO 1 1 _xO

Illc6 12 dS Illes 1 3 lOxeS dxeS is good for Black, whose bishop will find a more active post than the habitual e7.

9 •.• 1llc6 10 0-0 0-0-0 I I IlleJ 1 1 dS!? 1lle7 1 2 .d4 cS. 1 I • • .dS Il illel? 12 lOxg4 .xg4 is �ual. Il •• ..*.xel 13 -xel ll)xd4 Winning a crucial pawn, Latch-Kampars, corr. 1966.

H 7 b3

A similar idea to G was seen in the game Rauzer - I1yin Genevski, Leningrad 1936.

7 ••• lllf6 8 g4!? 8 ,j.e3 transposes to line C, and 8

,j.f4 to line A. 8 •• .J.e7 9 J.gl The g-pawn can also continue on

its route, 9 gS!? lDfd7 10 j,e3 (10

,j.g2 .*.xgS 1 1 .*.xe4 �6 12 .hgS .xgS 13 WO can be met by 13 . . . llfll 14 .e2 �d8 unclear) 10 . . . c6 11 dS 0-0 12 .d2 lDb6! as

the e3-bishop is required on the c l -h6 diagonal to defend the g­pawn. 13 0-0-0 with sharp play.

GUrd-Zschom. corr. 1989.

46 Main line: other seventh moves

9 ••• 0-0 10 .i.e3 c6 1 1 d5 And now, pinning the dS-pawn by 1 1 •••• 17 gives Black good chances.

I 7 b4!?

I thought that this was an idea of Grandmaster Yaacov Murey, one of the most original players in the world. He mentioned it to me over dinner one evening after a club match (we both played for Auxerre, in the French League), however I have since discovered that none other than the then World Cham­pion Alekhine played it twice in a simultaneous display in Lisbon, 1940!

7 ••• �f6 Others: a) 7 . . . J..g4?! Several of Murey's

games against a com�uter opponent went like this, 8 hS! .e6 (8 ...• es 9 J..e2 .i.xe2 10 .xe2 I£lffi 1 1 ItxI2 wins a pawn, although 1 1 . . .1Oc6 1 2 ltXIxe4 �xe4 1 3 �xe4 0-0-0 1 4 �g3 .dS generates plenty of play) 9 dS .i.xd I 10 dxe6 .l.xc2 might be

playable, then 1 1 �S 1t)a6 12 �3 .i.d3 13 .i.xd3 exd3 14 ltXIs with the initiative.

b) 7 . . . h6?! 8 hS 1Wh7 9 f3 exf3 10 .xf3 .es 1 1 .e3+ with a big plus.

c) 7 . . . hS?! 8 .i.gS again with the advantage, the gS square is weak (as Murey pointed out, this is reminis­cent of the main variation of the Lowenthal Sicilian).

8 b5! As played by Alekhine, 8 .i.gS is

how Murey intended to play, putting pressure on e4, 8 . . . .i.e7 9 .d2 0-0 10 0-0-0 with interesting play.

8 •••• 17 9 .i.g5 9 .i.f4?! .i.e6. 9 ••• .i.e6?! No better than 9 . . . .i.e7?! 10 �3!

0-0 1 1 .i.xf6 .i.xffi 12 �xe4 with an extra pawn, however, 9 . . . dS!? might be the best, avoiding problems with the e4-pawn, I 0 �S .e6 1 1 h6 c6 and Black seems to be OK, e.g. 1 2 bxg7 .i.xg7 1 3 .i.e2 �bd7 14 .i.hS+ �xhS I S WxhS+ �f8.

10 d5!? 10 �3 dS 1 1 h6 �bd7 12 hxg7

J..xg7 1 3 .d2 0-0-0 is fine for Black, Mainka-Pape, Bad Mergen­theim Rapidplay 1994, but 10 .i.xffi! i s rather more worrying, 1O . . . .i.xc4 1 1 J..h4! .i.xfl 12 .g4!, threatening mate in one and winning the e-pawn with check.

10 ••• .i.xd5 1 1 i£leJ!? .i.e6 12 .d2 .6? 13 J..d6 pf6 14 �xe4 1Dc6 15 .cl .i.e7 16 I£lg5 fxg5? 17 .xb8+ .i.f8 18 .i.d3 0-0-0 19 .xb7 winning in a few moves, Alekhine­Braumann, simul, Lisbon 1940.

3 Nimzowitsch's variation 6 lbe3

I e4 eS 1 100 f5 3 lOses .r6 4 d4 d6 S �4 rxe4 6 lDe3

It is perhaps not surprising that Nirnzowitsch held this move to be White's strongest, as the blockading knight formed an intrinsic part of his system. Nowadays, however, the line is not much in vogue, prefer­ence being given to sharper systems. Black has two continuations:

A 6 ••• c6 8 6 ••• lOc6 Other moves like 6 ...• g6 are

sometimes played, but often trans­pose into 7 liX3, in the previous chapter, i.e. 7 1Oc3 (7 .i.c4 or to other lines in this chapter, 7 . . . 1Of6 8 0-0 c6 for instance) 7 . . . 1Of6.

A 6 ••• c6 7 .i.c4

Nirnzowitsch's original idea, pro­voking . . . dS, gives White a target for his c4-blow, however, Black has

several ways to achieve equality. White sometimes plays 7 c4 instead, gaining space on the queenside, and, objectively speaking, I believe this is stronger: 7 ...• g6 8 .i.e2 1Of6 9 1Oc3 .i.e7 10 0-0 ( 10 b3 is also a logical choice, 10 . . . 0-0 I I .i.b2, Moskovic-Gabrans, Latvia, when 1 l . . ..i.e6!? [ 1 l . . .lOa6] 12 dS .i.f7 1 3 0-0 IObd7 14 f3 lhe8 I S fxe4 lOxe4 16 .i.d3 .i.gS 1 7 IOfS favours White, Moskovic-Gabrans, Riga 1985. 10 h3?! is not the most pertinent: 10 . . . 0-0 I I .c2 lOa6 12 a3 cS 1 3 dS 1Oc7 14 .i.g4 .i.xg4 I S IOxg4?, a misjudgement, [ I S hxg4 is superior] I S . . . lOxg4 16 hxg4 lhe8 17 .e2 .i.f6 1 8 :h3 .i.d4! 19 .i.e3 .f6 20 .i.xd4? cxd4 2 1 lOcil bS!? 22 b3 bxc4 23 .xc4 e3 24 fxe3 dxe3 25 lh2 .f2+! 26 lOxI2 exf2+ 0-1 , Tefedor-Hector, Manresa 1990) 10 . . . 0-0 1 1 f3 exf3 12 :xf3! (and now I prefer this move to have the possibility of bringing the bishop directly to d3, to molest the black queen; the alternative is 12 .i.xf3 1&6 1 3 .i.e2 [ 1 3 dS cS { 1 3 . . . .i.d7!? intending . . . :ae8, is more flexible} 14 a3 lOci7?! {too ambitious, 14 . . . 1Oc7} I S .i.e4 :xfl+ 16 .xfl with advantage, Pineau-SCnechaud, Avoine 1995] 1 3 . . . .i.d7 14 a3 :ae8 I S .i.d3 WIts 16 .xhS lOxhS 1 7 IOfS?! , Leenes-Yska, Groningen 1984, when 1 7 . . . .i.xfS 1 8 :XfS .i.f6 shows the d4-pawn to be under­defended) 12 . . . 1&6 13 .i.d3 WIts 14 .e2 Harris-Destrebecq, France

48 Nimzowilsch 's variation 6 lOe3

1 977, and now 14 . . . lle8 I S lOfS .i.f8 limits White's advantage.

7 dS 11g6 8 1Oc3 lOffi transposes into Chapter 2.

7 ••• d5 Black is also able to defer this

move: 7 . . .• g6 8 0-0 lOffi, for in­stance, waiting for White to unveil his plan. 9 0 (9 Ab3?! dS 10 c4 .i.d6! [going for a direct KO! 10 . . . dxc4 1 1 .i.xc4 Ad6 12 1Oc3 transposes into a line from Chapter 2] 1 1 cxdS lOg4 1 2 lOxg4 Axg4 1 3 • d2 h6 14 &3 .i.0 I S g3 and White's king was a problem throughout the game, which Black won nicely, Fossan-Eilertsen, RandabelJ 1989) 9 . . . dS 10 Ab3 exO 1 1 .xO .i.d6 1 2 c4 1lf8! 1 3 .i.c2?! (better to remove the queen from the HUe; 1 3 We2! 11e4! 14 1Oc3 is sharp, but promises some ad­vantage, 14 . . .• xd4 IS cxdS .i.g4 16 1Iel ) 1 3 . . . 1Oe4 14 .i.xe4?! dxe4 I S 1Ie2 :Xfl+ 16 1Ixfl lOa6 1 7 a3 lths 1 8 h3 .i.d7 19 1Oc3 1Ig6 with threats, Diepstraten-De Boer, corr. 1982.

8 .i.b3

. � �B ?t1 :I: � � �t1 :I: . _ ' � � -- ,

� :I: � � � � � \w;

I � t � � � �� t � � � r. , � �I

� � � � � ��� tz.J � .a. ¥�. '4 .a. � r� .e. r� .u. �� .u. � �� .u. �.!.S .�� '-" ��.M,ffi � � t:::!f.�� g �'9J � C:l

8 •• ..i.e6 Attempting to reduce the impact

of White's next. Other moves are: a) 8 . . .• g6!? 9 c4 dxc4 10 .i.xc4

(Black's idea, forcing White to cap­ture with the bishop as 10 lOxc4? .xg2) 1O . . . .i.d6 1 1 1Oc3 lOffi 12

a4?! (the immediate 12 O? allows 12 . . . .i.xh2 but 12 0-0 is simple and good, with 0 to come) 12 . . . aS 13 .i.d2 lOa6 14 �3 lOb4 I S O-O-O?! .i.d7 16 1lhel We7! , intending . . . bS with dangerous counterplay, Santo Roman - Destrebecq, Uzes 1988.

b) 8 ... Ad6?! 9 c4 dxc4 10 lOxc4 .i.b4+ (too many moves with this piece is a sign that something has gone wrong) 1 1 1Oc3 .i.e6 12 0-0 .i.xc3 13 bxc3 .i.dS 14 .i.a3 with advantage, Eglitis-Strelis, corr 1979 .

c) 8 . . . bS? 9 lOxdS! (Black's elabo­rate attempt to forestall White's c4 was more successful in Eglitis­Kozlov, corr. 1978: 9 a4 b4?! 10 O-O?! [ 10 lOxdS! is still good] 1O . . . Ae6 1 1 as .i.d6 12 0 �) 9 . . . cxdS 10 WhS+ g6 ( 1O . . . Wd7? is catastro�hic, 1 1 WxdS+ Wc7 1 2 .i.gS! Wd6 [ 1 2 .. .• g6?? 13 .i.d8 mate] 1 3 .xa8 .i.b7 14 1Ixa7 1Oc6 I S .cS ± Destrebecq-Kozlov, corr. 1982) 1 1 1IxdS 1Oc6 12 .i.gS .e6 , 1 3 .xbS!? ( 1 3 .xe6+ .i.xe6 14 dS should be simple enough) 13 . . .• d7? ( l 3 . . . a6! 14 _a4 .d7 I S dS lOes 1 6 .xe4 .f5 ofTers some chances) 14 .i.dS?! (giving Black another chance, 14 .i.a4 .i.b7 I S dS is deci­sive) 14 . . . .i.b7? ( 14 . . . lOxd4! I S .xd7+ Wxd7 16 .i.xa8 lOxc2+ 17 Wd2 lOxa 1 1 8 1Oc3 is good for White, but Black is still playing) 1 S .i.xc6 .xc6 16 .eS+ winning, Preikschat-Jaeckel, Porz 1984.

d) This variation, in general, suf­fers from too few practical trials. It occurs to me that 8 . . . lOh6!? might be playable, with the possible con­tinuations: 9 c4 dxc4 ( 10 lOxc4 1Of5 1 1 lOes .i.e6 or 1 1 . . .1O:I7!?) 10 .i.xc4 1O . . . 1Of5 1 1 dS? .i.b4+ 12 Wfl ( 12 10:12 lOxe3 13 ne3 • gS is little improvement) 1 2 . . . .i.cS 13 .c2?! .i.xe3 14 .xe4+ Wd8 IS .i.xe3 lle8 16 .0 lOxe3+ 17 ne3

.xb2 I S dxc6 .xal 19 .eiS+ �7 � I , Hellberg-Ringoir, Belgium­Norway 1993.

Or 9 0-0 �!? 10 c4 .i.d6 1 1 g3 Wh3 12 0 ( 12 f4 �g4) 1 2 . . . dxc4.

9 c4

9 •••• ., Bolstering eiS, but it might be

better to concede it, 9 . . . dxc4!? 10 .i.xc4 .i.xc4 1 1 �c4 �7 12 0-0 �?! (sharp, but 12 . . .• e6! 1 3 .e2 �ga; is very comfortable for Black) 1 3 �3 .e6 14 1ib3 (threat­ening a nasty discovered attack on the queen, and gaining a tempo) 14 . . . lIeS I S .i.f4 lOga; 16 1lacl and White has a strong attack on the c-line, Destrebecq-Grivainisl Hayward, corr. 1997

9 . . . .i.b4+?! is very popular, but why exchange White's passive bishop? I 0 �3 ( 10 .i.d2 .i.xd2+ I I .xd2 .17 12 cxeiS cxeiS 1 3 0 exO 14 0-0 �6 I S 1Ix0 �a; 16 �3, White has a slight, but persistent, positional advantage, Huewels­OOAT, Compuserve, corr. 1993) 10 .. . �7 I I 0-0 .i.xc3 12 bxc3 0-0 13 0 .g6 14 h3 which must have made Black doubt the wisdom of conceding his dark-squared bishop! Megier-Crimp, Australia 1995.

IO .e1 10 cxeiS may be superior,

1O . . . cxdS 1 1 0-0 lOa; 12 �3 .i.b4

Nimzowilsch 's variation 6 lOe3 49

13 0 .i.xc3 14 bxc3 exO IS h3!? (confining the black king to the cen­tre, at the cost of a pawn) I S ... �? ( l S . .. fxg2 is critical, 16 1le1 �dS 17 .0 lieS I S 1le2 when White has compensation for the pawn, but pos­sibly not enough) 16 .xO �1? ( 16 . . . �) 1 7 Lel lZdS?! I S �f5 winning, De Firmian - Church, Hollywood 19S5.

IO • • • �f6 1 1 0-0 .i.b4?! Black might be more successful

with I l . . .iObd7!?, answering 12 1&3 with 12 . . . � trying to fix the central pawns, and 12 O?! with 12 . . . � 1 3 1&3 .i.d6, but 1 1 . . . .i.d6! 12 �3 0-0 is surely the simplest, with a level game.

U .i.d1 .i.xd1 13 �xd1 0-0 14 f4!? dxc4 15 �u4 .e7 16 f5 .i.d5?!

Conceding a useful bishop, 16 . . . .i.f7 is fine.

17 �xd5 cxd5 18 M .d7? I S . . .• f7.

19 �xd5! �xd5 10 .u4 11d8 11 f6 pf6

Forced, in view of 2 1 .. .� 22 17+ Wt8 23 .xh7 +/-

II Ilf5 �h8 13 IIxd5 Ilea 14 IIxd7 lIu4 15 IId8+ Wg7 16 1lg8+ �h6 17 1ln 1-0

Nimzowitsch-8etin§, Latvia 1919.

50 Nimzowitsch 's variation 6 �3

B 6 ••• �

The modem preference, hitting d4. It was actually first proposed by Betinl after his defeat at the hands of Nirnzowitsch.

7 dS Conceding the cS square is not a

serious problem, but there are sound alternatives:

a) 7 �S .f7 with:

a l ) S .i.c4?!, hoping to win the as-rook but, as often happens in these type of positions, this tends to rebound. S .. .I&S 9 � (9 �ffi+? .xa; 10 �S+ .g6 1 1 .xaS .xg2 12 Ilfl, Golob-Thoma, corr. 19S2, when 12 . . . c6 is best, as 1 3 .i.xgS llxgS 14 �S+ g6 I S .xh7 Ilg7 wins for Black, i.e. 16 ..... .i.h3 17 llXl2 1lf7 menacing . . . e3, and winning) 9 . . . �c4 10 It)XaS b6 (the 1O . . . dS!? 1 1 0 exO 12 0-0 of

Crimp-Destrebecq, corr. 1999, may be playable, but leaves Black further from recovering the as-knight) 1 1 b3 ( 1 1 �3 .i.b7 12 �xc7+ .xc7 1 3 0-0 �a; 14 dS .i.e7 I S .e2 0-0. Black's two pieces are far stronger than the rook, Golob-Van Houten, corr. 19S2, and 1 1 c3 a6 12 b3 �S 1 3 dS .i.b7 14 .g4 WdS! must also favour Black) I I .. .�S 12 .i.d2 �6!? ( 1 2 . . . .i.b7 1 3 .i.xaS bxaS 14 .g4 �h6 is simpler, Black is clearly better) 1 3 dS �S 14 .e2 .i.b7?! (returning a 'piece for the attack, 14 . . . �a; I S WbS+ WdS is messy but good for Black) I S .xe4 .i.xaS 16 f4 �a; 17 .e2 .xdS I S fxeS dxeS ( I S . . .• xeS leads to a promising ending) 19 �3 .d4 20 0-0-0 �4 2 1 .g4?? (losing on the spot� 2 1 �b I defends, and leaves White in the driving seat) 2 1 . . ..i.a3+ 22 Wb I �xc3+ 23 .i.xc3 .xg4 0-1 , Smit-Diepstraten, Hilversum 1991 .

82) S lObe3 .i.e6 9 �c7+?!

!. � � " i ·� bB ·1101.� ... 'if.'Y .""�� ... �� ... -!"Z-J � - i�� ... �

.. ..... ? .";.J..{:@' .. �

.. . W�. �"'P� � �

� � � � �. t�J "' W@i � � p.u.:, .£ � mJfa &. 'rrv' - �" '� � � � � it �� it � �� it �j'[ .�� � �.M,Z '" � � � � � h t9" � >� b

(this theoretical continuation al­most loses by force, so perhaps White should prefer 9 �f4) 9 . . .• xc7 10 dS, now the theoretical choice is 1O . . . �a; ( I0 . . . J.fS!?, un­usual, but quite reasonable, 1 1 dxOO [ 1 1 �bS .d7] I I . . .bxc6 12 0 1Oa;!? 13 fxe4 [13 g4!?] 1 3 . . . �e4 14 .O! �xc3 I S bxc3 [ I S .xfS .e7+] I S . . . g6 16 .i.d3 .i.g7 17

.i.xf5 when, instead of 1 7 . . . 0-0?! sacrificing a piece for doubtful at­tacking chances, Torrecillas-Hector, Barcelona 19S5, 1 7 . . . gxf5 was per­fectly adequate, I S 0-0 0-0) I I dxe6 dS! (-1+ ECO), and now:

a2 1 ) 1 2 lOxdS "as+ 1 3 lik3 .i.b4 14 .i.d2?! (but 14 .i.e2 .i.xc3+ I S bxc3 lldS [ I S . . ... xc3+ 16 .i.d2 .eS should be OK for White] 16 .i.d2 "dS 1 7 llb I and White can hold with precise play) 14 . . . 0-0-0 I S "cl ? (Betin! orisinally analysed I S .i.e2 llxd2? 16 Wxd2 lldS 17 .e3 ltlds I S 1I'h3 IOf4, overlooking 19 "g4! .i.xc3+ 20 �fl and the threat of e7+ allows White to win the ex­change. Instead, I S . . . .i.xc3 16 bxc3 lld6 is quite satisfactory) I S . . . llxd2! 16 "xd2 lldS 17 "e3, ltldS? (ac­cording to the Betin! recipe, but 1 7 . . . lOg4! I S 1I'h3 e3 cuts the 'J!leen's defence of c3, and wins) I S Wh3 �f4 19 "e3 lOd3+! 20 .i.xd3 .i.xc3+ 2 1 bxc3? (2 1 �fl .i.xb2 22 llel [22 llbl?! exd3 23 e1? dxc2! ! 24 exdS="+ "xdS wins] 22 . . . exd3 23 cxd3 is unclear) 2 1 . . ... xc3+, winning, Reybum-Downey, corr. 19S0.

a22) 12 .i.gS 0-0-0 1 3 .i.bS d4 14 .i.xc6 ( 14 .i.xffi gxffi IS .g4 lOes 16 .f5 �bS 1 7 �xe4 .as+ I S �dl "xbS 19 .xffi .i.e7 ! ! 20 .xe7 d3 2 1 �cl dxc2 22 ltld6

Nimzowilsch 's variation 6 lOe3 51

lOd3+ 0-1, Pretto-Ortiz, corr. 1994, as 23 �c2 .cS+ 24 �d3 llxd6+ 25 �e2 .c4+ leads to a swift end) 14 . . .• xc6 I S .i.xffi gxffi 16 e7 .i.xe7 1 7 .g4+ �bS IS lOxe4 .i.b4+ 19 �d I llheS with an attack, Stockholm-Riga, 1934/36.

a3) S c4 logical, and best, White simply gains space:

S . . . .i.f5 (or S . . . lOffi 9 lOxffi+ [9 IObe3 .i.f5 transposes] 9 . . .• xffi 10 .i.e3 .i.f5 I I �e2 0-0-0 12 0-0 Wright-Boisvert, corr. 1973, with an edge to White) 9 lObe3 O-O-O!? (9 . . . �ffi 10 .i.e3 �xdS [ 10 . . . 0-0-0 transposes] I I cxdS lOe7 12 �3 [ 12 .a4+] 12 . . . 0-0-0 1 3 llc I Destrebecq-Downey, corr. 1997, 1 3 . . . a6!? is playable.) 10 .i.e3 (alter­natively, 10 b4!? has a.sgressive in­tentions, 1O .. . �ffi 1 1 &3 .i.g6 12 llbl [ 1 2 bS lOe7 1 3 .a4 is more di­rect] 12 . . . dS!? 1 3 cS �bS 14 .i.e2, Hanisch-Schneider, corr. 19S7, White's queenside build-up is mak­ing life very uncomfortable for Black. 10 .i.e2 IOffi 1 1 0-0 lOxdS 12 cxdS!? lOe7 1 3 .a4 �S, Schwartz-Dreibergs, corr. 1961, 14 f3 exf3 IS llxf3 with a plus, or 10 lObS lld7) 10 . . . lOffi 1 1 .a4 a6 (other than 1 1 . . .�S?! 12 lObS! [ 1 2 IOxffi .xffi 13 dS lOd4 14 0-0-0 cS I S dxc6 lOxc6 16 lOds is also very difficult for Black, Destrebecq-

52 Nimzowilsch 's variation 6 lOe3

Svendsen, COrT. 1997] 12 . . . lOes 1 3 �c7! lOxc7 14 dS lOxbS I S cxbS lOes 16 .xa7+ WcS 1 7 llc l+ Wd7 I S Jlc7+! Wxc7 19 b6+ Wd7 20 .i.bS+ We7 2 1 .i.gS+ analysis by Melchor.) 1 2 b4 lllbS?! 1 3 bS lOxdS 14 cxdS .eS I S lib I .i.d7 16 bxa6!? .i.xa4 1 7 a7, regaining his queen, with advantage, Destrebecq­Clarlce COrT. 1997.

b) 7 .i.bS

7 . . . a6 (I am not too keen on 7 . . . s.cJ1?, White can play S IlldS! [rather than S .i.xc6? .i.xc6 9 c4 .g6 10 0-0 .i.e7 1 1 lOc3 lOffi, Wasennan-Dreibergs, USA 19S7, and Black is doing fine] when S . . .• dS �ive, but probably forced as S ...• g6?! 9 lOxc7+ WdS 10 lOxaS .xg2 I I lln Illffi looks completely insufficient] 9 .e2 lOffi 10 IOxffi+ .xffi 1 1 dS lOd4 12 .xe4+ grabbing a pawn) S .i.xc6+ bxc6 9 dS (9 c4 _g6 10 illc3 ( 10 dS can now be met by 10 . . . cxdS as I I .xdS llbS and the e4-pawn is de-fended] 10 . . . lllffi I 1 0-0 .i.e7 12 • c2 0-0 [ 1 2 . . . Wfi!?] 1 3 f3 exf3 14 .xg6 hxg6 I S llxf3 .i.e6 and Black's bishop pair provides com­pensation for the structural weak­nesses, Krempl-Sireta, COrT. 1995. 9 0-0 .g6 10 f4!? lOffi I 1 f5 .n 1 2 c4 .i.e7 = Donoso-Cannona, Chile 1974) 9 . . . cS (9 . . . cxdS? 10 .xdS

wins the e4-pawn, but 9 . . . .i.b7 is a possibility, avoiding the fixing of the black queenside pawns) 10 c4 ( 10 illc3 .g6 1 1 0-0 .i.e7=) 1O .. .• g6 I I .a4+ ( 1 1 0-0 .i.e7 12 illc3 lOffi 13 lOe2 0-0 14 .c2 lOg4!? with interesting play, Malmstrom-Budovskis, COrT. 1994/9S) 1 l . . ..i.d7 12 .as lOffi!? 13 illc3 (White's queen is misplaced after 1 3 .xc7 .i.e7 14 0-0 0-0 but this is clearly critical) 1 3 . . . .i.e7 14 lObS? 0-0 IS illc3, handing Black two useful tempi, Johansson-GUrci, Sweden 1994, as I S lOxc7?? s.cJS 16 � llcS loses a piece.

c) 7 c3 is solid, and reasonably common in om games, 7 . . .• g6 S .i.e2 lOffi 9 0-0 .i.e7 10 illd2! 0-0 I I f3 exf3 12 lOxf3 WhS 1 3 lOh4 .eS 14 lOefS with a plus, Pantaleoni-Deslrebecq, Bagni di Luca 1979.

7 ••• lDe5

8 .i.el Stopping the black queen from

reaching its ideal square on g6 . a) Alternatively, S .i.bS+!? is

inaccurate, S . . . OO! (S . . . .i.d1?! 9 .i.xd7+ lOxd7 10 lOc3, when Black should reply 10 . . . lOcs ;I; keeping control of d4, rather than 10 .. .• g6?! I I .d4 when Black has the choice between losing his e-pawn, or allowing lObS: I I . . ..i.e7 12 .xe4

with advantage, Frommel­Dravnieks, 1988) 9 dxc6 bxc6 10 .i.e2 (10 .dS!? is more enterpris­ing, 10 . . . .i.d7 1 1 .xe4 IIb8 [not I LdS?! 12 �dS! .d6 1 3 lDdc3 1Df6 14 .e2 cxbS I S .i.f4 regaining the piece, with interest) 1 2 .i.e2 dS 13 .a4 .i.cS with plenty ofactivity) 1O .. . dS 1 1 �dS!? ( 1 1 �O .i.d6 is pleasant for Black, with his strong centre) I LcxdS 12 .i.bS+ .i.d7 1 3 .i.xd7+ �d7 14 .xdS IIc8 and White can get three pawns for his piece, Pintado-Busom, corr. I99S.

b) White has an interesting alter­native in 8 1De3 .g6 9 .d4 (9 .i.e2 1Df6 10 �O .i.e7 1 1 IDbS .i.d8 12 1lXJ4 � 1 3 f3 cS?! 14 dxc6 bxc6 I S fxe4?! [ I S f4! ) I S . . . .i.b6 with ac­tive play, Grey-Zemitis, USA 1974) 9 . . . 1Df6 10 IDbS �d8! (for tactical reasons this is best, 1O . . .• f7?! 1 1 IDxa7 .i.d7 12 IDbs IIc8 1 3 .i.e2 .i.e7, Black has some compensation for the pawn, but not enough, Serpi­Negrini, corr. 1994) 1 1 lllxa7 .i.d7 12 c4 ( 12 IDbS? reveals the point of no longer having the black king on e8: 12 . . . .i.xbS! 1 3 .i.xbS .xg2! [this combination was first given by Betim) 14 1Dxg2 1Df3+ I S �dl IDxd4 16 .i.c4 bS 1 7 .i.fl IDxdS with advantage, Lawrence­Andersen, corr. 1998) 1 2 . . . .i.e7 1 3 h3 ( 1 3 .i.d2?, cutting otT the queen's escape route, 1 3 . . . lDfg4 [a good move, with the right idea, but 1 3 . . . 1De8! is a killer, 14 lDdl { 14 IDbS .i.f6 wins the white queen} 14 . . . .i.h3 ! ! winning) 14 1Dx�4 1Dxg4 I S .i.e3 .t.f6 16 .d2 Illxe3 1 7 • xe3, Koemskoj-Manaspoetra, corr. 1990, when 17 . . . .i.xb2 1 8 IIbl .f6 is emphatic) 1 3 . . . 1118 14 IDbS .i.xbS I S cxbS IDfd7! (intending . . . lDes-dJ) 16 .i.e2?! IDes 1 7 �O lbb3 (and did I mention IDeS-b3, as well?) 1 8 axb3 IIxa I with a sound

Nimzowitsch 's variation 6 1De3 53

extra exchange, Pape-Niemand, corr. 1994196.

8 •••• 17 9 c4

The positional line, White can also go after the e4-pawn a&ain, much as in the previous note: 9 1llc3 1Df6 1O .d4 ( 10 h3 .i.e7 1 1 g4?! is inappropriate, I I . . . � 12 gS?! IDfg4! ouch! 1 3 IDxe4?! [ 1 3 �g4 .i.xg4 14 f4 is forced, if somewhat gruesome for White) 1 3 . . . 1Dxf2 14 1llxf2 .xf2+ I S �d2 .i.xgS � I , Evans-Niemand, corr. I 994/9S) 10 . . . cS! ( l 0 . . .• g6 is playable, al­though it does leave Black a tempo down on the line 8 1De3 .g6 9 .d4, as he has played . . .• f6-f7-g6, instead of ...• f6-g6, 1 1 lbbS �d8 [ 1 1 . . .cS? 1 2 dxc6 IDxc6 1 3 .d I forces the ugly move 1 3 . . . �d7 as 1 3 . . .• f7? 14 lDc4 blows Black apart, Littleboy-Svendsen, corr. 1985) 12 .c3!? [ 1 2 lbxa7 .i.d7 13 c4 certainly should be less risky now) 1 2 . . .• 17 13 �3 c6?! [ 1 3 . . . .i.d7) 1 4 �a7 �dS IS IDxdS .xdS 16 � is promising for White, Kozlov-Tiemann, corr . 1989) 1 1 dxc6 ( 1 1 .a4+ .i.d7 12 .i.bS a6 13 .i.xd7+ .xd7 14 .xd7+ Wxd7 I S a4 lIe8?! [ I S . . . g6 16 as .i.h6 looks right, this Benoni-type endgame is fine for Black, with the caveat that his e4-pawn can be a problem) 16 as g6 17 lla4! IDf7 18

54 Nimzowilsch 's variation 6 lOe3

b4! White has an edge, Kozlov­Destrebecq, corr. 1992; 1 1 �bS+ �d7 12 .a4 a6 amounts to the same) 1 l . . .�xc6 1 2 �bS ( 1 2 .a4?! dS 13 �bS �cS 14 0-0 0-0 I S c3 .i.d7 is great for Black, who will soon turn his attention to White's kingside, Sprenkle-Magee, USA 1998. But the 'unintuitive' 1 2 .dl ! �e6 1 3 f4! is superior-White must try to restrain the black d-pawn) 12 . . . .i.d7 13 .a4?! (once again White loses control ofdS, 1 3 .dl is best) 1 3 . . . dS 14 0-0, Sadeghi-GUrd, corr. 1990, 14 . . . �d6 with a wonder­ful position.

9 � �f6 10 c4 transposes, as does 9 0-0 �f6 10 c4 �e7 I I � .

9 ••• �f6 10 lDdl! In order to play 0, White must be

ready to recapture on 0 with a knight.

10 ••• �e7 In the game Graber-Leisebein,

1986, Black essayed the radical 1O . . . hS!? I I 0-0 gS 1 2 0 exO 13 �xO �O+ 14 lIxO g4 I S lift �d7 16 b4 .e7 and . . . 0-0-0 with double-edged play.

1 1 0-0 0-0

Il f4!? This seems more accurate than 12

0 , when Black can also continue 12 . . .• g6!1 1 3 fxe4 �fg4! (How­ever, 1 3 . . . �xe4?? is disastrous: 14 lIxfS+ WxfS I S �hS, deflecting the queen from its defence of the hap­less e4-knight, Van Vuuren-Van Bree, corr. 1982) 14 lIxfS+ WxfS I S �g4 �g4 16 �01! (White should try to obtain some control over eS by 16 �O �f6 1 7 .c2 IOes when Black i\as minimal com­�tion for the pawn) 16 . . . �f6 1 7 Illft Wg8 18 lib I IOes 19 lDe3 �d7 20 �d2 lIfS with a firm grip on the dark squares, as in Destrebecq­Elburg, corr. training game 1993.

1l . . • esO 13 �sO

13 ..•• g6?! Both 1 3 . . . �xO+ 14 �xO �d7

and 1 3 . . . Whs 14 ltXI4 ( 14 �eS .xeS) 14 . . .• g6 I S lDers Ad8 re­quire investigation, although in both cases White retains a small advantage.

14 �! .eB 15 �f5 Tasc R3o-Mephisto Berlin, corr.

training game 1994, when 15 •• id8 is best.

4 Bronstein's variation 6 i.e2

I e4 e5 2 li:)f3 f5 3 li:)se5 .f6 4 d4 d6 S liX4 fse4 6 .i.e2

In his excellent book 200 Open Games, David Bronstein goes into a deep discussion of how he came to discover the strong move 6 .i.e2 (which is odd, as it had already been analysed by BetinS). In essence, his idea is to stop Black's queen from settling on the important square g6, from where she simultaneously sur­veys g2, protects the e4-pawn, and also allows the king's knight to de­velop naturally on f6. The other ad­vantage of 6 '*.e2 is its flexibility: White can attack the black e-pawn with lDc3, and, if necessary, with 0 and yet is ready, should Black play .. . d5, to play ltle3 and c4.

Black has tried many replies:

A 6 ••• liX6 B 6 •••• d8 C 6 ••• dS?! D 6 •••• f7?! E 6 ••• bS?

... without ever finding one that is completely satisfactory.

A 6 ••• llX6

Immediately counter-attacking the white d-pawn.

7 dS This ' isolates' the black e-pawn,

whose subsequent defence causes the second player no end of prob­lems, and is really the only move worth considering, 7 c3 being rather too passive, and allowing Black an easy game, 7 . . .• f1 (7 . . . d5?! gives White a talJet, 8 ltle3 .i.e6 9 0-0 <ro-o [9 . . . .i.d6?! 10 0 exO I I .i.xO leaves Black in trouble, 1 1 . . .... 6 12 g3 <ro-o 1 3 li:)f5 .ffi 14 li:)xd6+ IIxd6 15 .i.xd5 and White has an extra pawn, and the better position, Goldgewicht­SCnechaud, Parthenay 1992J 10 0 � 1 1 fxe4 dxe4 12 .i.g4 .i.xg4 13 .xg4+ .xg4 14 �g4. White has an edge in this endgame because of the weakness of the e4-pawn, Bademian-Burgos, Mar del Plata

56 Bronste;n 's variation 6 s'e2

1993) 8 0-0 �fti 9 f3 exf3 1O :Xf3 s'e7 1 1 s'gS ( 1 1 lbe3 dS 12 �f5 gains the bishop pair) 1 1 . . .s'g4 12 Zlf4 �xe2 1 3 Wxe2 ()-(H) which is close to equality, Johner-Grob, corr. 1940.

7 ••• lOe5 Much the best square for the

knight, although 7 . . . lDd4?! has been seen: 8 1tX3! S,f5 9 lbe3 (9 �hS+ g6 10 s'e3 is temptin.8, but less �ood, 1O . . . s'g7! [ 1 0 . . . lOxc2+? 1 1 _xc2 gxhS 12 .a4+ is awkward] 1 1 �e2 [the obvious 1 1 s'xd4 _xd4 12 .xd4 s'xd4 13 lObS en­counters 13 . . . s'g7! 14 fuc7+ Wd7 15 fua8 gxhS when the knight is trapped] I l . . .lOe7?? [ I I . . .fue2 12 Wxe2 .f7 is fine for Black] 12 s'xd4 _xd4 13 .xd4 s'xd4 14 lObS winning material, Levene­Lynn, Dunedin 1983) 9 . . . lOxe2 (9 . . .• eS 10 fuf5 fuf5 1 1 s'bS+ may be objectively better, but is still difficult for Black, Watson-Pascute, corr. 1998) 10 .xe2 ()-(H) 1 1 0-0 ( 1 1 fue4! �xe4 12 _g4+) 1 1 . . . s'g6 12 lOxe4! WeS (as 12 . . . �xe4 is again met by 1 3 _g4+) 1 3 1tX3 lOfti 14 Wc4 wb8 when, instead of 1 5 lObS Zlc8 16 f4?! .e4 with some chances for Black, White would have retained a clear advantage, in the game Oren-Rittenhouse, corr. 1993, by the simple 15 s'd2 Zld7 (else lObs and s'c3) 16 Zlae I .

i � �li .l1 t �� � �:6 t

� }':t4 �. '� � . . ' < �

� � � � � ft � � .'-"'� t � � ��� � �

'� � � � ,� � � � .Q. '9'� .Q. � 1\ "'� .Q. ""$' � �� � ���� � �� �i '-"' � '��r.:1 � � t:::5J '"Z-.J �. ;.: 't:9' � W'� �

8 0-0 Alternatively: a) 8 fueS .xeS 9 0-0 (Obvi­

ously, 9 s'bS+?! is met by 9 . . . c6 10 dxc6?? .xbS, but, nevertheless, White can continue here with 9 �3 lOf6!? [I believe that Black would be better off avoiding the annoying check on bS, perhaps 9 . . . s'd7!?, or 9 . . . c6] 10 .i.bS+ Wd8 [ IO . . . wn!?, hoping to 'castle' by hand] 1 1 s'e3 a6 1 2 .i.e2 s'e7 13 0-0 S,f5 14 �d4 .f4 1 5 f3 e3 16 g3 .gS 17 .*.d3 Ziecher-Urban, Bundesliga 1989, when 1 7 . . . fudS 18 fudS s'xd3 19 .xd3 .xdS keeps White's advan­tage down to a minimum) 9 . . . �fti (9 . . . .i.f5?! transposes to the note to the next move) 10 c4 g6!? (the game Nilsson-Korhonen, corr. 1988, continued most bizarrely: 10 . . . hS? 1 1 f3 gS?! In for a penny! 12 1tX3 h4? 1 3 f4! .e7 14 fxgS 1-0) 1 1 1tX3 s'g7 12 f3 0-0 13 f4!? .e7 14 .i.e3 hS!? ( I4 . . . cS immediately may be the most accurate) 1 5 h3 a6 16' :c I .i.f5 17 a4 cS with a fairly level position, Elburg-senechaud, corr. 1994/95.

b) There is a further option in 8 1tX3, putting immediate pressure on e4: 8 . . . lOxc4 9 s'xc4 .g6! (this is more interesting than 9 . . . s'f5?! 10 s'bS+ when the black king must move, as 10 . . . .i.d7? loses to 1 1 s'xd7+ Wxd7 12 .g4+) 10 s'bS+ (if 10 O-O? s'h3 picking up the ex­change, although White has a fair deal of compensation: 1 1 g3 .i.xfl 12 .xfl ()-(H)!? [there is nothing drastically wrong with 12 . . . lOfti 1 3 .i.bS+ Wd8 for Black] 1 3 s'e3 Wb8? [this loses by force, although I do not really blame Black for missing White's next; 1 3 . . . a6 14 s'xa6 bxa6 1 5 .xa6+ Wd7 16 .a4+ We7 17 fue4 is dangerous, but Black does own an extra rook!]

14 .b6! A bolt from the blue! 14 . . .• eS I S .i.b5 .e7 16 .i.c6 a6 1 7 .c4 1-0, as 1 7 . . . �cS I S � and mate in five more moves, Huldin-Rantalainen, COlT. 19S3) 1O . . . �f7! ( 10 . . . .i.d7?! 1 1 .i.xd7+ �d7 12 0-0 �ffi transposes to a later note) 1 1 �fl?! �ffi 12 h3 .i.e7 1 3 �2 1118 14 .d4 �gS and the respective positions of the two kings speaks volumes, Browning-Sadeghi, COlT. 199 1 .

c) The popular S �3 transposes to Chapter 3.

8 ... �][c-t Black can thereby bring his queen

to the desired square g6, and develop his gS-knight to ffi. The alternatives:

a) S . . . .i.f5? ! seems the most obvi­ous but 9 �xe5 (9 �3 is also quite good, as 9 . . . 0-0-0 10 .i.e3 leaves the black king poorly defended, 10 .. . �bS 1 1 .d2 �c4 12 .i.xc4 .g6?! [ 1 2 . . . �7) 1 3 .d4 b6 14 a4 and White will crash through on the a-file, Canal Oliveras-Clarke, COlT. 1999) 9 . . .• xe5 (9 . . . dxe5!? needs to be examined, although 10 .i.b5+ c6 [ 1O . . . �dS!?) 1 1 dxc6 bxc6 12 .i.g5! .c6 1 3 .la4 certainly favours White) 10 .i.b5+! (again, 10 1Oc3 is a gOod alternative: 10 . . . �ffi [ 10 . . . 0-0-0 is a risky try) 1 1 f3! exf3?! 12 lhf3. White threatens lIe3, the black queen has nowhere

Bronstein 's variation 6 .i.el 57

to go, 12 . . . �g4 1 3 .i.f4 .ffi 14 �b5 �d7 I S lOxc7 Elburg-GUrd, COlT. 1994/95, and Black is losing, for 1 5 . . . �c7 16 IIc3+ WdS 17 .i.xg4 regains the piece) 10 . . . OO? (this loses by force, 10 . . . wn is not much of an improvement as 1 1 f3 �ffi 12 1Oc3 plans to expose the black king on the f-fiIe; 1O . . . �dS! is always awkward, but is best, 1 1 1Oc3 �ffi 1 2 LI [for once, 1 2 f3?! is not very effective, 12 ... exf3 1 3 lhf3? .i.g4) 12 . . . a6?! [driving the bishop to a good square, I prefer 12 . . .... g6) 1 3 .i.d3 Wd4 14 .i.e3! � 1 5 a3 .xb2 16 .i.d4 exd3 1 7 :'2 winning the queen for insuffi­cient compensation, Tatlow­Gocdhart, COlT. 1995) 1 1 f4! De­flecting the queen from its defence of c6, or opening the c-file, I l . . .exf3 12 lie I t2+ 13 �t2 .i.c4 14 �3! (this is simpler than 14 dxc6 .c5+ I S �g3 Wxb5 16 lhc4+ .i.e7 17 cxb7 .xb7 IS �3 �f6 19 IId4 0-0 which leaves White's king somewhat exposed) 14 .. .... f5+ ( 14 . . . �ffi 15 dxOO Wc5+ 16 J.e3 [ 16 �g3!? "'e5+ 17 .i.f4 �h5+ IS �t2 .xf4+ 19 �gl also seems very strong) 16 . . .• f5+ 17 �gl when Black was lost, e.g. 17 . . . bxc6, Rittcnhousc-Hcap, COlT. 1992, I S �xc4! 1Wxb5 19 �d6+ .i.xd6 20 .xd6) I S �g I 0-0-0 16 lhc4 cxb5 17 1Wd3 �7 IS .i.e3 a6 ( I S . . . WbS 19 �b5 a6 20 .i.a7+ �aS 2 1 .i.b6 1-0, Downcy-Clarke, COlT. 1997) 19 .i.b6 11d7?! 20 lhe7! 1-0, Svendscn-Clarke, COlT. 1999, 20 . . .• xd3 2 1 LS+ and mates.

b) S . . . .i.d7?! 9 1Oc3 .f5 (9 . . . �xc4 10 �c4! .e5 1 1 .i.xc4 wins a pawn for nothing) I 0 �3 .f4 1 1 .d4 �ffi 12 �g4 .f5 13 �ffi+ removing the c-pawn' s defender, Rittcnhousc-Grivainis, COlT. 1992.

58 Bronstein 's variation 6 �e2

c) 8 . . . hS!? 9 lik3 �fS 10 lOxeS dxeS ( 1O . . .• xeS 1 1 �bS+ Wd8 12 L I �ffi 1 3 0 h4 14 �gS �e7? I S f4 1 -0, Kozlov-Leibson. corr. 1979, was positively disastrous) 1 1 �e3 ( 1 1 �bS+ Wd8 12 LI .g6 is not too bad for Black) I I . . ..i.d6 ( 1 1 . . ..g6!?) 12 �xhS+!? risky, 12 . . . g6 13 1.g4 ( 1 3 �e2? � 14 h3 �xh3 wins) 1 3 . . . � 14 h3 �ffi I S O, Rittenhouse-Melchor, corr. 1992, and now IS . . . exO 16 �xf5 gxf5 17 .xO :tIS is not too clear.

9 �sc4

9 •••• g6 The move 9 . . .• g6 was chosen by

Black in the beautiful game, Bronstein-Mikenas, Rostov 194 1 :

9 . . . �e7 should be similar, or may even transpose, if White plays �bS+, 10 &3 ( 10 J.bS+ Wd8 1 1 1&3) 10 . . .• g6 1 1 .e2 (again 1 1 �bS+ is reasonable and 1 1 0, although apparently unplayed, deserves serious consideration, e.g. I l .. .exO 12 .xO �ffi 13 �d3 .g4 14 .e3, vel! reminiscent of Chapter I) I l .. .tllffi 12 �bS+ (ganging up on the e4-pawn by 12 lIel?! Kozlov-Miiller, corr. 1979, is less precise, as 12 . . . �h3 13 f3 O-O! 14 �xe4?! [14 fxe4? �g4] 14 . . . lOxe4 I S fxe4? [ I S �d3 J.h4 16 �xe4 .xg2+ is only slightly better for Black] I S . . . �h4 wins for

Black) 12 . . . Wn 1 3 0 a6! ( l 3 . . . exO is slightly less accurate, 14 .xO �g4 [ 14 . . .• xc2?? loses to I S �d3] I S .f4 :tIts 16 Ad3 �S 17 �e3 Wg8 1 8 .c4 J.d8 19 lIae I , White has more space, Svendsen-Melchor, corr. 1997) 14 '*'.4 bS?! (too weak­ening, 14 . . . exO I S .xO IIfS is al­most level) I S .i.b3 exO 16 .xO lIe8 17 �2! (e6 and c6 beckon) 1 7 . . . Wg8 18 �f4 .n 19 .*.d2 �g4 20 lIael 1.ffi 2 1 � �S? 22 IIxeS! dxeS 23 �xc7 winning quickly, Silavs-Cuba, Latvia 1963.

IO �b5+ Tempting, and almost unani­

mously played, yet 10 0 is also extremely logical, 10 . . . exO ( IO . . . �ffi?! 1 1 fxe4 J.e7 [the pawn can't be recaptured, I I . . .�xe4? 12 .*.d3 wins a piece] 12 .*.d3 .g4 1 3 .*.bS+ and Black's compensation for the pawn is non-existent, Genius S-Comet test game 1996) 1 I .xO �ffi 12 lik3 �d7 13 1.d3 .S4 14 .xg4 �xg4 I S �bS Wd8 16 �gS+ �e7 1 7 1.xe7+ Wxe7 18 lIael+ �S, Alvarez Villar - GHrd, corr. 2000, and as 19 lOxc7 lIac8 20 � J.xe6 2 1 dxe6 g6 regains the pawn, Black is OK.

IO ••• Wd8!

Black prefers to keep his Iight­squared bishop for use on the king­side (for instance • . . . 1.h3. or .. . .i.g4.

can sometimes prove useful re­sources). However, as Black has less space to manoeuvre, swapping a pair of pieces also has to be considered:

1O . .• �d7 1 1 �xd7+ �d7 12 10:3 ( 12 c4?! �e7! 1 3 10:3, Simcoe-Campbell, corr. 1978, and now 13 . . . �ftJ! intending . . . Axc3, or . . . �eS, and then .. . lOftJ) 12 . . . lOftJ 13 l%e I ( 1 3 1Oe2 -f4-e6 can also be considered) 13 . . . l%e8? ( 1 3 . . . cS) 14 • d4 �S IS .xa7 lOxdS 16 "xb7 IOxc3 17 bxc3 g6 1 8 �e3 �g7 19 �d4 and White has a significant ad­vantage, Borisek-Picco, Nova Go­rica 2000.

1 1 .i.r4 Bronstein's move, defending

against the threatened . . . �h3 (as .i.g3 would follow), but the bishop can prove a target here for a knight hop to hS.

I I f3! is objectively stronger, l l . . .exf3 1 2 .xf3 �e7 ( l 2 . . . lOftJ 1 3 .i.d3 .g4 14 10:3 .xf3 I S l%xf3 .i.e7 16 .i.f4 and White has the more pleasant position, Harper­Tatlow, corr. 1999) 1 3 1Ot3 IOftJ 14 h3 ! ( 14 �d3 .g4 should transpose to the .. previous game) 14 . . . �d7 I S .i.d3 WItS 16 .xhS IOxhS 1 7 �e3 IOffi 18 :ael c6 19 l%e2 (maybe 19 g4 h6, J first, is more precise) 19 . . . l%e8 20 l%fel �fS 2 1 �gS l%xe2 22 1Oxe2 h6 23 �xfO+ gxftJ 24 1Of4, White has a nagging edge, Ruggeri Laderchi-Kosten, email 2001 .

1 1 ••• lOr6 . l l . . .hS?! is overly aggressive, 12 f3 �f5 13 10:3 exf3 14 .xf3 �xc2? (exceeding!>' risky) I S �gS+! IOftJ ( l S . . .• xgS?? allows mate in one) 16 :ae l ! c6 (and now 16 .. .• xgS? is met by 1 7 .xfO+! and 1 8 l%e8 mate) 1 7 �xfO+ .xftJ 1 8 .e2 .d4+ 19 �hl �g6 20 l%xfS+ �c7 (20 .. . l%xfS 2 1 .e7+) 2 1

Bronste;n 's variation 6 �e2 59

�xc6 (2 1 l%xa8 and other moves also won, if somewhat mundanely) 2 1 . . . bxc6 22 IObS+ cxbS 23 .xbS l%e8 24 l%e7+ 1 -0, as 24 .. . l%xe7 25 .c6 mate.

However, 1 1 . . .�e7!? 12 f3 exf3 I � .xf3 �ffi 14 10:3 1Oe7 is an in�ting attempt to improve the black piece coordination, 1 5 �d3 .g4 16 .e3 �S and Black is do­ing well, Elburg-Malmstrom corr. 1994 .

ll lDc3 12 c4?! is worse, 1 2 . . . �e7

( 1 2 . . . IOhS 13 �g3 IOxg3 14 hxg3 �e7 to ftJ, seems simple, and good) 1 3 10:3 �f5 (and again, 1 3 . . . lOhS g!ves the bishop a hard time) 14 .c2 with chances for both sides, Terlecki-Tscharuschin, corr. 1990.

12 .... 6 13 �.4 b5!? 13 . . . lOhS 14 .d2 J..e7 IS �e3

l%fS is a logical line, intending to play . . . �ftJ.

14 �b3 Svendsen-GHrd, corr. 1990,

when, once again, 14 ... lOh5 is worth consideration.

B 6 .... dB

This seems very retrograde, but is in fact one of Black's two strongest possibilities.

7 0-0

60 Bronstein 's variation 6 J.e2

Other b"ies: a) 7 dS Gaining space and isolat­

ing the c4-pawn, but this is probably not a necessary precaution as Black will not be playing . . . dS himself for a while. There are two main alternatives:

a l ) 7 ... 1Oa;

8 lOt3! (In the I S-minute game Jackson-Mephisto, London 1992, after 8 .d4 J.e7 9 lOt3 J.f5 [9 ... cS is temptin.8] 10 lOe3 J.g6 1 1 0-0 0-0 12 IOg4 �7 1 3 J.e3 lOts 14 b4 lOxg4 I S J.xg4 J.a; 16 .c4, the computer showed its greedy nature with 16 .. . bS? [ 16 . . . lOd7 is level] 1 7 IOxbS J.xa I 1 8 lba I a6 19 �d6 cxd6 20 bxcS when White had more than enough compensation for the exchange and went on to win. 8 lOe3 g6!? 9 h4 [9 c4 J.g7 10 lOt3 0-0 = Fiorito-Borrmann, COff. I 983/8S] 9 ... h6 10 lOt3 J.g7 1 1 J.c4 lObd7 12 hS lOes 1 3 hxg6, Littleboy-Borrmann, COff. 1983, 13 ... lOxc4 14 lOxc4 .t.f5 with good play. 8 0-0 is rarely played in this position, but is important for trans- _ positional purposes, 8 .. . .t.e7 9 lOt3 [9 lOe3 0-0 10 0 exO 1 1 .t.xO lObd7 and, by bringing the knight to cS, Black solves his main develop­ment problem, Elburg-Downey, COff. 1994/96] 9 . . . 0-0 10 0 exO 1 1 lbO [ 1 1 J.xO] 1 1 . . .1086 12 J.e3,

Wamock-Hayward, COff. 1986, when 12 ... c6 and . . . lOt7 will oblige White to exchange his dS-pawn) 8 . . . .t.e7 and White does dispose of a wide choice at this juncture:

a l l ) 9 J.e3 0-0 10 lOd2!? cS!? (immediately giving up the pawn in return for an active position, if 1O • • • J.f5 then 1 1 g4! .A.g6 12 gS IOfd7 produces a fairly typical posi­tion for this variation, 13 lOdxc4 lOts 14 lOxcs dxcS I S .d2 lOd7 16 (H)-() a6, Coqhill-Elburg, COff. training game 1999, 1 7 h4-hS clearly favours White) 1 1 lOdxc4 � 1 2 :bl (perhaps 12 J.c l is more acc:urate, as it does leave., White a solid pawn up, 1 2 ... lOxc4 13 lOxc4 1086 [ 1 3 ... .t.h4?! 14 0-0 J.f5 I S lOd2 lOd7 16 lOc4 .c7 17 .t.f4 forces the dark-squared bishop to return, Ruggeri Laderchi­Cramton, COff. 1999] 14 0-0 J.f5 IS IOg3 .t.g6 16 .t.hS and Black has no compensation, Ruggeri Laderchi­Elburs, COff. traininS.- game 1998) 1 2 ... lOxc4 13 tOxc4 .t.f5 ( I 3 . . .• aS+! 14 lOt3 .t.ffi I S J.d2 .t.f5 may be a better continuation) 14 lOd2 1086 I S a3 lOt7 16 b4! the sharpest possibility, 16 ... lOxdS 17 .t.c4 .t.e6 18 100 �3! (this is bet­ter than 1 8 ... lOf4 19 .t.xe6+ lOxe6 20 .dS :a; 2 1 .t.gS lbO 22 pO .t.xgS 23 .xe6+ Wh8 24 0-0 with

advantage) 19 .*.xe6+ �h8 20 .d3! (20 .cl -..,S [20 . . . dS !?] 2 1 109 I dS 22 :b3 d4 looked awkward to me, and certainly not the sort of position to play against lonny) 20 . . . lOxbl 2 1 0-0 lOxa3 22 .xa3 .xb4 23 .d3 bS 24 c3 .as 2S lOgS .*.xgS 26 .*.xgS � 27 .*.a2 c4 28 .g3 dS 29 :dl .e6 30 .*.e3 as 3 1 .*.bl b4 32 :d4 (the black queenside is be­ginning to look menacing, so White stakes everything on the attack) 32 .. . bxc3 33 :M h6 34 h3 J:ab8 3S .*.g6 :b6 36 .*.c2? (as my opponent pointed out to me after the game: 36 J:xh6+! gxh6 37 .*.d4+ :ffi 38 .c7 .g8 [38 . . . �g8? and mate in S: 39 Wh7+ �fS 40 .*.cS+ J:46 4 1 �8+ etc.] 39 .xb6 .xg6 40 .*.xffi+ �g8 41 .d8+ �f7 42 .e7+ �g8 43 .*.xc3 wins) 36 . . . :ffi 37 :g4 .d6 38 f4 :b7 39 .*.d4 :bf7 40 .xc3 �g8 4 1 .*.xffi J:xffi 42 .xaS � 43 .xb6 J:xb6 44 :g6 :b2 4S .*.fS c3 46 :c6 d4 47 .*.g6 � . 48 :e6 and White easily rounded up the pawns, Kosten-Hector, Torcy 1988.

a 12) 9 g4 is very dangerous, White tries to displace the e4-pawn's defender. 9 . . . h6 (Black should probably stop the advance of the g-pawn thus, and then develop his minor pieces, waiting to see where White will castle. 9 . . . 0-0!? 10 gS lOe8?!, this avoids losing the e­pawn but the knight is horribly placed, [ 1O . . . lOfd7!? is more natu­ral, but Black's compensation after 1 1 lOxe4 lOc:s 12 lOXcs dxcS, Rug-

, geri Laderchi-Villar, corr. 1999, 1 3 h4 is not that obvious] I I h4 .*.fS 12 lOe3 .d7 [ 1 2 . . . .*.g6?! 1 3 :gl .d7 14 Ag4 .d8 I S hS is horrible for Black, Tatlow-Niemand, corr. 1995] 13 lOxfS .xfS 14 .*.e3 cS IS .*.g4 .eS 16 .*.e6+ with a crushing posi­tion, Liardet-Nyffeler, corr. 1992) 10 00 (White picks up the e-pawn.

Bronstein 's variation 6 .*.e2 61

10 gS?! hxgS I I .*.xgS .*.fS 12 lOe3 .*.g6?! 1 3 h4?! [ 1 3 :gl is more awkward] 1 3 . . . lOg8?! [ 1 3 . . . lOfd7] 14 he7 .xe7 I S .d4 J:xh4 16 J:xh4 .xh4 17 .xg7 �7 18 .xh7 hh7 19 lObs with an edge, GHrd-Kozlov, corr. 1994197) 10 . . . e3 I I fxe3 0-0 12 0-0 IObd7 13 h3 lOes 14 llX:4, Tatlow-Svendsen, corr. 1996, when 14 .. . lOxc4 I� .*.xc4 lOcI7 would replace the eS-knight. The looseness of White's pawns provides some compensation for the e-pawn.

a 1 3) 9 .*.gS is not effective here, 9 . . . .*.fS (9 . ..loxdS!? 10 .xdS .bgS I I lOxe4 might be playable, al­though it does leave Black some­what behind in development) 10 lOe3 ( 10 .d2 lObd7 I I lOe3 .*.g6 12 h4 h6 13 .*.f4 a6, fairly equal, Deacon-Cavendish, corr. 1991 ) 1O . . . .*.g6 I I .d4 lObd7 12 0 exO 13 .i.xO, Evans-Heap, corr. 1994/9S, 1 3 . . . lOes with equality.

a (4) 9 OO!? heads directly for the e-pawn, 9 . . . .*.fS 10 g4 .*.g6 I I gS e3 ! 12 fxe3 IOfd7 13 h4 lOes with some positional compensation for the pawn, Davics-Crimp, Aus­tralia 1995.

a2) 7 ... .*.e7 8 0-0 .*.fS!? (8 . . . lOf6 transposes to 7 ... lOffi) 9 lOe3 (the correspondence game Strautins­Borrmann, 1990, came to a rapid conclusion following 9 .i.g4!? .*.X&ct 10 .xg4 IOffi! I I .xg7 :g8 12 Wh6 .d7 1 3 :el :g6 14 .d2 [ 14 � c6 I S dxc6 lOxc6 is very risky for White] 14 . . . J:xg2+ Ih-Ih, as I S �g2 .g4+ leads to perpetual check) 9 . . . .t.g6 10 .t.hS ! �ffi!? (Latvian Gambit players like using an open h-file! ) 1 1 hg6+ hxg6 12 lOc:3 ( 12 O?! exO 13 .xO lObd7 14 lOc:3 lOes I S .e2 .d7. Black is very well placed, with the eS-square for his knight, 16 .t.d2 0-0-0 17

62 Bronstein 's variation 6 �e2

."S c6! 18 .as �b8 19 Zlae I lIhS 20 lIf4 J:ldh8 21 h3? [2 1 10ft] 2 l . ..lIxh3 ! 22 gxh3 Wxh3 23 �t2 lObs 0-1 , Niemand-Krantz, corr. 1995) 12 .. .• d7 13 Wd4 cS 14 Wc4 lIh4 IS f4 ( I S g3?! Wlt3! 16 gxh4 IOg4 17 lOxg4 _xg4+ forces a per­�tual check, but I S O a6 16 lOxe4 l&e4 17 fxe4 �fti 18 Wd3 appears strong) I S . . . lOa6 16 lOxe4 0-0-0 17 g3?! ( 1 7 lOgS) l7 . . . lIhS 18 lOgS ll)g4 when White had problems turning his extra pawn to account, Kozlov-Krantz, corr. 1993).

b) 7 1Oc3 dS! (the logicafl'iposte now that White cannot avail himself of the flanking blow c4. 7 . . . lOfti? is much worse: 8 �gS �fS 9 �hS+! g6 10 lOe3! gxhS [or 10 . . . �e6 I I ll)xe4 with a good pawn more] I I IOxfS �e7? 12 lOxe7 Wxe7 13 ltldS 1-0, Rittenhouse-Elburg, corr. 1990) 8 lOes IOfti and:

b I ) 9 .i.gS �e7 I 0 �hS+!? (White invests a tempo in order to weaken Black's kingside dark squares, to 0-0 0-0 I I 0 exO 12 �xO c6 is far from frightening for Black, 13 .d3 [ 1 3 iOC2 IObd7 14 c4?! Elburg-Keller, corr. 1994. Black can now gain the upper hand by 14 .. . lOxeS! I S dxeS IOg4] 13 . . . b6?! allowing a tactic, [ 1 3 . . . lObd7 is OK] 14 IOxc6!? IOxc6

I S �xfti lIxfti 16 �xdS+ �h8 17 lIxfti �xfti 18 �xc6 .xd4+, re­gaining one pawn, with some draw­ing chances because of the bishop pair, Shortland-Crimp, corr. 1996) 10 . . . g6 1 1 �e2 �fS 12 0-0 lObd7 13 0 exO 14 .A.xo lOxeS?! (initiat­ing a sharp tactical sequence, much in keeping with Hector's 'no-holds­barred' style, but 14 .. . c6 was some­what simpler) I S dxeS lOe4 16 �xe7 .xe7 17 .xdS lOxc3 18 .xb7 .cS+ 19 �h I lId8 20 bxc3 0-0 2 1 c4 �e6 22 �dS lIxfl + 23 lIxfl �xdS 24 cxdS Wc4! and Black had just enough initiative to make a draw, Ivanovi�-Hector, Metz 1988.

b2) 9 0-0 �e7 (Black could probably try 9 ... c6, or 9 . . . -*.fS to avoid the problems associated with �hS+ in this line) to �hS+! g6 I I �e2 c6 12 �h6 1Obd7 ( l 2 .. . �fS 13 o IObd7? [of course, 13 . . . exO was better] 14 fxe4 dxe4?? [and 14 .. . �xe4 I S IOxe4 dxe4 was also superior to the move played, al­though 16 �c4 or 16 l&d7 gives the advantage] I S lIxfS 1-0. as I S . . . gxfS 16 ,j,hS+ leads to matC), Rittenhouse-Hayward, corr. 1990) 1 3 0 exO 14 IOxo IOg4 I S Wd2 IOxh6 16 .xh6 �m 17 _f4 Wfti 18 .d2 �e7. Black is reasonably placed, Downey-Heap, corr. 1994/96.

7 ••• lOr6 Averkin's suggestion 7 . . . �e7

might be more accurate as 8 �gS is thereby avoided, e.g. 8 0 (8 1Oc3 dS 9 lOes 10ft) transposes to 7 1Oc3) 8 . . . exO 9 �xO liSfti transposing to 7 . . . lOfti.

Otherwise, 7 . . . dS?! is best met by 8 lOe3, c4, 1Oc3, etc. rather than 8 lOes 10ft) 9 �gS �e7 10 c4 c6 I I 1Oc3 �e6 12 Wb3 Wb6 13 .xb6 axb6, Liepins-Gubats, Latvia 1963,

although even here White has a small edge.

8 13 8 .i.gS (the most common, but

avoided by 7 .. . J..e7) 8 . . . J..e7 9 �3 (9 0 exO 10 J..xO (H) I I liX:3-Keres, transposes to a later varia­tion) 9 . . . 0-0 (9 .. . c6 10 c4 is a more common move order, but can Black forego . . . c6?) 10 c4 (White is target­ing the e4-pawn, I 0 � dS I I c3 c6 12 Wb3 Wh8 [ I 2 . . . aS !?] 1 3 0 exO 14 lLlxO Correa-Kozlov, corr. 1976179, when Black could have played 14 . . . lLlbd7 I S :ae l IDc4 16 J..xe7 .xe7 1 7 J..d3 lLldffi) 1O . . . c6 (far from forced) 1 1 liX:3 lObd7 12 .c2 lIe8 13 :ad I ( 1 3 O?!, Aalzum-Reurich, corr. 1982, allows the surprising tactic 1 3 . . . lLldS!? 14 J..xe7 lLlxe3 with near equality) 1 3 . . . dS? (the only chance is 13 . . .• aS when 14 .i.xffi lLlxffi I S lLlxe4 lLlxe4 1 6 .xe4 .xa2 appears fine for Black) 14 cxdS cxdS I S lLlbS, winning material, Zier­Tiemann, COrr' 1 979.

8 dS transposes to 7 dS. 8 .•• ex13 9 J..dJ J..e7 10 llk3 10 .i.hS+!? is a typical idea,

weakening the black kingside dark sguares. 1O . . . g6 ( 1O . . . lOxhS!? I I WxhS+ g6 12 .dS IIfS might be playable) I I J..gS !? (the most fun, although 1 1 J..O 0-0 1 2 J..h6 1If7 1 3 lLlc3 c6 14 dS cS I S J..f4 favours White, Harabor-Downey, corr. 1995) l l .. .lLlxhS (critical, 1 1 . . .(H) 12 J..O c6 1 3 lLlc3 lDa6 14 _d2 with a typical white space advan­tage, Strautins-Kozlov, corr. 1990 and l l ...gxhS!? 12 .i.xf6 .i.xffi 13 .xhS+ We7 14 lLlc3 certainly looks like it could be winning for White!) 12 .i.xe7 ( 1 2 lIe l !? &6 1 3 lDas lLles 14 J..xe7 .xe7 I S dxeS dxeS is also unclear) 12 . . .• xe7 13 lIe l .i.e6 14 dS (H) I S dxe6 lLlc6 16

Bronstein 's variation 6 .i.e2 63

lLlc3 lLlf4 and the e6-pawn may be just a weakness, Strautins-Krantz, corr. 199O.

10 ... 0-0 1 1 .i.g5

This position was considered by Keres to be advantageous for White.

1 I ... lLlbd7 1 1 . . . c6 is more common, but after

1 2 dS White is better, e.g. 12 .. .'�h8 ( l 2 . . . cS leads to an unfavourable Benoni-type position) 13 _e2? (missing Black's threat, 13 .i.f4 maintains a plus, and 13 _d2 like­wise, Harper-Van de Velden, corr. 1999) 1 3 . . . lLlxdS! 14 .i.xdS .i.xgS I S J:lxfS+ _xfS 16 IIfl .i.ffi and Black has a solid extra pawn, Herg-Loon, corr. 1982. l l . . .h6!? 12 .t.f4 dS is possible.

12 lilds 1 2 _e2 is stronger, 12 . . . 1If7 13

J:lad I _fS 14 lLlbS .i.d8 and Black is under pressure, Berg-Den Hertog, corr. 1982.

12 ••• lLlsdS 13 .t.sdS+ Wb8 14 .i.u7 _u7 15 IIsf1+ _sIB 16 .13 _s13 17 .t.s13 lOf6 18 lie I Wg8 19 lLle3

19 lIe7 Wf8 20 J:lxc7 �8 traps the rook, although White will obtain sufficient compensation.

19 •••• 20 lilds .i.d7 21 cl c6 1l 1llsf6 pf6 ll Wfl as 24 Wg3 bS 25 a3 b4! 26 axb4 axb4 27 (sb4 IIbS

64 Bronstein 's variation 6 �e2

. . . which is equal, although Black managed to grind out the win on move SS in the game Fransson­Hector, Gausdal 1990.

C 6 ••• d5?!

This move is a mistake, providing White with a target.

7 � This puts immediate pressure on

dS and is much stronger than 7 lbeS?! , which, although appearing to be more aggressive, achieves lit­tle: 7 . . . �d6 S f4 �+!? (S . . . lbe7 is simpler) 9 g3 �3 10 �fl .e6 I I h4?! lbe7 Black is at least equal, Comer-Johnsrud, corr. 1999.

7 ... fOe7 None of Black's alternative

moves are satisfactory, but I prefer this to the attempt to hold on to the centre by:

a) 7 . . . c6 S c4 (S (H) does not have quite the same ·bite', although after S . . . �d6 9 0 � 10 g3 .e7 1 1 c4!? [ 1 1 fxe4] I I . . .lOft; 12 lOc3 �e6?! 1 3 cxdS cxdS 14 fxe4 [ 14 "'3 wins a pawn] 14 . . . dxe4 I S Illi3 White still had the better play in Gaard-Strelis, corr. 1992) S . . . .i.b4+ (Black hereby gains time, but he will be obliged to concede his dark­squared bishop, and this brings its own problems. Anyway, S . . . �e6?! 9 lOc3 .f7 fails to defend the centre, 10 cxdS cxdS 1 I �g4!, the d­pawn's defender is exchanged, 1 1 . ..�xg4 12 .xg4 lOc6 1 3 lOexdS :dS 14 .xe4+. Black has zero compensation, Ruggeri Laderchi­Scherman, corr. 1999) 9 lOc3 lbe7 10 (H) (H) (Black might just as well get his king into safety, the alterna­tives are no better: 1O . . .• d6? 1 1 -"3 �xc3 12 bxc3 dxc4 13 �xc4 with advantage Strautins-Budovskis, corr. 1970, is even worse, 10 . . . lOf5? loses on the spot, I 1 lOcxdS! cxdS 1 2 lOxdS forking queen and bishop, Strautins-Krustkains, corr. 1990, and 12 . . .• d6? is no defence, 13 �f4, 1O . . . �xc3 1 1 bxc3 �e6?! [ I I . . .(H) 12 "'3 transposes into the next note] 12 0 .g6 13 fxe4, Strautins-Grava, corr. 1970/1 , is al­ready too dismal for words, if now 13 . . .• xe4? then 14 �O � IS g3 -.u 16 �g2 wins) 1 1 "'3! ( 1 1 cxdS �xc3 12 bxc3 cxdS [ 1 2 . . . lOxdS 1 3 "'3 .f7, Strautins­Nyman, corr. 1990, seems worse af­ter 14 0] 13 � is less precise, as Black now has the c6 square avail­able, 1 3 . . . lObc6 14 0 with an edge, Schuttrich-Mathe, corr. 1973) 1 1 . . .�xc3 12 bxc3 .f7?! 13 .h3 :dS 14 �xe7 .xe7 I S cxdS with a big advantage, Silavs-Vildavs, Latvia 1963, is fairly typical of the problems facing Black here: behind

in development, he is incapable of countering White's pressure.

b) 7 . . . -*.e6 S c4 -*.b4+ 9 lOc3 �7 10 (H) -*.xc3 I I bxc3 1Obc6 12 f3 and Black is already losing, Curdo­Hayward, New Hampshire 1976.

c) 7 . . .• dS!? is a better idea, S 0-0 1Of6 9 c4 dxc4 (the best, for, as usual, holding on to the centre just wastes time, and gives White a ready-made target) 10 .t.xc4 lOc6 I t .t.bS a6!? ( 1 1 . . .-*.d7 12 lOc3 -*.d6 is playable) 12 -*.xc6+ bxc6 13 lOc3 with a plus, Elburg-Budovskis, COrT. 1994/9S.

8 c4 This is strong, but S (H) is also

reasonable; S .. .lt�bc6, counterattack­ing d4, 9 c3 -*.d7!? 10 1Wb3?! ( 10 f3 is more testing, as the black queen is exposed on the f-tile, e.g. 10 . . . exf3 I I .i.xf3 .d6 12 -*.xdS! IOxdS 13 �+ g6 14 .xdS, winning a good pawn) IO . . . Q-(H)!? I I IOxdS IOxdS 12 .xdS -*.15 13 1Wb3 -*.d6 14 -*.e3 �4 I S g3 �3 16 llXI2 hS with some counterplay, Nilsson­Svendsen, COrT. 1991 .

8 ••• dxc4 Rather than defend the d-pawn,

which as we have seen in a previous note is too passive. Black can also by the counterattack S . . . 1Obe6 9 lOc3 .xd4 (9 . . . dxc4!? 10 dS IOd4 is worth a try), now there is much to be said for the simple 10 cxdS (bet­ter than the 10 .i.d2?! of From­Harju, COrT. 1975n7, when 10 .. . dxc4 I t lObs .dS is unclear) 1O . . .• xdl + I I .t.xd l IOb4 1 2 (H), winning the e4-pawn, 1 2 . . . -*.15 13 -*..4+ -*.d7 14 .t.xd7+ Wxd7 MalmstrOm-Krustkains, COrT., I S :cil etc.

9 lOc3 -*.f5 10 .t.xc4 10 g4?! -*.g6 I I h4 is too loosen­

ing, 1 1 . . . h6 1 2 -*.xc4 1Obc6 13 lObS

Bronstein 's variation 6 -*.e2 65

Q-(H), winning quickly, Niermann­Elburg, COrT. 1999.

10 ••• lObe6 I I dS! The alternative I t 1Oxf5 IOxI5!?

12 dS Q-(H) is �ite double-edged. 1 1 ••• 1Oe5 II •• 4+ -*.d7 12 . . . WdS!? 13 1I'b3 1I'b6 14 0-0 14 1Oxe4. I

14 ••• 1Oc8 IS IOxe4 IOllc4 US IOllC4 .xbl 17 axbl -*.b4 18 .t.dl .t.xdl 19 10cxdl 0-0

. . . when White's tatty pawn struc­ture allowed Black certain drawing possibilities, Rittenhouse-Downey, cOrT. I992.

D 6 •••• f7?!

In an ideal world, this is where Black should like to put his queen.

66 Bronstein 's variation 6 lLe2

Unfortunately, White appears to be able to build up a large advantage almost by force.

7 liX3! As this move is strong, there is no

need to concern ourselves too much with the alternatives, most of which allow Black a relatively easy game. In brief:

a) 7 0-0 also promises some ad­vantage, 7 . . . 1Oft» 8 lLgS (8 lLlc3 c6 [8 ... .i.e7 9 AgS transposes] 9 lLf4 .e6 and now, instead of 10 lOe3 dS I I lLeS?! lObd7 = Trutmann-Purins, corr. 197 1 172, 10 dS! was good) 8 . . . .i.e7 9 lLlc3 lLf5?! (9 . . . lObd7 is considered in the main line) 10 lOe3 .i.g6 1 1 lLc4 .18 12 0 exO 13 .xO c6 14 :ae l with a consider­able advantage, the black king is stuck in the middle, Van Willigen­Keijzer, corr. 1982.

b) 7 dS lOffi 8 lLlc3 .i.e7 9 lObS lLla6 10 lOe3 0-0 I 1 0-0 lLd7 =

Strelis-Maly, corr. 1970. c) 7 lOe3 lOft» 8 0-0 lLlc6!? 9 0

exO 10 lLxo lLd7, unclear, Simila­Zalit� Sweden 1978.

d) 7 c3?! lLe6!? 8 b3 lLld7 9 0-0 lOgf6 10 1Obd2 0-0-0 I I lOe3 dS 12 a4 hS 1 3 0 • g6 14 f4 lOg4 with the makings of a strong attack, Karlsson-Zalitis, Sweden 1978.

7 ••• lOf6 Black must defend the e-pawn,

and 7 . . . dS? loses to 8 lOes We6? 9 .i.g4, whilst 7 . . . lLf5?! 8 lOe3 is most awkward, for instance 8 . . . lLg6 9 0-0 (9 Ac4 .d7 10 lLdS wins a pawn) 9 . . . lLld7? (9 .. . lOffi 10 0 with advantage.) 10 .i.c4 .e7 1 1 lOeds .d8 12 lObS winning, Downey­Logunov, corr. 1994/9S, as the de­fence 1 2 . . . lIc8 meets 13 lOf4 lLf5 14 .i.e6 which is awful for Black.

8 .i.g5 lObd7 There is little choice, 8 . . . J.f5?! 9

lOe3 J.g6 10 lLc4 .d7 1 1 lLxffi

gxffi 12 lLdS wins a pawn, 1 2 . . . c6 1 3 lLxc4 dS 14 lLxg6+ bxg6 IS .0 .17 16 0-0-0, Schultz Pedersen -Kozlov, corr. 1976 and Black has no real compensation.

9 1Ob5 9 0-0 is also dangerous, 9 . . . lLe7

10 lObS (forcing the black king to move, although 10 0 also looks good) I 0 .. .'�d8 1 1 0 exO 1 2 J:lx0

� h6 13 lLh4 a6 14 lLlc3 gS I S lLg3 .g7 16 a4, Downey-SCnechaud, corr. 1994I9S, and now 16 . . . lLlhs 1 7 lLf2 lOf4 is playable.

9 ••• �d8 10 .d1 lLe7 1 1 0-0-0 b6!?

Not a good move to play against any member of the Littlewood fam­ily: it allows White to wrap up the game in typical, incisive fashion. To be fair to Black, his position was already quite unpleasant. I I . . .a6 12 lLlc3 lIe8 keeps things intact for a while, though.

11 lLle5!?

1l ••• dxe5 13 dxe5 lOa8? 13 . . .. xa2 14 lLlc3 .al + I S lObi

.a4 is not so clear. 14 lLg4 lLxg5 15 .xg5+ .e7 16

.f4 b5 17 lLxd7 lLxd7 18 .xe4 lIbS 19 .c6 �e8 10 lOsc7+ �dS 11 e6 lOf6 11 esd7

1-0, P .Littlewood-Kindermann, London 1978.

E 6 ••• bS?

Black plays a game of bluff, since if White doesn't take this pawn for fear of opening the h-fiIe, Black will be able to develop his queen on g6 after all, the hS square now being additionally defended by the rook. That said, if White does take the pawn (and he should) the open h­file does sometimes come in useful.

7 .i.sbS+ If instead 7 liX3 then Black

should reply 7 . . .• g6 (rather than 7 . . . .i.f5?! 8 0 exO 9 .i.xo c6 10 0-0 dS? [ 10 . . .• g6 was more cir­cumspect, although still not very �ood) I I .i.xdS! cxdS 12 lOxdS .c6 13 1I'e2+ .*.e6 14 1I'xe6+ 1i'xe6 IS liX7+, winning, From-Kemer, corr.) when the game Destrebecq­Jurgenson, corr. 1992, went on 8 lOe3 IClf6 9 0-0 lOc6 10 0 exO I I .i.xf3 .i.d7 12 lOcds lOxdS 1 3 .*.xdS 0-0-0 14 .i.f1 .e4 which looks OK for Black.

7 ••• g6 8 .i.g4 Possibly 8 .i.e2 is more accurate. 8 ••• .i.fS 9 llX3 c6 10 .i.r4

Bronstein 's variation 6 .i.e2 67

10 dS iCld7 I I .i.e2 liXs 12 .i.e3 .e7 13 .d4 is also very good for White, Van der Hauw-Diepstraten, corr. 197S.

10 •• ':b4!? The h-file! 1 1 .i.sfS :d4?! 11 .*.se4?! Both 12 .i.c8! and 12 lOxe4 look

better, and with great advantage. Il ... dS 13 .i.sg6+? 13 .i.xdS! . 13 •••• sg6 14 �? .sgl � K1oss-Diemer, corr. 19S9.

F Apart from the five moves mentioned, Black might also be able to essay 6 ••• �7!?

7 dS The idea is to answer 7 0-0 with

7 . . . 1Clf5, hittin� d4, and 7 liX3 with 7 . . . dS as 8 &3 c6 is fine since White can no longer play c4 .

7 •• ifS 8 llX3 lOa6 9 .i.e3 cS?? Losing a piece, 9 . . .• f1 is consid­

erably better. 10 g4! .i.g6? 1 1 gS 1 -0, Crimp-Clarke, corr. 1998, the

queen is lost.

5 Leonhardt's variation 4 lLlc4

This variation was recommended by the Genoan master Leonhardt ( 1 877- 1934) but was also mentioned by Cozio in 1760. The difference between this and the previous chap­ters is that White will have the pos­si�i1ity ofattacking Black's e4 pawn With d3 at some point. Instead of playing strategically, White aims for rapid piece play . . This is now White's most popular

hne. 4 ••• fxe4 Neither 4 . . . bS? S /.De3 fxe4 6 liX:3

with advantage, nor 4 . . .• e6?! S /.De3 f4 (S . . . fxe4 6 .i.c4 .g6 7 d3 exd3 8 .i.xd3 leaves White a useful lead in development, Van de Pol-De Boer, Dutch Junior ch 1990) 6 lOds .xe4+ 7 .e2 with advantage, Pupols-Zemitis, corr. 1970, repre­sent improvements in any way.

5 /l)cJ

The immediate S d3 is also played sometimes, and is not · quite as in­nocuous as it appears. S ... exd3 (In the game Foltys-Petrov, Podebrady 1936, Black, probably the strongest Latvian chessplayer before the ad­vent of Tal and numbering Alekhine amongst his many victims. played S . . . dS!? 6 �3 [but 6 eJts+ .f7 7 .xf7+ *xf7 8 · �3 Halstead­Smith, Lansing ch 1985 gives White a smallllus] 6 . . . .i.e6 7 dxe4 dxe4 8 �3 b4 9 .t.d2 �7 10 �e4 .i.xd2+ 1 1 �d2 0-0 12 IOe4 .xb2 =) 6 £xd3 dS 7 1I'hS+ (after 7 /.De3 Black should be OK as com­pared to the main line he has not wasted a move with his queen for instance 7 . . . £e6! 8 0-0 �' [or 8 . . . .i.d6] 9 c4 dxc4 10 £xc4 £d6 and . . . 0-0-0) 7 . . .• f7 8 .xf7+ *xf7 9 /.DeS+ ci>ffi 10 100 .i.d6 1 1 j,gS+ ci>f7 12 liX:3 c6 1 3 0-0 IOffi 14 life I £g4 and Black is fine, Hunon­Destrebecq, Brest 1978.

At this juncture there is a wide range of options but only the fol­lowing three come anywhere near being satisfactory.

A 5 •••• 16 B 5 •••• e6 C S •••• .,!

Nevertheless, other moves are played:

a) S . • . <:6?! this is surprisingly popular, 6 lOxc4 (this time Black is well placed to meet 6 d3?! dS 7 /l)e3 .i.b4 8 d4? [8 .i.d2] 8 . . . /l)e7 9 a3?! 0-0 10 .d2 .i.d6 and White is in trouble, Dagendhart-Krull, corr. 197 1 , but 6 .g4!? is interesting, 6 . . . �h6 7 .xc4+ .i.e7 8 .eS, Castelltort-Julbe, Barcelona 1996, although Black obtains some play for the pawn, 8 . . . dS 9 .xf6 .i.xf6 I 0 �6+ Wd7 1 1 lOxc8 IIx(8) 6 . . .• e6 7 �S+ (7 .e2!? appears tempting, but 7 . . . dS! 8 �6+ Wd8 9 �gS .xe2+ 10 .i.xe2 .i.xd6 1 1 �f7+ We7 12 lOxh8 and can the knight escape? 12 . . . �f6 [ 1 2 . . . .i.e6! is more a<:curate] 13 g4! this pawn will save the knight, 1 3 . . . lOa6 14 .i.xa6 bxa6 15 gS �g4 16 g6 .i.f5 17 �f7 .i.xh2 18 f3 .i.g3+ 19 Wfl hg6 20 fxg4 Wxf7 2 1 d3 and White is better, Malmstrom­Goedhart, corr. 1995) 7 . . . g6 (7 . . . Wd8? 8 .eS dS 9 �gS! , Black is in a bad way, Matwienko-80mer, corr. 1974) 8 .eS (8 .e2!? is still possible, as the g6-move has its ad­vantages, and its disadvantages: 8 . . . dS 9 �6+ [now 9 �6+?! is worse, 9 . . . Wd8 10 �gS .xe2+ 1 1 .i.xe2 .i.xd6 12 �f7+ We7 1 3 lOxh8 .i.eS and the knight is captured] 9 . . . Wd8 10 lOxb7+ Wc7 1 1 .xe6 .i.xe6 12 � .i.c8 1 3 d4!, Crimp­Svendsen, corr. 1997. White is win-

Leonhard, 's variation 4 liJc4 69

ning as 1 3 . . . .i.xb7 14 .i.f4+ Wb6 15 lOxb7 Wxb7? 16 .i.eS wins the rook) 8 . . .• xeS 9 �xeS dS (9 . . . d6 1O �! Wd7 [IO . . . dS 1 1 �6+ Wd7 1 2 �f7 dx<:4 13 .i.x<:4] 1 1 .i.e2 Wc7 1 2 d4 and Black still cannot play . . . dS, Mi&ala-Melchor, corr. 1998) 10 �gS lOh6 1 1 d4 .i.f5 12 .i.d3 .i.xd3 13 �xd3 and one can't really say that Black has much in the way of compensation for his pawn, Whiteside-Silverman, corr. 1980.

b) S . . • /l)e7? clogs up the kingside, , 6 /l)e3!? (this is better than 6 d3?! ,

Bendig-Leisebein, corr. 1987, when Black can play the obvious 6 .. . cxd3 7 hd3 dS with a fine position. However, 6 �S+ g6, Frommel­Leisebein, corr. 1987, 7 lOxc4 .e6 8 .eS leads to'a favourable ending for White) 6 . . .• eS . 7 d4 exd3 8 .i.xd3 dS 9 0-0 .i.e6 (9.,.c6 is 00 better: 10 :Cl .c7 [ 10 . . .• d6 1 1 /l)exdS! cxdS 1 2 �bS .c6 1 3 .i.f4 is also crushing] 1 1 /l)exdS! cxd5 12 �xdS with a winning advantage Vitomskis-Krustkalns, Latvia 1973) 10 :C l ! ( 10 f4 .d6 1 1 f5 .i.f7 12 �g4 is less convincing, Pachman­Florian, Prague 1943) 1O . . .• d6, Miiller-Grott, corr. 1986, and others, and now 1 1 .i.f5 wins immediately, i.e. 1 1 . . .d4 ( 1 1 . . .<:6 1 2 � .eS 13 �!) 12 iObs � 13 lOxd4.

c) S . . . d6?! 6 d3!? (this is reason­able, as is 6 /l)e3 .g6 7 d3 exd3 8 .i.xd3 .f7 9 0-0 [9 .i.<:4 .i.e6, the point of playing an early . . . d6, 10 .i.xe6 .xe6 1 1 0-0 �6 12 �S 0-0-0 13 :Cl .f7 14 b4 �f6 and White has a space advantage, Kozlov-Grobe, corr. 1991] 9 . . . lO:6 10 f4!? �h6 1 1 .0 .i.e7 1 2 �S with a plus, Reitstein-Grivainis, corr. 1957, but the evident 6 �S+! g6 7 .as wins the e-pawn for noth­ing) 6 . . . exd3 7 .i.xd3 .f7 (perbaps

70 Leonhardt 's variation 4 liJc4

7 . . . .i.c6!? is playable, e.g. 8 1fhS+ .17, Mozota-Parzefall, corr. 1996, Black is not too badly placed) 8 0-0 ltk6 9 Ilel+ Wd8 (9 . . . .i.c6 10 .g4 lDd8 I I .i.f5 with advantage, Morgado-Korsmaa, corr. 1975) to .i.gS+?! .i.e7 I I .i.xe7+ �gxe7 12 lD04 dS?!, Gaard-Kozlov, corr. 1992, when 13 1De3 favours White:

d) S . . . ltla6?! 6 d3 (more accurate than 6 �xc4 .c6 7 1fhS+ 86 8 .eS .xeS 9 �eS �b4 10 Wdl .i.g7 with a very slight compensation for the pawn, Destrebec:q-Kozlov, corr. 198 1 ) 6 . . . .i.b4 7 .i.d2 .e6!? (but 7 . . . exd3 8 .i.xd3 1De7 9 0-0 leaves White well ahead in development) 8 1De3 (8 dxc4!? �f6 9 .e2 wins a pawn) 8 . . . �f6 9 dxc4 �xe4 (9 . . . .i.xc:3?! 10 Axc3 �xc4? I I .i.xg7 .17, Dravnieks-Svendsen, corr. 1992, and now 1 2 .d4 wins on the spot, Black only has one check!) to .i.c:4 .i.xc3 I I .i.xe6 .i.xd2+ 12 Wfl dxc6 13 1fhs+ WiS 14 .0+ with a winning advantage, Budovskis-Muller, corr. 1979/8 1 .

e) S . . . ltk6?! 6 lDds (6 .&.4 is also strong, 6 . . . dS !? 7 1fhS+ g6 ' .xdS .i.c6 9 Wbs 0-0-0 10 ltlas with a quick white win on the cards, Malmstrom-Triantafillopoulos, corr. 1997) 6 . . .• d8 7 .g4 �f6 8 �xf6+ .xf6 9 .xc4+ .i.e7 10 1De3 and Black's compensation for the pawn is very slight, O'Bee-Howes, corr. 1995.

f) S . . .• f5?! 6 1De3 .eS 7 d4 exd3 8 .i.xd3 �f6? (8 . . . .i.b4 9 .i.d2, Morgado-Atars, corr. 1977, and now 9 . . . �f6 to 0-0 0-0 leaves Black worse, but not terminally so) 9 0-0 ltk6 to Ile I and Black is al­ready lost, 10 . . . Wd8 I I It)eds .d4 1 2 .i.e3 1fh4 1 3 g3 .h3 14 �f4 1fh6 I S �+, winning the black queen, Krustkains-Lonsdale, corr. 1986.

A 5 •••• C6

6 d3 Thematic. If Black captures on

d3, then .i.xd3 will gain time on the black queen.

a) Leonhardt's original intention was to continue with 6 0 but, as BetinA pointed out, 6 . . . .i.e7! is then good: 7 fxc4? .i.h4+ 8 g3 .i.xg3+ 9 bxg3 .xg3+ 10 We2 dS 1 1 lOC3 d4 + .

b) White can also play 6 lDdS!? and this is quite dangerous, 6 . . . ltla6 (according to Bucker, 6 . . . Wd8 is a better bet, 7 d3 [7 d4 d6 8 �f4 .17] 7 . . . exd3 8 .xd3 .xd3 9 .i.xd3 ltk6 10 .i.f4 d6 I I 0-0-0 �f6 and White has a development advantage, but the black position is solid) 7 d4! (7 ltke3 .i.d6 8 .i.e2 �f6 = Atars­Tomson, corr. I 9741S) 7 . . . exd3? (7 . . . c6?! 8 lDes .c6 9 �f4 .f6 to .i.xa6 bxa6 I 1 0 with a dear ad­vantage; a better chance is offered by either 7 . . . d6 8 ltke3 c:6 9 .i.xa6 cxdS 10 .i.bS+, the doubled black d-pawn is a weakness, Pickles­Slaven, corr. 1996, or by 7 . . .• fT 8 ltke3 c6 9 .i.xa6 cxdS 10 .i.bS with similar problems) 8 .i.xd3 ! .xg2?! 9 1fhS+ (9 .e2+! also wins by force, 9 . . . .i.e7 10 .i.c4 1fh3 I I .i.gS .c6 12 �f4 .17 13 lDes .is 14 �! dxe6 IS Wbs+ c:6 16 .i.xc6+ Wd8 1 7 O-()-()+ Wc7 18 1ld7+!

.i.xd1 19 .xb1+ Wd6 20 llX4+ WcS 2 1 �S+ Wd4 22 .i.e3 mate, Biicker) 9 . . . Wd8 10 .i.gS+ .i.e1 1 1 li:lxe1 /l)xe1 ( l 1 . . ..xhl+ 12 Wd2 .xa I 1 3 li:lxg8 mate) 1 2 0-0-0 h6 13 .i.xe1+ Wxe1 14 /l)es and the black king is not long for this world, Fantoni-Richard, email 1999.

c) 6 /l)e3?! allows Black a simple route to equality, 6 . . . /l)fti 1 d3 .i.b4 8 .i.d2 Spiel-Antusch, Bayern 1988, and now 8 ... .i.xc3 9 .i.xc3 0-0.

6 •.• .i.b4 Black will lose his queen if he

tries 6 . . . exd3?! 1 .i.xd3 .xgl?? 8 �S+ Wd8 9 .i.gS+ /l)fti 10 .i.e4, and 6 . . . /l)fti?! 1 dxe4! (1 /l)xe4 /l)xe4 8 dxe4 .xe4+ 9 /l)e3 .eS!? 10 .i.c4 bS!? Biicker) 1 . . . /l)xe4?! (Black's compensation for the pawn is insufficient after the superior 1 . . . .i.b4 8 f3 .i.xc3+ 9 bxc3 d6, Rozzoni-Zaniratti, corr. 2000) 8 .e2!? (8 .i.d3 ! .i.b4 9 0-0 .i.xc3 10 bxc3 is very effective and without risk [for White, that is! ). After 10 . . . dS 1 1 /l)e3 .n 1 2 /l)xdS .xdS 1 3 .i.xe4 etc.) 8 . . . .i.b4 9 f3 dS 10 /l)d2 .i.xc3 1 1 bxc3 0-0 12 fxe4 White has some problems develop­ing his bits, but it is difficult to be­lieve that Black has enough compensation for a piece.

7 l.d2! White can transpose into B with 1

dxe4?! .xe4+ 8 lh3 .i.xc3+ 9 bxc3 /l)e1 10 .i.d3 (or 10 .i.c4 as per NCO, but this offers White nothing).

7 •• ..i.xc3 Universally played, as the alterna­

tives are worse: 1 . . . exd3?! 8 .i.xd3! .xgl 9 1Whs+ (White can also con­tinue 9 .e2+ WfS 10 .i.e4 1I'h3 1 1 /l)ds .i.xd2+ 12 .xd2 with advan­tage, Wei-GGAT, corr. 1993) 9 ... Wd8 (9 ... WfS) 10 0-0-0 (not 10 .i.e4?! /l)fti 1 1 Wh4 .g4 nor 10

Leonhardt 's variation 4 lix4 7/

.i.gS+ /l)fti) 10 . . . /l)f6 1 1 Wh4 .g4 1 2 .xg4 /l)xg4 13 :bgl (or 13 .i.gS+ .i.e1 14 .i.xe1+ Wxe1 IS :bgl d6 16 f3 /l)fti 11 IIxg1+ and Black is in difficulty, Strautins­Jac:kson, corr. 1991 ) 1 3 . . . /l)xt2 14 :del (14 .i.gS+ .i.e1 IS .i.xe1+ Wxe1 16 1lde l+ Wd8 11 /l)es offers good chances, too) 14 . . . lIe8? I S .i.gS+ .i.e1 16 :Xe1 li:lxd3+ 1 -0, Elburg-Michalek, corr. 1988. .-

1 . ..It)fti?! loses a pawn: 8 li:lxe4 .i.xd2+ (8 . . . /l)xe4 9 .i.xb4) 9 /l)exd2, Schuh-Hohnemann, corr. 1990. ....

8 .i.xc3 /l)r6 The game Smyslov-Kamyshov,

Moscow 1944, continued 8 .. . dS?! 9 /l)eS (9 /l)e3 /l)e1 [9 . . . .i.e6?! 10 dxe4 dxe4 1 1 .d4 is strong] 10 dxe4 dxe4 1 1 .d4 is a good alter­native) 9 . . .• £5 10 dxe4 .xe4+ 1 1 .i.e2 lDf6 12 0-0 c6?! ( 1 2 . . .• £5 is better, but 13 .i.d3 .f4 14 g3 ! 1Wh6 I S .i.b4 keeps the king in the centre, with a clear plus) 1 3 .i.hS+ Wf8 14 lie I 1Wh4 IS .i.g6!? /l)a6?! ( I S . . . /l)bd1) 16 .e2 .i.h3?! 1 1 /l)f3 1-0.

9 .i.xf6! 9 dxe4?! is much less convincing,

9 . . . li:lxe4 10 .i.d3 0-0 1 1 0-0 li:lxc3 1 2 bxc3 .fti ( 1 2 . . .• gS! 1 3 /l)e3 c6 14 f4 .cS I S .f3 dS 16 Whl /l)d1 1 7 Ilae I /l)fti is very comfortable

71 Leonhardt's variation 4 llk4

for Black, Grobe-Robins, corr. 1972) 1 3 �S g6 14 .as �6 (l4 .. . c6? I S .a3! lOa6 16 iOd6 cuts the black position in half, Dubois­Strautins, corr. 1973) I S lOe3 c6 16 Axa6 bxa6 17 J:ae I and White has a small pull, Morgado-Eckenfels, corr. I968.

9 ••• pf6 This disfigures the pawn struc­

ture, but otherwise Black sheds a pawn, 9 . . .• xffi?! 10 dxe4 0:0 1 1 f3 (the simplest, 1 1 .d2?! bS ! 1 2 .dS+ �h8 13 .xa8 .xf2+ 14 �dl 10:6 IS lOd2 Aa6 16 .xfS+ • xfS, Jackson-Svendsen, corr. 1994/96, might favour White on 17 a4, but it is messy, while 1 1 .dS+ �h8 1 2 0-0-0 10:6 [ 1 2 . . .• xf2? loses: 13 Ad3 .f4+ 14 �b I .f7 I S .xf7 lIxf7 16 Ilhfl loss of ma­terial is inevitable. the black queen­side pieces are too far away, Grobe-Kozelek, corr. 1976] 1 3 f3 transposes) 1 1 .. .10:6 12 .dS+ �h8 13 0-0-0 d6 14 .d2 .i.e6 I S a3 as 16 lOe3 and White has an extra pawn. and more space, Budovskis­Hempel, corr. 1970.

10 M! 10 dxe4?! .xe4+ I I lOe3 ( 1 1

Ae2?! .xg2 1 2 Af3 .gS 1 3 .e2+ �d8 14 �3 .as+ I S c3 d6 :f Siklosi-W esterinen, Copenhagen 1988) 1 1 . . .1Ib4+ 12 cl .xb2 13 lid .a3 is far from clear, Gebuhr­Vitols, corr. I 972n4.

10 ••• �d8 Black is in trouble, 1O . . . c6?! 1 1

dxe4 .xe4 12 .d6 .cS 1 3 0-0-0 with advantage Filipovs-Zarins, corr. 1991, or 10 . . . d6 1 1 Ae2 0-0 12 AhS 1Ig7 1 3 dxe4 with a pawn more, Budovskis-Grivainis, corr. I 977n9.

1 1 d:le4 .:le4 11 .d1 lOc6 13 0-0-0 d6 14 lMs :m IS f3 .eS 16 Ac4

16 •••• 15 17 f4 Black's position is extremely

uncomfortable. Gaard-Jensen, corr . 1990.

B S •••• e6

6 M! Stronger than a) 6 d3 Ab4 (6 . . . exd3+!? 7 lOe3

dxc2 is risky, but playable; 8 .xc2 c6, else White plays .A.c4, and Black will be unable to castle, [8 . ..It�ffi?! 9 Ac4 .eS 10 0-0 Ad6 1 1 g3 1&6 1 2 lllrs with dangerous threats, Pannt­Baudoin, corr. 1988] 9 Ad3 lOe7 [9 . . . 111ffi 10 0-0 Ae7? 1 1 lllfS 0-0 12 Ah6 gxh6 1 3 Ilae I and Black is crushed, Miiller-Gagsch, 9 . . . dS?? 10 AfS We7 1 1 Axc8 d4 1 2 .e4! iOd7 13 Axd7+ �d7 14 .Jtd4+' 1-0, Prat-Bravo, corr. 1986] 10 0-0 lOa6!? [ IO . . . dS 1 1 .dl !? d4?! { I 1 .. .iOd7 is safer} 12 Ac4 .ffi 13

� .eS 14 .d3 �g6 I S f4! lOxf4 16 Axf4, Sidenko-Collier, email I99S, Black is in trouble, as 16 .. .• xf4 17 �S .17 loses the queen after I S �to+) 1 I -*.xa6 bxa6 12 -*.d2 dS 13 life I .d6, Morgado-Atars, corr. 1967, and now Lein likes the positional 14 1084, but Black has chances here) 7 dxe4 -*.xc3+ (7 . . . �to!? S �3 [S f3 dS 9 �3 c6 transposes to a later line) S . . . �xe4 9 .i.c4 lOxc3 10 bxc3 -*.xc3+ 1 1 -*.d2 .eS 12 lib I .d4 and Black was hanging on, Elburg­Kozlov, corr. 1992) S bxc3 .xe4+ 9 �3 �7 with a choice:

a I ) 10 -*.d3 (chases the queen to a safer square, but it avoids Black's princ�pal defensive idea: . . . -*.e6) 10 ... Wh4 1 1 -*.c4 �bc6 ( I l . . .c6!? is possible as, although it weakens the dark squares, the king will be able to castle. 12 -*.a3 dS 1 3 -*.d3 0-0 14 0-0 with chances for both sides, Morgado-Strelis, corr. 1974177) 12 0-0 ( 12 a3?! i s less logical, 12 . . . �S 13 .ta2 IIffl 14 0-0 d6 and Black has a solid position, Kozlov­ISvendsen, corr. 1990) 12 . . . d6 (or 12 . . . �S!? 13 lie I d6 14 .d4 .xd4 I S cxd4 lOxc4 16 �S �dS 17 -*.gS h6 I S -*.xe7+ �d7 19 -*.h4 IS 20 -*.g3 Bucker) 13 g3 .to 14 �I IIfS IS f4 �dS 16 IIbl IIbS, ·unclear with complicated play'­Bucker, 17 -*.d3 gS !?

a2) 10 -*.c4 d6 1I 0-0 (+/- NCO) ' 1 . . .-*.e6 (exchanging one of the bishop pair is the positionally right decision) 12 .-..S+ ( 1 2 -*.d3 .c6!? [ 12 . . . �) 1 3 IIbl �7 14 �+ [ 14 -*.bS .cS I S -*.xd7+ �d7 16 Axb7 -*.xa2! is fine, as 17 .g4+? -*.e6 I S .xg7 sets White up for tac­tics on g2, I S . . .• c6 19 IIb2 IlagS with a useful attack) 14 . . . -*.17 IS .gS, Nigg-Agarkoff, corr. 1939, IS . . . -*.xa2 with fair chances; 12

Leonhardt 's variation 4 1Oc4 73

-*.b3 might be White's best, as he will iron out his pawn structure when Black captures on b3) 12 . . . �d7 1 3 -*.xe6+ (13 �S+ �cS 14 -*.xe6+ .xe6 IS IIb l b6 16 a4 �bc6 equal, Sb'elis-"I:omson, corr. 1974. Black's queenside is solid as a rock) 13 . . .• xe6 14 c4 �bc6 I S IIbl ( I S -*.b2 g6 16 � IIhffl 17 .xh7 1117 I S Wh4 �r5) I S . . . b6 and Black doesn't have any problems, Castelli­Gabrans, corr. 1970.

b) 6 .-..S+ leads nowhere: 6 ... g6 7 .dS .xdS S lOxdS 1086 9 �3 c6 10 -*.xa6 cxdS 1 1 .i.bS �ft). Sawicki-Lewandowski, 1997.

6 ••. �f6!? Black can take measures against

the threat of -*.c4 by playing 6 ... c6 and this may be stronger: 7 d3 -*.b4 (7 . . . dS?! S dxe4 dxe4 9 -*.c4 .eS 10 -*.xgS! AxgS 1 1 �4 with advan­tage, Grobe-Sb'elis, corr. I 974/S, 7 . . . �f6 S lOxe4 lOxe4 9 dxe4 .xe4 10 -*.d3 .eS 1 1 0-0 dS 12 lie I � 1 3 c4 -*.b4 14 -*.d2 -*.xd2 IS .xd2 .gS 16 f4 is crushing for White, Elburg-Nilsson, corr. 19S5) S dxe4 (S -*.d2 exd3 [S .. . dS?! 9 dxe4 d4 10 �f5; S .. . �ft)) 9 hd3 �to 10 0-0 0-0 1 1 .i.c4!? dS 1 2 �xdS cxdS 13 lOxdS lOxdS 14 hb4 lidS? [ 14 . . ':f5 is the only chance) I S lie I , White wins, Marin-San Pedro Fraga, corr. 1975) S ... �ft) (S . . .• xe4 9 -*.d3 � 10 �f5 .ft) 1 1 0-0 dS

74 Leonhardt 's variation 4 fi)c4

12 :e l+ � 1 3 .g4 is very good for White, Krustkains-Kozlov, corr. 1979) 9 f3 (9 Ad2 might offer more, 9 . . . lCixe4 10 lCJxe4 lIxe4 1 1 Ad3 Axd2+ 12 .xd2 .f4?! , Gnirk-love. corr. 1995, and White can infiltrate on d6 by 1 3 lCJc4! .xd2+ 14 Wxd2) 9 . . . dS 10 exdS cxdS 1 1 .d4 lCJc6 12 AbS 0-0 ( l 2 . . . AaS! 1 3 0-0 0-0 is interesting as the bishop will come to the sensi­tive gl -a7 diagonal) B Axc6 .i.xc3+ 14 bxc3 ( 14 .xc3 is more logical) 14 . . . bxc6 I S 0-0 is better for White, Diepstraten-<>ren, corr. 1992.

7 �c4 1 d3?! is worse, 1 . . . �b4 S Ad2

dS? (this is tricky, but incorrect, S . . . lCJc6! is stronger, 9 dxe4 lCixe4 10 �c4 Axc3 ! 1 1 �xe6? [ 1 1 0-0 is superior] I I . . ..i.xd2+ 12 WfI dxe6 13 .g4 Axe3 14 flxe4 �cS with three pieces and a solid position for the queen. Black is winning, Risto-Kilpela, corr. 1993) 9 dxe4? (but White had a lot of opportunities to improve, for instance 9 lCJexdS! lCJxdS [9 . . . exd3+ 10 lCJe3 dxc2 1 1 flxc2 is also good for White] 10 Whs+ Destrebecq, and 9 lCJcxdS! Axd2+ 10 Wxd2 li)xdS 1 1 Whs+ is also convincing) 9 .. . d4 10 lCif'5! (in the games Eiken-Svedenborg, Nor­way 1969, and Doroftei-Vaisman,

Soutcheava 197 1 , White was crushed after 10 lCJcdS dxe3! 1 1 lCJxc7+ Wf1 12 lCJxe6? [ 1 2 fxe3 .xe4 13 c3 Ad6 is less dear] 1 2 . . . Axd2+ 13 We2 Axe6 14 h3 lCJc6 :J=) 10 . . . Wb6 (this time, 10 . . . dxc3? is a mistake, 1 1 lCJx17+ � 12 lCJxe6 cxd2+ 13 We2 xe6 14 c3) 1 1 lCia4 .c6 12 lCJxg7+ �, Melchor-Svendsen, corr. I 99419S, when 13 Axb4 .xa4 14 .xd4 is critical.

7 •••• eS

8 d3 The most dan.serous. Others: a) S lCig4!? I&g4 9 .xg4 Ab4!

10 lCixe4? (a typical • computer mis­take', White wins a pawn, but loses his head! 10 .xe4?? Axc3 is even worse, but 10 Wdl !? is a possibility, and 10 lClds is best, 10 . . . lCJc6 [Bucker p'refers 10 . . . lCia6!? 1 1 We2 c6 1 2 f4 Wd6 13 lCixb4 lCJxb4 14 d3 We7 I S dxe4 dS with _play for the (?8wn] 1 1 0-0 Ad6 1 2 Wh4 Ae1 13 l&e7 lCixe7 14 :el dS I S Ab3 and Black's centre is difficult to maintain, Kozlov-Destrebecq, corr. 19S I ) 1O . . . lCJc6 (menacing . . . dS) 1 1 flf3 IOd4 1 2 .d3 ( 1 2 "f1+? WdS 1 3 Ad3 lCie6! threateni�.B to win the queen, 14 .f3 dS I S Wg3 lCif4 16 WgS+ .xgS 17 lCixgS :eS+ I S Wdl h6 19 lCif7+ We7 20 lCJes Wft) 2 1 lCif3 Ag4 22 h3 Axf3+ 23 gxf3

lDxd3 24 cxd3 and White's position is a mess) 1 2 . . . d5 1 3 c3 dxc4 and Black is better, Owens-Kravitz, email l998.

b) 8 0-0 .i.b4 (8 ... .i.d6!? 9 g3 b5!? 10 lDxb5 .i.b1 I I �xd6+ .xd6 12 d3 1Oc6 13 �f5 .f8 14 dxe4 �5 doesn't look completely correct, Atars-Saavedra, corr. 1969) 9 d4 exd3 to .xd3 .i.xc3 I I bxc3 d6 1 2 .i.b2 1Oc6 13 life I �d8, with a typi­cal position where Black's superior pawn formation provides ample compensation for his misplaced king, Krustkains-Strelis, corr. 1990.

c) 8 d4 exd3 transposes. 8 ••• exdJ 9 0-0 �d8 Black doesn't need his two most

,important pieces on the same open file.

10 .i.xdJ to .xd3 c6 I I f4 Wh5 12 �d2

.i.c5 13 �4 �g4 14 h3 .i.xe3+ 1 5 .i.xe3 �xe3 16 .xe3 Borrmann­Kozlov, corr. 1 99 1 , when 16 . . . b5 might bt best, e.g. 1 1 .i.e2 ( 11 �b6?! 1le8 1 8 .a3 axb6!? 19 .xa8 �c1 20 �d3 �6, trapping the fqueen) l 1 . . . 1Ie8 1 8 .i.xh5 IIxe3 with equality.

10 ••. d6 1 1 lIel

Leon"ardt 's variation 4 liJc4 75

C 5 •••• n!

This idea of Gunderam's is by far the most popular move at this stage. Black is ready to solve his develop­ment problems by gambiting his e-pawn.

6 � a) For once 6 d3 is well met by

6 . . . d5 1 1Oe'3 .ffi! (but not 1 . . .• e6 8 Wh5+! g6 9 �g6 .xg6 10 .e5+ �f7 I I �d5 which is very messy; however 1 . . .• f5 is also fine, 8 f4, with the threat of g4, 8 . . . exf3 9 �xO �ffi 10 .i.e2 .i.c5 equal, Bering-Nilsson, Copenhagen 1993) 8 d4 (if 8 �g4 .e6 9 .i.e2 Endzelis-Kozlov, corr. 198 1 , then, instead of 9 . . . exd3?!, Black could have essayed 9 .. . d4!? with the two lines to lOxe4 h5 or to �b5 Wb6 I I dxe4?! .i.xg4, winning a piece in both instances) 8 .. . c6 9 .i.e2 .i.d6 10 �g4 .i.x8� (more straightforward than 10 . . . 1fh4) I I .i.xg4 �1 12 0-0 0-0 13 .i.e3 �f5, regaining a bishop, with comfortable equality, Clayton-Peters, corr. 1995.

b) There is a major alternative, though, in 6 d4 �ffi (the others: 6 . . . .i.b4 1 �5 .£5 (1 . . .• e6?! 8 Wh5+! 8..6 9 �g6 �ffi 10 � IIg8 I I fl}f4 {simpler than I I �5 lOc6 12 .i.c4 d5 13 .i.b5 lbg2 14 .i.g5 with advantage, Weber-Fox, corr. 1989} 1 l . . ..f5 { I l . . .• n 12

76 Leonhardt 's variation 4 ll:x:4

.t.e2 with advantage} 1 2 .t.e2 dS 13 0-0 c6 14 f3 Black has nothing for his pawn, Zink-Lopez, corr. 1 993] 8 .t.<:4 liJe7 9 .t.f7+ Wd8 10 g4 .ffi 1 1 gS "f5 is unclear, Haas­Schlenker, corr. 1 988; 6 . . . dS !? 7 liJes "e6 8 lOxdS!? "xdS 9 -*.<:4 "d6 10 'WhS+ g6 1 1 .t.f1+ We7 12 .t.xg6, Maier-Rafnung, oorr. 1984, when 1 2 . . . LOffi! 13 .t.gS "xd4 14 .t.e8 -*.e6 I S LOg4 �g7 16 c3 "dS is unclear) with:

b I ) 7 liJes "e6 8 �g_S (8 �<:4 dS 9 lOxdS? lOxdS 10 WhSt g6 1 1 lOxg6 "xg6 1 2 "eS+ 'lOe7 13 "xh8 "xg2 =1= Gabrans-Zemitis, oorr. 1970173) 8 . . . -*.b4 9 -*.<:4 (9 .t.e2 0-0 10 0-0 dS is b3 below, but in this move order, 1O . . . �xc3 1 1 bx<:3 d6 1 2 lO<:4 b6 is simpler, with a balanced position in both cases, or 12 . . . lOiS!?) 9 . . . dS (better than 9 . . ... CS?! 10 �xffi gxffi 1 1 LOg4 "g6 12 liJe3 f5 1 3 IOeds with advantage, Dubinsky-T chebotariev, USSR 1968) I 0 �xffi (going in for the complications, although 1 0 �b3 is not so bad, 10 . . . 0-0 [10 . . . c6 1 1 0-0 LObd7 12 �xffi lOxffi 1 3 f3 exf3 14 "xf3 0-0 is safer, Trigan<:e­Tatlow, corr. 1995) 1 1 0-0 �xc3 12 bx<:3 lO<:6 1 3 f3 exO 14 .xf3 lOxeS [ 14 . . . liJe7? I S :ael .d6 16 <:4 �e6 1 7 cxdS LOfxdS 1 8 .e4 with advan­tage, Grobe-Magee, corr. 1 989] I S

dxeS "xeS 16 �xffi l%xffi 1 7 "xdS+ .xdS 1 8 ' .i.xdS+ Wf8 which is quite equal, MalmstrOm­Budovskis, corr. 1997) 10 . . . -*.x<:3+ (this move <:an be played at various moments, without c�ging any­thing) 1 1 bxc3 gxffi 1 2 WhS+ We7 1 3 "lOg6+ ( 1 3 �xdS? .xdS 14 LOg6+ Wd6 IS .xdS+ WxdS 16 lOih8 .i.e6 1 7 l%b 1 Wc6 [ 1 7 . . . lOi7!? 1 8 l%xb7 Wc6 J.h-1h, Yagolnitzer-Elburg, oorr. training game 1999, but Black can continue without risk as his minor pieces have good light S«iU8fCS to settle on) 1 8 <:4! �x<:4 19 l%b4 bS 20 L<:4+ bx<:4 and Black is better, Grobe­Elburg, oorr. 1994/9S) 13 . . . hxg6 14 .xh8 dx<:4 IS dS!? (better than IS o-O?! lOi7! [or I S . . . b6 16 dS .CS { I6 .. ... d7? 1 7 .g7+ Wd8 18 .xffi+ .e7 19 "xg6 with advan­tage, Grobe-Svendsen, corr. 1990} 1 7 f3 e3?! { 17 . . . lOi7!? 1 8 fxe4 .gS} 1 8�4! 'Wd7 19 "g7+ Wd8 20 "xffi+ We7 2 1 .xg6 and again White is well on top, Yagolnitzer­Elburg, corr. training game 1999) 16 f3 [ 16 dS?! .17 {there is nothing obviously wrong with 16 . . .• xdS!?} 17 :ael f5 18 f3 .ffi 19 "'7+ Wd6 20 fxe4 liJes and Black is slightly better, Grobe-Krantz, oorr. 1994I9S) and instead of 16 . . . b6!? 17 dS?! [ 17 fxe4) 1 7 . . .• xdS 18 fxe4 .f1? [ 1 8 . . . .i.b7! is strong, 19 .xa8?? .<:5+] 19 eS ! lOxeS 20 :ael .e6 2 1 .g7+ with a powerful attack, Grobe-Mel<:hor, oorr. 1991 , there is 16 . . . e3 ! 1 7 l%ael LOb6 18 "g7+ Wd6 19 .xg6 00 =1=. How­ever I S 0-0-0 lOi7 16 Wh7+ Wd6 1 7 "xg6 lOb6 18 h4 CS ; is interest­ing) I S . . .• CS ( l S . . .• eS!? is worth oonsideration: 16 .g7+ [ 16 .xc8 .xc3+ 17 We2 .xc2+ 1 8 Wfl "d3+ 19 W9 l .xdS =1=) 16 . . . Wd6 1 7 .f8+ Wd7 18 0-0 lOa6 19 :adl

.e7 and Black has the better chanc:es, Grobe-Clarke, corr. 1994/96) 16 0-0 lLld7 17 f3 .xdS?! (Black <:811 try to keep files closed with 1 7 . . . e3 !? I S :ael .eS 19 �6 �d6 [ 19 . . . gS 20 f4 pf4 2 1 lbf4 is also unclear] 20 :Xe3 .xdS 2 1 :tfe l , Grobe-Jackson, corr. 1994/9S, and now 2 1 .. .gS 22 f4 b6 is unclear, or go in for the complicated 17 .. . exf3!? I S :ael+ �d6 19 lbf3 .xdS unclear) I S fxc4 .f7 19 eS! f5 20 e6 �e6 21 :tadl .fS 22 :tfe 1+ again with a strong attack, Grobe-Clarke, corr. 1991 .

b2) If instead 7 -*.e2 dS S lOes (S lOe3?! c6 9 f3 exf3 10 -*.xf3 -*.d6 1 1 0-0 0-0 12 -*.d2 .i.e6 13 .e2 lDbd7 14 :ael :ad =, Zschorn­Clarke, corr. 1991) S . . .• e6 9 .i.gS then, as well as 9 .. . -*.b4 transposing, there is the simple 9 . . . c6! 10 0-0 -*.d6 =.

b3) 7 -*.gS is actually the most common move, but normally trans­poses, 7 . . . .i.b4 S -*.e2 (S lOes .e6 transposes to the above, but S . . .• e7 ,is also a possibility, 9 -*.<:4 d6 10 -*.f7+!? WfS 1 1 .i.hS .i.xc3+ 1 2 bxc3 dxeS 1 3 dxeS .xeS 14 .dS+ lOes I S .xcS .xc3+ 16 -*.d2 .xa I + 1 7 �e2 lD<:6 I S .xaS lLld4+ 19 �e3 lDxc2+ 20 �e2 lLld4+ Ih-1h, Grobe-Melchor, corr. 1994/9S; perhaps Black <:811 risk 20 . . .• eS) S . . . dS 9 lOes (9 lDe3 0-0 10 0-0 c6 1 1 f3 exf3 1 2 .i.xf3 .i.x<:3 13 bx<:3 .g6 equal) 9 . . .• e6 (or 9 . . .• e7) 10 0-0 O-O!? (chancing his arm a little, but if 1O . . . .i.xc3 1 1 bxc3 0-0 then 1 2 <:4 is JlOSSible, al­though the reply 1 2 . . . 1&6 seems reasonable) 1 1 .i.xa;!? (this is com­pletely obscure, but if 1 1 f3 then not 1 1 . . . exf3? 12 .i.xf3 c6 when, instead of 13 lDe2? lDbd7 14 .i.xa; =

Gyuricza-cartsson, corr. 1975, there is 1 3 .i.xa;! lba; 14 .i.g4 lbfl+?!

Leonhardt 's variation 4 liJc4 77

I S .xfl .eS 16 -*.xcS .xcS 17 .f7+ �hS I S :tfl +1-, but I I . . .-*.xc3 12 bxc3 lDbd7! 13 lDxd7 [or 13 fxc4?! lDxc4 14 lbfS+ lDxfS =] 13 . . . .i.xd7 14 .i.xa; lba; I S fxc4 lbfl +) I I . . .gxa; 1 2 lDxdS fxeS 13 lDxb4 ( l3 lDxc7 .g6 14 lDxaS -*.h3 I S g3 is also far from certain) 1 3 . . . aS 14 b3, blow for blow!, 14 . . . �hS IS .i.<:4 .g6 ( I S . . .• d7?! 1 6 lLlds bS 17 f3 ;t) 16 lLlds -*.g4 17 .i.e2 ( I cannot find anything wrong with 1 7 .d2!? .i.f3 I S g3 but it is very risky) 1 7 . . . -*.h3 I S g3 lD<:6 19 dxeS -*.xfl 20 .i.xfl lDxeS 2 1 .d4 .g7 22 �g2 :adS 23 .xc4 lDg4 0-1 , Lindgren-Hector, Rilton Cup 1990.

c) 6 lDeS!? is tricky, but 6 . . .• f5! 7 d4 (7 lDg4 c6 S lDC3 .f7 9 lDxc4 transposes to the main line) 7 . . . exd3 S lDXd3 lDa; 9 -*.f4 d6 10 -*.e2 .i.e7 1 1 0-0 0-0 is nearly level, Forte­Ruggeri Laderchi, corr. training game 1999.

d) and naturally not 6 lDxc4?? dS 7 lOes .f5, winning a knight.

6 •.• c6!

Calmly leaving the c4 pawn to its fate. What exactly does Black 0b­tain in return? Objectively, very lit­tle, just a slight lead in development and a lot of fun. In other words, all the things that he is denied in the other lines!

lLlfe3 .i.d6 14 .d3 lLlc7 I S .i.d2 lLlg6 16 .i.O?! h3 1 7 0-0-0 lOit4 1 8 .e2 .i.xg4 0-1 , Parraga-Zcmitis, COlT.) 9 d4 h4 10 lLlc2 .i.d6 1 1 1Oc3 .i.c6 12 .i.e2 lLlh6!? 1 3 lLlg4 lLlfS 14 b3 lLld7 IS .i.e3 lLlxe3 16 lOxe3 0-0-0 1 7 .d2 AIdS 1 8 .i.o lLlb6 19 0-0-0 .i.f4 20 .d3 lOc4 with sufficient compensation, Cursoux­Kostcn, Aurcc Rapidplay 200 I ) 9 . . . .i.d6 10 d4 lLlc7 1 1 .i.d3 g6!? (Black might forego this and aim to castle long, instead, i.e. 1 1 . . . .i.c6 12 c3 [ 1 2 .e2 menaces lLlfS, so 1 2 . . . g6 transposing] 1 2 . . . lLld7 1 3 �3 0-0-0 ;) 12 .e2 0-0 1 3 lLlcn lDa6!? ;t Kozlov-Budovskis, COlT. 1978179.

8 •••• 16 9 lLlO . 9 h4!? is occasionally played, but can make it difficult for White to castle short, 9 . . . .i.d6 (9 . . . h6!? 10 �S+ g6 1 1 .e2 lLlc7 1 2 lLlO .i.g7 1 3 d4 0-0 14 hS, Schott-Kempf, COlT. 1994, and now 14 . . . gS is very playable for Black) 1 0 d4 lLlc7 1 1 Ad3 h6 1 2 lLlo .i.c6 1 3 .i.d2 0-0 14 .e2 lLlg6 IS .i.xg6 .xg6 and the light-squared bishop provides some

, compensation for the pawn, Walthcr-Strautins, COlT. 1974.

9 ••• .i.d6 Probably best, and certainly the

most popular. Others: a) If Black wants to castle quccn­

side, and, despite Nunn's opinion, I still think that this is not a bad idea,

Leonhardt 's variation 4 1Oc4 79

he should make his plans now: 9 . . . .i.c6 10 d4 lLld7 1 1 .i.d3 (Sim­plest. 1 1 g3 0-0-0 [ 1 1 . . . .i.d6 trans­poses to a later line] 1 2 .i.g2 gS! 1 3 0-0 hS 1 4 h4 [ 14 lLlxdS!? .i.xdS I S .i.xgS is possible] 1 4 . . . gxh4 I S lOxh4 .i.e7 Black has kingside attacking chances, Kinnes-Kozlov, COlT. 1977 1 1 c4 .i.b4+ 12 .i.d2 .i.xd2+ 1 3 .xd2 lLlc7, Tiemann­Jackson, COlT. 1989, 14 �!? with a plus.) 1 1 . . .0-0-0 ( 1 1 . . ..i.d6 will transpose here if Black continues with . . . 0-0-0, and to the main line if he castles short) 1 2 O-O-Nunn's recommendation, not fearing 1 2 . . . gS!? ( I 2 . . . .i.d6 1 3 c4 lLlc7 14 cxdS cxdS IS .i.d2 1Oc6 is possible, Malmstrom-Strautins, COlT. 1997, Black will first play his king to b8, and then start his kingside march) 13 c4 as 1 3 . . . g4?! ( 1 3 . . . dxc4 14 lOxc4 [14 .i.xc4, Zaniratti-Van de Velden, COlT. 1999, 14 . . . g4 I S lLld2 .i.xc4 16 lLldxc4 hS] 14 . . . 8.4!? is su-�rior) 14 cxdS! cxdS I S l&g4 ( I S &S is also OK, IS . . . lOxeS 16 dxeS .xeS 1 7 lOxg4 .g7 1 8 .el ! 1 -0, Kurc-Orr, COlT. 1997, as 1 8 . . . .i.xg4 19 .i.f4 menacing Zlc l +, leads to a decisive attack) I S . . . .i.xg4 16 .i.gS is strong 16 . . . .i.xO 1 7 .i.xffi .i.xd I 1 8 .i.xh8 and White will capture on h7 and create three passed pawns.

b) However, 9 . . . d4? is definitely misguided, Black concedes the c4-square, 10 lOc4 bS 1 1 .e2+! Wd8 ( 1 1 . . . lLlc7 12 10ceS lLld7 13 c3 dxc3 14 dxc3 does not help, White has an extra pawn, and the better position, Nimzo-Comet test game 1997) 1 2 lLlces ( 12 d3?! is less ac­curate, 1 2 . . . .i.b4+ 13 Wdl h6 14 lLlces, although White is still clearly better, Peters-Bie, COlT. 1982) 12 . . . .i.d6 1 3 d3 (+/- NCO) 1 3 ... h6 14 .c4!? ( 14 .i.d2 Wc7 I S .c4 is safer, Zccha-Traut, COlT. 1987, as

80 Leonhardt 's variation 4 liJc4

White can castle long when the black rook comes to eS) 14 . . . 1Oc7 IS .xd4 lOfS 16 .c4 l:[eS, Grobe­Szilagyi, corr. 19S3, when 1 7 d4! lOxd4 IS .xd4 lOd7 19 �f4 lOxeS 20 0-0-0 Wc7 2 1 .xd6+! .xd6 22 :Xd6 Wxd6 23 lOxeS is crushing.

IO d4 fiJe7 I 0 . . . �c6 will transpose.

1 1 c4 The most important move,

although there are two reasonable alternatives:

a) 1 1 g3 �c6 (if 1 1 . . .0-0 1 2 �g2 lOf3 [ 1 2 . . . .i.c6 transposes] 1 3 h3 ! lOxe3 14 .i.xe3 followed by c3, �3, 0-0-0 with advantage, whilst the attempt to stop the king from castling, 1 1 . . .b6?! 1 2 .i.g2 h6, should not be met by 13 c4?! dxc4 14 0-0 0-0 [ 14 .. . lOd7!?] I S lOg4 .fS 16 lOgeS, Tiemann-Krantz, corr. 1991 , when White is only slightly better, but by B lOcS 0-0 [if 13 ... .i.xeS 14 lOg4 with advantage] 14 f4, and now Black erred with 14 . . . lOg6? when I S lOxdS! cxdS 16 .i.xdS+ wins material, Malmstrom­Elburg, corr. 1994/9S) 1 2 .i.g2 lOd7 13 0-0 0-0 (unfortunately, the natu­nl 1 3 . .. 0-0-0 allows 14 lOgS! hS?! [if 14 . . .• xgS?! I S lOc4 .ffi 16 lOxd6+ WbS 1 7 .i.f4 WaS leaves the d6 knight trapped, but there is no way of getting at it, i.e. I S l:[e I lOg6

{or I S . . . h6 19 .e2 with advantage} 19 �gS with advantage; 14 . . . �8S?! I S lOg4 .g6 16 f4 h6 1 7 lOh3 .i.c6 I S lOCs and White has a dominating position; so perhaps Black's best bet is 14 . . . �f3 I S lOXf3 lOxf3 16 c3 h6 1 7 lOO gS, but White is better] I S l:[e l �gS 16 lOO? [ 16 h4 is good] 16 . . . gS � Borrmann-Krustkalns, corr. 199 1 ; 1 3 . . . gS? is a blunder, 14 lOxgS! winning quickly, Morgado­Grivainis, corr. 1977, but 1 3 . . . h6!? is the best alternative, 14 b3 0-0-0 I S c4 gS 16 .i.b2 hS, with fair attacking chances, Budovskis­Svendscn, corr. 1995) 14 b3 ( 14 �d2 is more passive, 14 . . . bS [ 14 . . . lOfS is also reasonable, I S �g4 .g6 16 lOfeS lOxeS 17 dxeS? { 17 lOxeS is forced} 1 7 . . . �cS . { 1 7 . . . lOxg3 ! is p()SSibly even stronger} I S �f4?! lOh4! winning material, Borrrnann-Ticrnann, corr. 19S5] I S �c3 lOfS 16 lOg4 .g6 1 7 lOfeS .i.xeS I S lOxeS lOxeS 19 dxeS lOh4! with good chances on the light squares, Korchnoi-Dcstrebccq, simul, Lyon 1979) 14 . . . lOg6!? (I4 . . . :'c8! IS .i.b2.lDf3 is also pos­sible, and should be similar to the 14 .i.d2 lines) I S .i.b2 hS?! ( I S ... :'eS rust, is more accunte) 16 h4 :'eS 1 7 lOgS! ( 1 7 c4!? is more common, but allows Black a strong · attack: 1 7 . . . �g4!? I S lOxg4 bxg4 19 lOh2 lOxh4 20 gxh4?! [20 lOxg4! is stronger, 20 . . .• dS!? 2 1 gxh4 {21 .i.hl !?} 2 l . . ..xh4 22 .i.cl? {22 0 lOffi 23 lOxffi+ :Xffi 24 l:[e I is more testing} 22 . . . lOffi 23 lOxffi+ l:[xffi 24 f4 :b6! and Black can force a dnw-Biicker. 2S .0 Wh2+ etc.] 20 . . .• xh4 2 1 f4 g3 22 lOO WhS 23 .i.c I lOffi and the attack eventually won through, Morgado-Gundenm, corr. 1970) 1 7 . . . lOf"4. Dicpstraten­Pot, corr. 1975, and now there is nothing wrong with taking the

piece: I S gxf4 .xf4 19 lOO .i.g4 20 lOxg4 bxg4 2 1 .i.c I .fS 22 lOgS Black has almost nothing for the material.

b) I I .i.d3 (H) (or 1 l . . ..i.e6 12 h3 !? [ 1 2 (H) transposes] 1 2 . . . lOd7 1 3 lOg4 [or first 1 3 .e2 0-0-0 14 lOg4 .bg4 I S bxg4 h6 16 gS with advantage, Grivainis-Hel!J), corr. 199 1 ] 1 3 . . . .i.xg4 [ 1 3 . . .• f7?! 14 lOgS] 14 bxg4 We6+ I S .e2 .xg4 16 WO ! lOffi 1 7 :b4 .d7 I S lOgS with advantage, Strautins­Destrebecq, corr. 1990/92) 1 2 (H) (or 1 2 h3 lOg6 1 3 .i.xg6?! White stops the knight coming to f4 by eliminating this piece, but the white light-squared bishop is an imp()rtant

. piece. [ 1 3 lOg4 is better, 1 3 . . .• e7+ 14 WO !? { 14 .e2 lOh4 I S lOxh4 .xh4 1 6 .i.e3 is simpler} 14 . . . .i.fS I S .i.gS, Strautins-Filipovs, corr. 1992, and now I S . . .• f7 is fine for Black] 1 3 . . .• xg6 14 lOg4!? [ 14 .d3?!, Grivainis-Krantz, corr. 199 1 , should be met by 14 . . .• xd3] 14 . . . .i.xg4 [but 14 . . . .i.fS and .. . lOd7 is more ambitious] I S hxg4 .xg4 16 .d3 h6 1 7 Wfl , Strautins­Downey, corr. 1990/93, 1 7 . . .• fS = .

12 c4 lOg6!? [ 1 2 ... .i.e6! 13 �3 may transpose into the main line] 1 3 hg6 .xg6 14 cxdS cxdS I S �3 lOc6?! 16 .xdS+ WhS, Tanner­Leisebein, corr. 19S9, and now 17 (H) questions the correctness of Black's _play) 12 . . . lOg6 1 3 c3 ( 1 3 .i.xg6 .xg6, Gross-Zemitis, San Francisco 1965, is also fine for Black) 1 3 . . . lOf4 14·lOel?! ( 14 .i.c2 retains the bishop) 14 . . . lOxd3?! (I4 . . . �6! threatens mate on h2, IS g3 00+ 16 Wg2 lOxf2!? [16 . . . lOf4+!? 17 gxf4 J:h3+ I S WgI .xf4 looks strong, but White can defend by 19 .i:xh7+! Wxh7 20 �S+ exchanging queens, with a slight plus] 1 7 hfl hfl+ I S Wxfl

Leonhardt 's variation 4 lDe4 81

.xh2+ 19 101g2 .i.xg3+ 20 WO .i.h3 with a strong attack) I S .xd3?! b6! 16 g3 .i.a6 1 7 c4 bS ! I S b3 bxc4 19 bxc4 dxc4 ; Krustkalns­Downey, corr. 1993.

1 1 ••• 0-0 Alternatively, 1 l .. ..i.e6 12 cS ( 12

cxdS lOxdS 13 lOc4 is also liked by NCO, 1 2 �3 (H) transposes) 12 . . . .i.c7 1 3 .i.d3 (H) ( 13 . . . llXl7?! 14 .i.d2 (H) I S .c2 is given as advantageous for White by NCO, and, indeed, I S . . . .i.h3!? 16 Q-(H) .i.xg2 17 .i.xh7+ WhS I S lOx� .xf3 19 :he I ! does seem good) la .i.d2 1Og6! (again this typical idea, the knight heads for f4) I S .c2 .i.h3! 16 .i.xg6 bxg6 17 0-0 ( 17 :gl .i.xh2!) 1 7 . . . .i:xg2 I S lOxg2 .xO 19 .xg6 lOd7 and Black is only slightly worse, Elburg­Rozzoni, corr. 1999.

12 �3 .i.e6 Destrebecq's suggestion of

1 2 . . . WhS!? has found no takers, 1 3 .i.d2!? ( 1 3 cxdS?! cxdS 14 lOxdS lOxdS I S .xdS .i.b4+ is unneces­sarily risky, 1 3 .i.e2 is preferred by NCO) 1 3 . . . dxc4 14 .i.xc4 lOd7 I S 0-0 lCJb6 16 .i.d3 .i.e6 1 7 .c2 is better for White.

13 c5 To prepare the queen's escape, 13

.xb7 1O:l7 14 .a6!? ( 14 cS .i.f4 is probably best, and transposes, but not 14 .i.e2? lOb6! I S cS [ I S .a6? .i.b4+ 16 Wd I .i.cs is worse] I S . . . .i.cs trapping the queen) 14 . . . lOfS!? ( 14 . . . .i.b4+ I S .i.d2 .i.xd2+ 16 Wxd2 is possible, as the white king's awkward position will provide Black with counte!play) I S .i.d2?! lOxd4 16 lOxd4 Wxd4 17 .xc6 .f4 IS cxdS lOes 19 .a6 .xf2+ 20 Wdl :f4 and Black has a powerful attack, Zaniratti-Bartsch, corr. 2000.

13 •• ..*.f4 14 .xb7 lOd7 15 �

82 Leonhardt 's variation 4 qx4

' ! ' NUM and NCO. Otherwise, 1 S .i.d3 gS 16 h3 ! (so

that when Black plays . . . g4, White will obtain the open h-file for his rook; on 16 h4?! g4 1 7 �gS .i.xgS I S .i.xh7+ '*'g7 19 hxgS .xf2+ 20 '*'d 1 Itf4 Black has dangerous tacti­cal possibilities against the white king, Priede-Grivainis, corr. 1970) 16 . . . hS ( 16 . . . ltfbS!? is an interesting alternative, 1 7 .a6 hS I S .a4 g4 19 hxg4 hxg4 20 �g 1 .i.xe3 2 1 .i.xe3 :Xb2 22 Ith6 and the active black rook, and open f-file will pro­vide various tactical possibilities for Black, Rozzoni-Koetsier, corr. 1999) 1 7 Wb4 g4 (Black cannot af­ford to hang around, 17 . . . aS?! I S .c3 a4 19 .i.d2 .i.c7 20 0-0-0 and the white king has found safety, Koetsier-Elburg corr. 1999) I S hxg4 hxg4 19 �gl ltaeS 20 �2 .i.c7 2 1 Itn .i.fS 22 .c3 .i.aS! 23 .xaS .i.xd3 24 �g4 .g6 2S �3 .i.xe2 26 '*'xe2 �fS (the black attack con­tinues unabated) 27 Wb4 .xg2 2S '*'c:l3 :Xe3+! 29 .i.xe3 .e4+ 30 Wc3 �e3 3 1 Itgl+ �g2 and Black is winning, Elburg - Ruggeri Laderchi, corr. training game 1999.

15 ••• r.; I S . . . labS 16 .c2 ± NCO. 16 b3 b5 Black will try to blast the f-file

open. 17 .cl

Eyeing h7, as does 17 .d3 �fS (perhaps Black should play 1 7 .. ':17, guarding h7, as in the main line) I S .i.e2 ( I S .i.d2! intends long castling, I S . . . �xe3 19 .i.xe3 .i.fS 20 .d2 and White is fine) IS . . . ltaeS 19 �2!? .i.xc l 20 :Xci �g7!? (20 .. . g4 2 1 hxg4 hxg4 22 � g3 23 fxg3 .g7 24 g4 lOh6 2S �f3 .i.xg4 and Black maintains the initiative, Ruggeri Laderchi-Elburg, corr. training game 1999) 2 1 �3 &4 22 hxg4 hx�4 23 �7+ '*'fl 24 Ith6 pf3 2S Itxffi+ �xffi 26 .d3 fxe2 is unclear, Tatai - Ruggeri Laderchi, corr. 2000.

17 .. .ltn! Bucker's move, Black guards h7

to revive the threat of .. . g4. Other moves have been tried:

a) 1 7 . . . ltaeS IS .i.d2 � 19 .c3 84 20 hxg4 �e3 2 1 fxe3 .i.g3+ 22 Wd I .i.xg4 and Black does not have enough compensation, Tait-Elburg, corr. I999.

b) 1 7 . . . �fS I S .i.e2 ( I S .i.d2! is stronger, I S . . . g4, otherwise White simply castles, 19 hxg4 �xe3 20 .i.xe3 .i.xe3 [20 . . . hxg4?? 2 1 �7 mate is the point of White's earlier queen move] 2 1 fxe3 .i.xg4 22 0-0-0 and White has a winning advantage-NCO) I S . . . ltaeS 19 .i.d2 ( 19 O-O? • out of the frying pan .. . ' 19 . . . .i.xe3 20 .i.xe3 g4 2 1 hxg4

bxg4 22 tOes IOxeS 23 dxeS 1fh4 and White will soon be mated along the h-line, Faraoni - Ruggeri Laden:hi, corr. training game 1999) 19 . . . l&e3 20 -*.xe3 .i.xe3 2 1 ne3 -*.f5 22 .d2 g4 23 bxg4 hxg4 24 lDh4 -*.h7 2S ()-(H) .gS and Black will regain one of his pawns and re­tain an active set-up, Faraoni­Ruggeri Laderchi, corr. training game 1999.

18 -*.d2 g4 19 bxg4 bxg4 20 IOgI -*.eS! !

21 dxeS Black also has good compensation

after 2 1 tOe2 .xf2+ 22 Wdl -*.hS 23 .d3 1Of6 24 -*.el 1Oe4!, and af­ter 2 1 ()-(H) -*.xd4 22 -*.b4 as 23 Jhd4 axb4 24 tOe2 Jha2-my analysis. . 21 •••• xf'2+ 22 Wdl lOf5 23 -*.e2 23 tOe2 .xe3! 24 .a4 .e4 (or

24 . . .• xeS 2S .xg4+ .g7) 2S .xc6 :am 26 .xe6 IOg7 obliging White to part with his queen.

23 •••• xe3! 24 -*.xg4 .d4 25 -*.xf5 -*.xf5 26 .c3 -*.g4+ 27 100 .xc3 28 -*.xc3 IbO 29 pO -*.xO+ 30 Wc:2 -*.e4+ 31 Wd2 -*.xbl 32 Jbbl

32 e6?! IOxcS 33 Ilxhl 1Oe4+ 34 Wc2 IOxc3 3S Wxc3 lleS 36 :c 1 Wg7, winning the e-pawn, although a draw is probable.

32 ••• lOxc5 33 Ilb6 1Oe4+ 34 Wd3

Leonhardt 's variation 4 fiJc4 83

This is also likely to be a draw. Many thanks to Stefan Bucker for this original analysis.

C2 7 d3!

This idea of Budovskis, refusing the pawn and playing for the attack, is the modem preference.

7 ••• nd3 8 -*.xd3

8 ..• d5 Black cannot really do without

this move, but might be able to delay it a little:

a) S . . . -*.b4!? 9 0-0 -*.xc3? (White's dark-squared bishop will l1"!ake Black regret this; 9 . . .tt�f6 is interesting, as if 10 -*.d2 [ 10 IOg4--Elburg] 10 .. . dS might be su­perior to S .. . dS 9 0-0 Ab4, and 9 . . . dS transposes to S . . . dS) 10 bxc3 tOe7 1 1 -*.a3 dS ( 1 1 . ..0-0 12 -*.d6! is quite hopeless) 12 Ilel 0-0 ( l2 . . . WdS 13 -*.xe7+ .xe7 14 IOxdS +/- Elburg-Diepstraten, corr. 19S5) 1 3 -*.xe7 .xe7 14 -*.xh7+! Wxh7 I S �S+ WgS 16 10xdS win­nin� the black queen, 16 . . .• xe l+ 1 7 :Xe 1 cxdS the black pieces are incapable of helping, SCnechaud­Schirmer, corr. 1992, e.l. I S :Cs 1OcI7 19 .xdS+ WhS 20 WhS+ WgS 2 1 lle7 1Of6 22 .g6.

b) s . . . IOf6 9 0-0 -*.cS?! (the bishop is exposed here; better 9 . . . -*.b4 as above) 10 IOg4! O-O?

84 Leonhardt 's variation 4 �4

( 1O . . . lDxg4 1 1 .xg4 0-0 is better) I I lDxft)+ .xft) 1 2 �S! even stronger than ( 12 .i.xh7+ �h7 13 ebS+) 1 2 . . . .i.xf2+ 1 3 Whl g6 14 .e2 eb4 IS lOe4 winning, Tener­Grobe, corr. 1971172.

9 0-0 If Black is careful, then 9 .e2

�resents no problems: 9 . . . .i.e6 10 1Of5 1ixI7 I I lixI4 ( 1 1 .i.f4 �S 1 2 1ixI6+ .i.xd6 13 .i.xd6, Purins­Kozlov, corr. I 977n9, 1 3 . . . lDxd3+ 14 .xd3 0-0-0 =.) 1 1 . ..0-0-0 12 0-0 ( 12 lDxe6 lieS 1 3 0-0 lIxe6 is fine for Black) 1 2 . . . lDcs 1 3 lDxe6 lDxe6 14 lIel lieS and Black is very close to equalising, Niemand-Tatlow, corr. I99S.

And thus we attain the critical p0-sition for Black in this variation. Black has tried most legal moves, but only three of them come close to being satisfactory,

C21 9 . • id cn 9 .. id6 A brief round-up of the others, in

order of playability: a) 9 ... .i.b4?! 10 lDcxdS! (Critical,

the other moves are fme for Black: 10 li)e2 is calmer, but not very ef­fective, 10 . . . li)e7 [ 1O .. . lOft)!?] I I c4 0-0 [ 1 1 . . . .i.e6 12 1rb3 dxc4 13 lDxc4 lOa6 14 lOf4, Budovskis­Melchor, corr. 1993, 14 . . . JldS!=, and 14 a3 !? bS IS .xb4! 0-0 might

favour White, but only slightly] 12 1rb3 .i.d6 1 3 cxdS lDxdS?! [ex­changing the e3-knight helps White's development, 1 3 ... cxdS! 14 1Oc3 .i.e6!? { l4 . . . JldS} IS .xb7 1Obc6 brings Black excellent play for the pawn] 14 lDxdS cxdS I S lDc3 .i.e6 16 lObS with a slight edge, �ozlov-Gllrd, corr. I 99419S. 10 .A.d2!? might be good, lining up threats along the e I-aS diagonal, but 10 lie I is less effective here, 1O ... li)e7 [ IO .. . .i.e6? loses-see 9 .. . .i.e6] and now l l li)exdS?? cxdS 1 2 lDxdS? .xdS 13 .i.g6+ loses to 1 3 ... WdS) 10 . . . cxdS I I lOxdS the attack is very dangerous, but I don't know if there is a forced win:

a l ) 1 l .. ..i.cS!? 12 .i.f4 Wf8 1 3 .i.c4 lDc6 14 .i.e3! .i.d6 I S 1Ob6 .i.xh2+ 16 Whl .ft) 17 lDxcS! ( 1 7 .i.cS+ WeS I S lDxaS 1rh4 and White will have to face a dangerous discovered check) 1 7 ... 1rh4 ( 1 7 ... JlxcS!? I S .i.xgS .i.eS 19 .g4) I S .0+ lOft) 19 .i.e6 .i.eS+ 20 �3 .xh3+ 2 1 ph3 and Black cannot exploit the cS-knight's posi­tion, and so is worse, Malmstrom­Minglia, corr. 1997.

a2) I l .. .LS? 12 b4 .i.dS (this loses, but what else? 1 2 ... .i.e6 13 lIel li)e7 14 .i.c4 1-0, as 14 . . . US I S Jlxe6, Krustkalns-Borrmann, corr. 1991 , 12 . . . WfS 13 .i.c4 .i.xb4

14 Ab3 is also hopeless for Black) 13 :el + wm 14 .i.c4 .d7 I S bS! Ae7, Littleboy-Borrmann, corr. 1991 , when 16 .i.a3! Axa3 1 7 .0+ is decisive.

a3) 1 1 . . .• xdS?! 1 2 Ag6+ bxg6 13 .xdS lOc6 14 c3, Svendsen­Sireta, corr. 1997, Black has three pieces for the queen, but his open king is a real problem.

a4) l l . . .lOC6!? 1 2 IOxb4 (alterna­tively, 1 2 .e2+ � 1 3 lOxb4 1Oxb4 14 Ac4 .ffi?! [ 14 . . . oo! I S :e 1 Ad7] I S Ae3 1&6 16 Acs+ fiJge7 1 7 :rei Af5 I S lIadl :eS 19 Ild6 .gS 20 :Xc6! bxc6 2 1 .eS hS 22 he7+ 1-0, Svendsen-Downey, COff. 1997) 12 . . . lOxb4 13 :el+! (or 1 3 AbS+ 1&6 14 :el+ IOge7 I S "Axc6+ bxc6 16 AgS 0-0 17 Axe7 .xt2+ IS Wh 1 when the opposite­colour bishops are unlikely to save Black) 1 3 . . .1&7 14 AgS 0-0 I S he7 .xt2+ 1 6 Whl lOxd3 1 7 .xd3 :eS I S Ah4 .17 19 :xeS+ • xeS 20 :e 1 Ae6 2 1 .d6 Wf7 22 .c7+ WgS 23 .xb7 and White's two extra pawns should be enough, Laderchi Ruggeri - Elburg, corr. 2000.

as) 1 1 . . . Ad6? transposes to 9 . . . Acs 10 lOexdS.

b) 9 . . . Ae6!? is reasonable, avoids all sacrifices on dS, or c4, and may prove to be one of Black's best lines. 10 :e 1 and:

Leonhardt 's variation 4 liJc4 85

b'l) 10 . . . Ae7 1 1 lOe2! (seems to be more potent than 1 1 IOg4 lOd7 12 .e2 Axg4 [ 1 2 . . . I&S!?] 1 3 .xg4 0-0-0 [ 1 3 . . . lOgffi 14 .e2 1&5 I S Af5 g6 16 Ah3 d4?! 1 7 b4 lOa6 I S Ah6! +/- Elburg­Schoppm�er, corr. 19S7] 14 Af5 IOgffi 1 S Wh3 Acs with chances to equalize, Kozlov-Grobe, corr. 1977nS) 1 1 . . .AcS!? ( 1 l . . .lOd7?! 12 lOd4 0-0-0 1 3 lOxe6 .xe6 14 1Of5 .ffi [ 14 . . .• 071 I S :Xe7 1-0, Kozlov-Strelis, corr. 1977, as IS . . . lOxe7 16 lOd6+] I S .g4! with advantage Budovskis- Kozlov, corr. 19S0) 1 2 fiJg4 ( 12 IOf4!? .xf4 13 1Of5 .xf2+ 14 Whl Wf7 I S JUt picks up the black queen, van den Berg-Hage, corr. 19S5, although Black has almost sufficient coml!CD­sation) 1 2 . . . Ax.g4 ( 12 . . . lOe7 13 �f4 Acs 14 lOes Wffi I 5 �S+ g6 16 �6) 1 3 IOf4+ lOe7 14 .xg4 0-0 I S Ae3 with advantage, according to Bucker, although I S . . . Axe3 16 :Xe3 1Of5 doesn't look so terrible .

b2) 10 . . . lOd7 1 1 Af5 O-O-O?! ( 1 l . . .Axf5 12 lOxf5+ WdS avoids the loss of a pawn) 12 Axe6 .xe6 13 lOexdS with advantage, Elburg­Diepstraten, corr. 1990.

b3) 1O • . • Ab4? 1 1 1Of5 (not bad, but 1 1 Af5! Axf5 12 lOexdS+ � [ 1 2 . . . Ae7?! 1 3 :Xe7+! lOxe7? 1 4 liJc7+ WfS IS .dS+] 1 3 lOxb4 just wins, and I l lOexdS cxdS 12 .g4 is also pretty effective) 1 1 . . . Wd7 1 2 lOd4 :i.cs 1 3 lOxe6 Axt2+ 14 Whl Axe I I S 1&5+ winning, Grivainis­Alloin, corr. 19S5.

b4) 10 ... lOe7!? 1 1 lOe2 lOd7 12 lOd4 0-0-0 is playable, Diepstraten­Meester, Hilversum 1984.

bS) 10 ... AcS? the pressure on f2 does not equal the pressure along the e-file: 1 1 Af5! Axe3 ( I l . . .Axf5 12 lOxf5+ lOe7 1 3 .g4 is also hor­rible, but not 1 3 lOxe7? Axt2+) 1 2

86 Leonhardt 's variation 4 fiJc4

.i.xe6 .xf2+?! ( 1 2 . . . .i.xf2+ 13 Whl • f6 14 lIe2 lDe7 I S .i.xdS is supe­rior, but Black can hardly hope to survive) 1 3 Whl d4 ( 1 3 . . . .i.xcl 14 .i.xg8+ Wd8 I S .i.xdS) 14 .i.xe3 dxe3 I S .i.xg8, with a decisive advantage, Tiemann-Alloin, corr. 1986, as I S .. . llxg8 16 .d4 �7 1 7 IIxe3+ is calamitous.

c) 9 .. . 1Of6?! fails to keep control of fS, 10 1Ie1 .i.e7 l l lOfS .i.xfS 12 .i.xf5 lObd7 (obviously 1 2 . . . (H)?? 13 .i.e6 is out of the question, and 12 . . . g6? 13 .i.e6 .g7 14 .i.c8 a6 IS .i.e3 is also crushing, Diepstraten­Purins, corr. 1983/8S) 1 3 :l.f4 ( 1 3 .d4 WfB [ 1 3 . . . Wd8] 14 b4 h6 I S .i.a3 .i.d6 [ I S .. . aS !?] 16 11e6 is also rather good, 8orrmann-Lonsdale, corr. 199 1 ) 13 .. . WfB ( l 3 . . . 0-0-0? 14 .d4 lOhS I S lObS! 1 -0, Sinke-De 8oer, corr. 1986; 1 3 . . . lIfB 14 .i.e6 1Whs I S g4 1Wh4? ( l S . . .• g6) 16 .i.g3 .gS 1 7 .e2 d4?! 1 8 .l.xd7+ lOxd7 19 lOe4 +1- Kozlov-Lonsdale, corr. 199 1 , but 1 3 . . . Wd8 may be the best chance) 14 lDe2 lObs I S .i.xd7 lOxf4 16 lOxf4 .xf4 1 7 lIe3 with advantage, Kozlov-Gaard, corr. 1989/9 1 .

d) 9 . . . lOh6?! 10 ltiexdS!? ( 10 lIel is simpler) 10 . . . .i.cS? (he might as well take the piece, ...... though 10 .. . cxdS 1 1 .i.bS+ .i.d7 1 2 lIel+ Wd8 is not very pleasant) 1 1 .e2+ Wd7, Grobe-Schwibbe, corr. 1976, 12 .i.xh6 cxdS 1 3 .g�+. winning.

e) 9 . . . d4? 10 .i.c4 Wd7 ( l0 . . .• g6? 1 1 .xd4 .i.d6 12 .i.xg8 Ilxg8 13 1Oc4 1 -0, Elburg-Korhonen, corr. 1988. Or 1O . . .• f6 1 1 lOe4 .eS 1 2 lOg4 with advan�ge) 1 1 1Whs+ Wd8 12 IIdI lOf6 13 Wh4 cS 14 b4! b6 IS bxcS bxcS 16 � .i.e7 17 lOgS and Black is on his knees, Diepstraten-Alloin, corr. 1 983.

f) 9 .. . .i.e7?! will transpose else­where after both 10 lIel .i.e6! (or

10 . . . 1O(6) and 10 lDexdS cxdS I 1 ltixdS .i.d6 1 2 lie 1+ .

C21 9 •• "'*.CS

In this main line Black is nearly lost, and there being no obvious im­provements for Black, this move must be considered doubtful.

10 ltia4 Played to put pressure on Black's

centre with c4. Others: a) 10 lDexdS!? this is less effec­

tive here although White still gets two pawns and an attack for his �iece, 1O . . . cxdS 1 1 ltixdS ( 1 1 lObS?! is worse here than in the position with the bishop on d6, as nothing is attacked, 1 1 . . . 1Of6 [ 1 1 . . . .i.fB? transposes to 9 . . . .i.d6 10 �xdS] 1 2 lIel+ [ 1 2 .i.e3 .e7 1 3 ..*.f4 (H)! 14 1Oc7 lOg4 counterat­tacking against f2] 1 2 . . . WfB 1 3 .i.e3 .i.xe3 14 IIxe3 � and White's compensation is insufficient) with:

a l ) 1 1 . . .1Oc6! ( 1 1 . . . .i.d6? 1 2 1le1+ WfB [ 1 2 . . . lDe7 13 .i.c4 .fB? 14 llxe7+ 1 -0, Littleboy-Riegsecker, corr. 1986, � transposition] 13 .i.c4 .g6 14 Illf4 1-0, Diej)straten­Tiemann, corr. 1990; 1 1 . . .1llf6?! 12 lOxf6+ gxf6 13 Ilel+ .i.e7 14 .i.c4 �7 IS .i.dS .g7 16 lIe3 with pressure along the central files, Malmstrom-Niemand, corr. I99S) 12 .i.f4 ( 1 2 1Oc7+!? .xc7 13 1Whs+

.f7 14 .xcS .i.e6!? Tatai-Ruggeri Laderchi, corr. 2000, [ l 4 . . .• e7!? IS .c3 .16 16 lIe l + lOge7 17 �3 .i.CS! may be better; if White takes the b-pawn, then Black can castle short] I S lIel fijge7 [ I S . . . O-O-O?! 16 b4 �bS 17 bS JldS IS .e3 lOdS 19 .i.c4 wins the exchange, although Black will have some play] 16 .i.gS JldS, unclear) 1 2 . . . �f8 ( 12 . . . .i.b6? 1 3 :el+ fijge7 14 .i.c4 .CS?! [al­though if 14 . . .• g6 then I S {jjxe7 {jjxe7 16 .dS is crushing anyway] IS (jjxe7 fijxe7 16 lIxe7+! forcing mate, MalmstrOm-Diepstraten, corr. 1994) 1 3 .i.c4 ( 1 3 /O;7? .xf4! 14 {jjxaS fijf6, Black already has two pieces for a rook, and will also win the as-knight eventually, I S .i.bS .i.d7 16 .i.xc6 .i.xc6 17 .dS+ �t7 I S .c7+ [White exchanges queens, for if I S .xhS? .e4 and there is no defence against mate] I S . . .• xc7 19 (jjxc7 lIcS and the three pieces soon triumphed. Niemand-Malmstrom. corr. 1995) 1 3 . . . .i.e6 14 .i.e3 ( 14 .i.gS fijd4!? I S b4 fijf6 16 bxcS .i.xdS 17 .xd4 .i.xc4 and the position is messy. but Black is better, De Jong-Sireta. corr. 1994/9S) 14 . . . .i.d6 I S fijf4?! .i.xf4 16 .i.xe6 .i.xh2+ 17 �xh2 .xe6 IS :el .CS and Black i s winning, Diepstraten-Sireta, corr. 199419S.

a2) 1 1 . . . .i.e6!? is the alternative, 12 lOf4 ( 12 .e2?! �dS 13 .i.c4 [White's attack soon evaporated on 13 lid I .i.xdS 14 .i.e4 fijf6 I S .i.xdS (jjxdS 16 c4 lieS! 17 .i.e3 .i.xe3 I S fxe3 .e7. Laderchi RU88.eri - Elburg, corr. 2000] 1 3 . . . lllf6 14 .i.gS .i.xdS I S .i.xf6+ .xffi 16 .i.xdS fijc6 17 c3 �c7 and Black has overcome the worst of his problems, Kessel-Elburg, corr. training game 1999) 1 2 . . . .i.d7! ( l 2 . . . fije7?? 13 (jjxe6 .xe6 1 4 �S+ 1-0. Diepstraten-Kozlov,

Leonhardt 's variation 4 liJc4 87

corr. 1991) 13 .e2+ fije7 14 .i.c4 .CS ( l4 . . .• f8!?) IS .i.e6! ( I S lOdS?! �dS! 16 fijf4 fijbc6 17 fije6+ .i.xe6 I S .i.xe6 .eS 19 Jld1+ �c7 -/+ Diepstraten-Krantz, corr. 1991) I S . . . he6 16 {jjxe6 �t7 17 fijc7 fijbc6? (Black should prefer 17 . . . {jja6 IS g4 .cS 19 .0+ �g6 although White is better) I S g4! De­viation! I S . . .• eS 19 .0+ Wg6 20 .i.f4 .16 2 1 .e4+ � 22 Wc4+ 1-0, Nater-Torre, email l999.

b) 10 lIel For once, this is not in the least bit troublesome for Black: lO . . . fije7 1 1 {jja4 .i.d6 ( 1 1 . . . .i.b4!? 1 2 c3 .i.d6 13 c4 transposes, de Wit-Diepstraten, Hilversum 1990) 12 c4 ( 1 2 .e2 0-0 == Diepstraten­Tiemann, corr. 1984) 12 . . . 0-0 13 .c2 �S 14 g3 .i.h3 IS .i.e2 .t7 and, with the central pawn and pressure along the Hile, Black is better. Makkinga-Diepstraten, Breda 19S7.

c) lO fijg4 fije7 ( l0 ... lOd7!?) I I fijes .ffi 12 �S+ g6 1 3 .sS .xgS = Mercadal-Melchor, corr . 19S7/SS.

IO .. id6 Although not particularly good,

10 . . . .i.e7?! is playable. I I c4 fijffi 12 cxdS cxdS ( 1 2 . . . {jjxdS!? 13 (jjxdS [ 1 3 .i.c4 .i.e6 14 .i.xdS cxdS I S fijc3 wins a pawn, for minimum compensation] 1 3 . . . cxdS 14 .c2, Svendsen-Littleboy, corr. 19S6, with a plus) 13 fijc3 .i.e6 14 .i.CS! .i.xCS I S {jjxCS fijc6 16 {jjxe7 {jjxe7 (maybe 16.:.�e7 17 :e1+ � is best, although White has much the better game. Sireta-Dic:pstraten, corr. 1997) 1 7 .a4+ lllc:6 (or 1 7 . . . lOd7 IS .i.gS fijc6 19 lIfe1+ � Kozlov-Melchor, corr. 1991 , when 20 �S � 2 1 .i.e3 c:I4 22 hc:14 (jjxc:14 23 �+ �gS 24 .xc:14 will win) I S lIe1+ �d7 19 .i.gS JIbeS 20 �S and, again, the

88 Leonhardt 's variation 4 1Dc4

d-pawn is lost, Malmstrom- ' Diepstraten, COrT. 1997.

10 .. . Axe3? Black cannot afford to concede the defender of his dark squares so easily, I I .i.xe3 fi:Je7 1 2 lIel 0-01! 1 3 AcS lieS 14 .e2 bS ( 14 . . . WfS I S Axh7) IS 1Oc3 as 16 .e3! controlling a7, and winning, Budovskis-Svendsen, COrT. 1994/9S; 16 Axe?? Jla7 is the point of Black's play.

1 1 c4

1 1 ••• lDe7 Black must try to complete his

development. Others: a) 1 1 . . . d4?! presents Black with

overwhelming difficulties: a l ) 1 2 10:2 cS 1 3 b4! (not 1 3

lIe l+?! lik7 14 b4 0-0 hitting 12, I S bxcS .xf2+ 16 Whl Axh2 17 1WhS? [accepting the draw by 1 7 Wxh2 1Wh4+ I S �gl .12+ is more prudent] 17 . . . IOf5 and Black is winning, Svendsen-Tiemann, COrT. 19S7) 1 3 . . • b6 (it is all much of a muchness here: 1 3 . . . cxb4 14 'Oxd4 •

1Oc6 [ 14 . . . iOe7 I S lObS .f6 16 Ab2 ±, Kozlov-Tiemann, COrT.

·

1990] I S lIel+ IOge7 16 'Oxc6! bxc6 17 cS Ac7 I S AgS AdS 19 .e2 and Black can hardly move, Gllrd-Sakellarakis, email 1995, or 13 . . . Ad7 14 'OxcS AxcS I S bxcS 1Oc6 16 Ab2 0-0-0 1 7 'Oxd4! with advantage) 14 Ae4 ( 14 bxcS fust,

amounts to the same, 14 . . . bxcS IS .i.e4 Ab7 16 'OxcS AxcS 1 7 Axb7 .xb7 I S 1Whs+ 1 -0, Vermeulen­Taylor, COrT. 1995) 14 . . . Ab7 I S Axb7 .xb7 16 bxcS bxcS 1 7 'OxcS Axh2+ ( 1 7 . . .• c7 I S lOe6 .f7 19 .xd4 1-0, Kozlov-Elburg, COrT. 1990) IS �h2 .c7+ 19 WgI .xcS 20 lIe l + fi:Je7 2 1 .i.a3 unclear, Kozlov-Jackson, COrT. 1991 .

a2) 1 2 cS? Axh2+ 1 3 Wxh2 dxe3 14 fxe3 !? (14 Axe3 1Of6 is equal, Krustkalns-80rrmann, COrT. 1992, ... lOg4+ is a threat, and Black can use the dS-square) 14 . . . 1Of6 IS e4 Ag4 (the most popular, but others are also adequate: I S . . . O-O!? 16 eS IOg4+ 1 7 WgI .e?? [Black has at least a draw after 17 . . .• dS! I S JlxfS+ Wx fS 19 .ft +?! { 1 9 Ac2 .xd I + 20 Axd I 'OxeS 2 1 Ac2 is very drawish} 19 . . . WeS 20 Axh7 .d4+] I S Ac4+ WhS 19 AgS Jlxft+ 20 .xft .eS 2 1 lIdl /Od7 22 e6, with advantage, Elburg­Tiemann, COrT. 1990; I S . . . Ae6!? 16 Wg I lOb<i7 1 7 Af4, otherwise Black will establish a knight on eS, 1 7 . . . 0-0-0 IS Ad6, 8orrmann­Gllrd, COrT. 1991 , when Black should exploit the latent d-fiIe pin: I S . . . Ac4! with good chances) 16 .el ( 16 .c2 1Wiis+ 1 7 WgI lOb<i7 IS Af4 fi:Jes 19 Jlae I 0-0-0 :f 8orrmann-Malmstrom, COrT. 1992, 16 "'3 Ae6) 16 ... 1Obd7 1 7 eS! ( 1 7 .g3 1Whs+ I S WgI 0-0 19 .i.e3 JlaeS 20 lIf4 �S ;, Melchor­Krantz, COrT. 19S7; whenever the knight successfully reaches the im­portant eS-sguare, Black will have a plus) 1 7 . . . 'WhS+ I S Wgt .xeS ( I S . . . 'OxeS?! 19 Af5! IOfd7 20 hd7+ Wxd7 2 1 Af4 OO?! 22 .g3 favours White but IS ... /OdS 19 e6 iOes is interesting) 19 .i.f4 .xe I 20 Jlaxe l+ WdS 2 1 Ad6 and White

has enough play for the pawn, Budovskis-Svendsen. corr. 1990.

b) Meanwhile I I . . .Ae6 is interest­ing, 1 2 liX:3! ( 1 2 cxdS?! cxdS 13 liX:3 liX:6! [ 1 3 . . . lOe7 transposes to the main line] 14 lOexdS [with the knight on c6, 14 /t)bS?! is no prob­lem for Black, 1 ... . . Ab8] 14 . . . .t.xdS IS It)xdS .xdS! 16 Ag6+ bxg6 1 7 :el+ [ 1 7 .xdS Axh2+ draws] 1 7 . . .• eS 1 8 j.f4 /t)ge7?! [ 1 8 . . .• xel+ 19 .xel + Ae7 is rea­soD8ble, after all Black has got two pieces and a rook for his queen, e.g. 20 .c4 /t)f6 21 .xg6+ eMs with chances for both sides] 19 AxeS AxeS 20 h3, Svendsen-Reinke, corr. 1994/95, White is better, the black pieces lack co-ordination, and the king is not safe) 1 2 . . . lOe� ( 1 2 . . . lt)d7? 13 cxdS cxdS meets Bu­dovskis' suggestion 14 f4! with ad­vantage, 12 . . . dxc4? 13 lt)xe4! Axc4 14 Axc4 Axh2+ IS Wxh2 .xc4 16 :e I + Wf7 1 7 l%c4 with a decisive advantage, 1 2 . . . d4?! 1 3 It)c4 Ae7 14 liX:2 cS I S l%e I liX:6 16 /t)gS AxgS 1 7 AxgS �d7 1 8 b4 is not very pleasant for Black, Vitomskis­Grivainis, corr. 1997) 1 3 cxdS cxdS transposes to the main line.

c) 1 l . . ./t)f6?! is doubtful, 1 2 cxdS (not 12 liX:3?! d4 1 3 /t)f3 .i.xf3 14 Axf3 .xc4 with a reasonable posi­tion) 1 2 . . . cxdS 1 3 liX:3 - liX:6 ( 13 . . . Ae6 14 Af3 with advantaae to White, who wins the d-pawn .. in; 13 . . . 0-0 14 liX:xdS It)xdS I S IiDtdS Ae6 16 liX:3 is similarly uninspir­ing, Owens-Baier, email 1998) 14 /OCxdS It)xdS IS Ac4! (but not I S It)xdS? .xdS! [ I S . . . Ae6!? is a risky winning try, 16 liX:3 0-0-0 1 7 a3 lOes 1 8 Ae2 Ab3 19 .e I :be8 20 /t)bS and Black's open king out­weighs his active pieces, Niemand­Svendsen, corr. 1991 ] 16 Ag6+ bxg6 1 7 .xdS Axh2+ drawing)

Leonltardt 's variation 4 lik4 89

I S . . . Ae6 ( l S . . . Axh2+? 16 Wxh2 .c7+ 1 7 �gl It)xc3 leaves the black king stranded in the centre, 1 8 bxc3 Ad7 19 AgS lOe7 20 .d4 1-0, Svendsen-Crimp, corr. 1997) 16 AxdS 0-0-0 1 7 Axe6+ .xe6 18 �S h6 19 Ae3, Budovskis­MalmstrOm, corr. 1997; Black's compensation for the pawn is non-existent.

t:z � 12 cxdS cxdS 13 llX:3 ( 1 3 l%el?

0-0 14 .c2 .-..5 IS h3 /t)bc6 Ih-Jh Logunov-Sireta, corr. 1994196, but Black has a very active, and clearly advantageous, IQP position) will transpose to the main line if Black I!lays 13 . . . Ae6, and 13 . . . /t)bc6 14 lOcxdS It)xdS transposes to (c), above, as IS Ac4 wins a pawn, still, Black �ight have· more chances here than in the main line.

12 :el?! 0-0 transposes to the note to 10 l%el .

t:z.-Ae6 12 . . . d41 1 3 It)c4 Ac7, Svendsen­

Clarke, corr. 1987, 14 'llX:2 and the d-pawn is lost, as 14 . . . /t)f3 IS lOgS .f6 16 l%e I + is awful for Black.

13 udS udS 14 /t)bS lOcI Reaching the critical position,

Black has been forced into playing an ugl¥ knight retreat; how can White exploit this?

15 f4!

90 Leonhardt 's variation 4 fiJc4

As I suggested, seven years ago, this is very unpleasant for Black, and verges on a forced win. The onward march of the f-pawn will weaken Black's defence of dS, and his king position. Otherwise:

a) Black is not doing too badly af­ter I S �d6+?! which wins the bishop pair, but helps Black's devel­opment, I S . . . �d6 16 b3 ( 16 .c2 is best, and transposes to the next note) 16 . . . 1Oc6 1 7 .i.b2 o-<H>!?, Dravnieks-Destrebecq, corr. 1993.

b) I S .c2 is the best alternative, I s . . . 1Oc6 16 ltlfS ( 16 ltlxd6+!? �d6 17 .i.xh7 o-<H> IS ;el � 19 ltlfl defending h2, when Black's compensation is not enough, MalmstrOm-Clarke, corr. 1997) 16 . . . .i.bS ( l 6 . . . .i.eS!? 1 7 f4 .i.bS may be more accurate, as the weak­ness of the gl-a7 diagonal may im­prove Black's prospects, e.g. I S Itel 0-0 19 ltlfd4 .i..d7 20 .f.xh7+ �hS 2 1 g3?! gS! hitting h7, and f4) 1 7 .i.d2!? ( 17 ltlbd4?! <H> I S ltlxc6?! .i.xfS 19 .i.xfS .xfS 20 .xfS ItxfS = Diepstratcn-Grivainis, corr. 19S7; 1 7 Itel ! is recom­mended by NCO, and in BCM Nunn gave 1 7 . . . �f8 [ 17 . . . <H>?! I S ltlfd4 wins the h7-pawn] I sMcs+ �gS 19 .c3 although Black is still fighting after 19 . . . :i.d7) 1 7 . . . <H> ( 17 . . . .i.xfS?! I S .i.xfS ltlSe7 19 :Sel �S [ 19 . . . a6 20 1lld4!? .i.xh2+? {20 . . . �S} 2 1 �h2 �d4 22 .a4+ IOdc6 23 lbe7+! �e7 24 Itel+, winning, Kozlov-Pape, corr. 1991/93, {by transposition} as 24 . . . �dS 2S .i.gS+ �c7 26 lte7+ �e7?? 27 .i.f4+ forces mate] 20 h3 <H> 2 1 .i.g4 � 22 g3 .ffi 23 .i.c3 and White has an edge, Kozlov-Sireta, corr. I 994/9S) I S ltlg3 ( 1 S ltlfd4!) I S . . . g6 19 .i.c3 ltlSe7 20 :Sel a6 2 1 ltld4 ltlxd4 22 hd4 and now Black got himself

into a tangle with 22 ... ltlc6?! 23 .i.cS lieS 24 .e2 .i.eS 2S f4 .i.d4+ 26 .i.xd4 ltlxd4 27 .eS � MalmstrOm-Destrebecq, corr., but both 22 . . . ltlfS and 22 . . . .i.f4 are fme.

c) IS Itel ?! <H> 16 .c2 �S 17 �d6?! �d6 IS .c7, Riegsecker­Edwards, corr. 19S6, I S . . .• f7 is fairly level.

d) I S Illd4 .i.d7 16 ltldfS .i.eS 17 �3 .i.xfS IS .i.xfS and White has the bishop pair, Logunov-Reinke, corr. I 994/9S.

15 ••• 0-0 The choice is not a pleasant one: a) I S . . . g6? 16 fS .i.d7 17 fxg6

.g7 I S ph7 with a huge advantage Svcndscn-Melchor, corr. 19S7/SS.

b) I S . . .• d7 16 .c2 1Oc6 17 hh7 ltlb4 IS .g6+ 1-0, Svendsen-Sadeghi, corr. 1991 , al­though Black can play on here.

c) I s . . . 1Oc6? 16 fS .i.d7 17 ffi .i.e6 (this is hopeless, but so is 17 . . . g6 I S ltlxdS) I S ng7 .xg7 19 .i.fS .i.xfS 20 ltlxfS .i.cS+ 2 1 �h I .d7 22 .el+ �dS 23 .c3 and Black's position is in tatters, Budovskis-Clarke, corr. 1995.

16 15 16 .c2 is also possible, and

strong, if less to the point, 16 . . . �S ( 16 . . . h6? 1 7 fS .i.d7 IS ltlxdS wins a pawn, senechaud-Elburg, corr . 1994/9S) 1 7 ltlxd6 ltlxd6 IS fS ItcS 19 .a4 .i.f7 20 .f4 and White is clearly better, 8enechaud-De long, corr. 1994197.

16 ••• .i.d7 17 ltlsd6! Introduced by Budovskis, and

very effective. 1 7 ffi 86 ( 17 . . . gxffi? I S ltlxd6

�d6 19 lOxdS with a crushing ad­vantage Leisebein-Svendscn, corr. 19S7) I S ltlg4 ( I S .i.e2 .i.e6? [ I S . . . .i.xbS! 19 .i.xbS .i.cs 20 �hl .i.xe3 2 1 .i.xe3 ltld6 avoids the worst] 19 .i.g4! .i.xg4 20 ltlxg4

�cS+ 2 1 Wh l .d7 22 f7+! Wh8 23 b4 the opening of the a l -h8 diago­nal spelt the end of the road for Black. Svendsen-Jackson, corr. 1994/95) 1 8 . . . �cS+ 19 Wh I hg4 20 .xg4 �6 2 1 �f4 leaves Black tied up, Svendsen-Griffiths, corr. 1996. '

17 ••• �xd6 18 b3! White plans to pin the d6-knight

by .b3. Lein gives 1 8 ��2 (menac­ing Ab3xdS) 1 8 . . . Ac6 19'ffi g6? (a terrible move, 19 . . . lDd7 is clearly superior) 20 Ab3 � 2 1 ltlg4 Wh8 � Ah6 IIg8 23 .i.xc4 dxc4 24 �g7+ forcing mate.

18 •••• f6 Black has no decent move; a) 1 8 . . . .i.c6 19 .i.a3 1Id8

(19 . . . lIe8 20 ltlg4 ltle4 2 1 ffi'is not much of an improvement) 20 ffi pffi 2 1 ltlg4 1 -0, Budovskis-Sireta, corr. 1994/95.

b) 1 8 . . . ltle4? 19 ltlxdS AxfS? 20 IlxfS 1-0, Budovskis-Vitols, corr. 1994/95, and the rook is untouch­able because of the fork on e7

c) 18 . . . �6 19 .i.a3 .ffi 20 �dS .d4+ 2 1 Wh I AxfS (2 1 . . .• xdS?? 22 Ac4) 22 hfS �fS 23 AxfS IlxfS 24 .xd4 ltlfxd4 25 IlxfS+ WxfS 26 lie I Black can resign, Budovskis-Melchor, corr. 1994/95.

19 .i.a3 1If7

Leonhardt 's variation 4 1Oc4 91

19 . . .• d4 20 .0 .xd3 2 1 .i.xd6 is no help.

20 .0 .e5 11 lIadl ltlc6 II ltlxd5! ltld4

22 ... �fS 23 Ac4 Wh8 24 �7 wins the exchange.

23 .n .xd5 1-0 Kosten-Elburg, email 200I . as 14

.i.xd6 b5 (24 . . .• xd6 25 .i.c4 nets an exchange) 25 .i.f4 .i.c6 26 .i.bl wins the pinned knight.

ell 9 .. .Ad6

10 llel Alternatives: a) The combination 10 ltlexdS!?

cxdS I I ltlbS is also dangerous, ( 1 1 �dS? fails in this position: I I . . .Axh2+! [ 1 l . ...xdS?! 1 2 .i.g6+ hxg6 13 :'1+ {or 1 3 .xdS when 13 . . . .i.xh2+ draws, 14 Whl In-In, Borrmann-Crimp, corr. 1995) 1 3 . . .• eS unclear] 12 Wxh2 .xdS 13 .e2+ [if 1 3 .i.g6+?? hxg6+ is check!] 1 3 . . . ltle7 14 .i.c4 .d6+ IS gl llfS 16 .i._Ss ltlbc6 17 life I .cS 1 8 Ae3 .eS :J: Diepstraten­Downey, corr. 1992) 1 1 . . .... c7 (the anti-developing 1 1 . . . .i.fS?! is the most common, but White appears to win almost by force: 12 lIe l+ Wd8 [ 12 .. . Wd7!? 13 -*-.sS { 1 3 Ac4?! ltlf6 14 .i.gS �c6 I S 1&l4+ �b6 and the king was wandering towards safety. Borrmann-Melchor, corr. training

92 Leonhardt 's variation 4 fiJc4

game 1997} 13 . . . lOffi 14 �xffi! gxffi IS �c4 �c6 16 .0 with a strong attack, GUrd-Magee, COlT. 1999] 1 3 �gS+ IOffi 14 �c4 .d7 [ 14 . . . �e6 I S Jlxe6! .xe6 16 �xdS � 1 7 �xb7+ IObd7 1 8 .*.xffi+ .xffi 19 .*.xa8 winning easily, Melchor-Budovskis, COlT. 1997] I S .*.xdS .*.e7 16 Jlxe7! �e7 17 .d4 lle8 1 8 llel+ Wf8 19 .xffi+! The exclamation mark is for beauty! 19 . . . gxffi 20 �h6+ .g7 2 1 �xg7+ �g7 22 Jlxe8 1-0, Schlocgel­Taufratshofer, COlT. 1992. Both 1 1 . . .�b4?! 12 c3, and 1 1 . . .�cS?! 12 .*.e3 only make matters worse) 12 llel+ �d8 1 3 .*.gS+ IOffi 14 .*.c4 �e6, Kristensen-Nicholls, COlT. 1998, and now I S Jlxe6! .xe6 16 .*.xdS .eS (the threat to h2 is a use­ful resource) 17 .*.xb7+ 1Obd7 1 8 IOxc7 �c7 19 .0 .xgS 20 �xa8 Ilb8! is unclear.

b) 10 1Oe2 is quite feasible, 1O . . . lOffi (Black doesn't have to woIT}' about the e-file, here, 1O . • . &7!? I I c4 0-0 1 2 cxdS cxdS [the IQP position is the most inter­esting for Black, but 1 2 . . . lOxdS!? 1 3 IOxdS .xdS is also possible] 1 3 1Oc3 IObc6!? [an interesting pawn sac, 13 . . . .*.e6 14 lObS .*.f4 { 14 . . . 1ld8} IS g3 a6 16 1Od4 �d6?! {if 16 . . . �c7?! then 17 IOxe6 .xe6 1 8 .*.xh7+ wins a pawn, but 16 . . . .*.eS is fine} 1 7 1Oxe6 .xe6 1 8 .*.c2 /Od7 19 IOxdS! IOxdS 20 �b3 107ffi 2 1 .*.gS with a small advan­tage, Kozlov-Morgado, COlT. 1977n9] 14 IOexdS IOxdS I S �c4 .*.e6 16 .*.xdS :ad8 17 �xe6 .xe6 1 8 .e2 .xe2 19 lOxe2 �cS and Black has the initiative, Kozlov­Krantz, COlT. 1994/95) I I c4 (White can also occupy the fS-square, I I IOfS .*.xh2+?? [but not this! 1 l .. ..*.xfS?! 12 .*.xfS 0-0 13 IOd4 lle8 14 1lel .*.eS I S �e6! Jlxe6 16

1Oxe6 .xe6 17 f4 wins the ex­change, although Black will have some compensation, but 1 l . . .�eS is best] 12 � �S+ 13 �gl �xfS, Vyskocil-Van Willigen, COlT. 1997, and now White could have won a piece on the spot by 14 .el with the threat oflOg3+) 1 1 . ..0-0 12 cxdS cxdS 1 3 IOd4 (13 �3 �e6 14 lObs can now be met by 14 . . . �cS, as c7 is defended by the queen) 13 . . . 1Oc6 14 IOefS �eS IS .*.e3 and the position is equal, Kozlov-Niemand, COlT. 1994/96.

lo ... 1Oe7 Now White has a choice between

two dangerous sacrifices:

Clll I l lOexd5! C111 1 I 1Oc4!?

If White plays another eleventh move Black will be better.

Clll I l lOexd5!

(! NCO) White's attention has re­cendy focused on this more direct way of assaulting the solid black centre:

1 I ••• cxd5 11 lObS 12 IOxdS?? is a blunder,

12 . . .• xdS 0-1 , Skrastins-Clarke, COlT. 1990, as 1 3 �g6+ �d8 leaves the queen defended.

12 ••• 0-0

With: a) 12 .. . .i.f4?! is the only playable

alternative, but it is very difficult for Black. 1 3 .i.xf4 ( 1 3 g3? is worse: 13 . . . a6 14 .i.xf4 axbS I S .i.xbS+ �bc6 16 .i.d6 .i.e6 1 7 .e2 Wd7 ;) 13 . . .• xf4 14 g3 ( 14 -..S+?! WdS I S :Xe7 Wxe7 16 1lel+ �S 1 7 g3 .a; I S .xdS+ .i.d7 19 lLld4 WcS 20 � lleS 2 1 .i.fS .i.xe6 22 .i.xe6t Wc7 23 .as+ b6 24 .dS, Salaske-Schreyer, East Gennany 19S3, when 24 . . . IldS is strong. as 25 .uS? �6 traps the queen) I •. . .• a; I S �7+ WdS 16 �aS b6 (l6 . . . i0bc6?! 1 7 .i.e4 .i.d7 I S .i.xdS �dS 19 .xdS WcS 20 :ad I IldS 2 1 .d6 .xd6 22 :Xd6 looks pretty miserable for Black. by the time he captures the knight, assuming he manages to, White will have gained a decisive advantage on the king­side, Malmstrom-Kozlov, corr. 19S5) and now:

a l ) 1 7 c4! .i.b7 I S �b6 axb6 19 adS ("White has three pawns to add to his rook, against two black pieccs"-Biicker.) 19 .. . 1lf8!? (Black cannot afford to go pawn hunting: 19 . . .• xb2?! 20 lle2 .a3 2 1 .i.bS .cS 22 .i.c6 .i.cs 23 Ilcl .d6 24 eb3, the alternative is 19 . . . �7!?, 20 .i.e4 Ilf8 [20 . . .• xb2?! 2 1 d6 ItX6 22 1lbl .a3 23 Ilel is strong] 2 1 .d2 but with the queens on the

Leonhardt 's variation 4 flx:4 93

board, the black king is a defmite liability) 20 .d2 .d4!? (Black de­cides that exchanging queens is his best bet, as then his king may turn out to be strong in the centre) 2 1 :ad I �7 22 .i.e4 (personally, I would �refer 22 .c2) 22 . . .• xd2 23 1lxd2 �fS 24 .i.g2 �6 (Black wants to find some dark squares for his knights, but White is on the ball) 25 1le6! Ila; 26 llde2 �S 27 :xa; pa; 2S b4 IOa4 29 Ile6 IOes 30 a3 bS 3 1 1le4! and White aims at the h7-pawn, Tiemann-Kosten, email 2001 . Black's position is difficult.

a2) 1 7 .e2 (+/- NCO, although it doesn't say why-a computer assessment, perhaps?) 1 7 . . . �6 ("with a little advantage to Black" (Budovskis) 1 7 . . . .i.b7?? I S �7) I S c4!? (SuJ8csted by Melchor, I S b4 lleS 19 We3 .i.b7 20 �b6 �4 2 1 f4 .xb6-MalmstrOm) I S . . . lLld4! ( I S . . . lleS?! 19 cxdS? [ 19 f4! .i.b7 20 cxdS avoids problems on 13] 19 . . . �4 20 "'5 �f3+ 21 Wg2 g6! 22 .xf3 [22 .xh7 �gS wins] 22 . . . .i.h3+ 23 Wxh3 .xf3 Black is better; I S . . . .i.b7?! 19 cxdS �dS [after 19 . . . lLld4 20 .xe7+ .xe7 2 1 llxe7 Wxe7 22 �7, Black will re­capture the d-pawn, but remains worse] 20 .i.e4 �e7 2 1 �b6 axb6 22 :adl qd White will probably regain the two pieces for his rook due to the pin) 19 .xe7+ .xe7 20 llxe7 Wxe7 2 1 �7 dxc4 22 .i.xc4 Black appears to have quite enough play for the pawn after 22 .. . IldS 23 �S+ Wf8! (23 . . . Wd6 24 1Oe3, and on 24 . . . .i.b7 25 Ildl ) 24 lldl ltlf3+, the weakness of the white kingside light squares is critical for Black's counterplay, 25 Wg2 .i.&4 26 Ile I IOes 27 1Oe3 (27 1lc3 .i.fS) 27 .. . .i.f3+ 2S Wgl lld2 29 b4 gS 30 a4 .i.e4 3 1 h3 1013+ 32 Wg2 hS and White's extra pawn is meaningless.

94 Leonhardt 's variation 4 1Oc4

a3) 1 7 lhe7?! .xe7 1 8 .i.c4 .i.b7 19 .i.xdS �c8 20 .g4+ llXl7 saves the knight from immediate capture, but does not resolve the problem of how to extricate it, 2 1 .fS (2 1 .f4 lDcs 22 lIdl .i.xdS?! [22 . . . lId8! 23 .i.xb7+ �b7 24 lhd8 .xd8 25 .f3+ �b8 is more straightforward) 23 IIxdS �b7 24 .f3 .e4 25 .xe4 lbxe4 26 IId7+ �a8 27 lhg7 and White has slightly better chances with his three pawns and active rook. for the knight, Van Gameren-MalmstrOm, corr. 1999) 2 1 . . .IIfS 22 .i.xb7+ �b7 23 .dS+ �b8 24 a4 .cS about equal, MalmstrOm-Svendsen, corr. 1997.

b) 12 ... �d7?? was not one of my better ideas from the first volume!

1 3 .g4+ �c6 14 .a4 IIfS. Unfortunately, IS .i.gS! wins: 1 5 . . .• xfl+ ( I S . . . .i.cS 16 .i.xe7 .xfl+ [if 16 . . . .i.xe7 1 7 lOxa7+ �c7 1 8 lbbS+ �d8 19 .xa8 is pos­sible, as the back rank is defended) 17 �hl .i.h3 1 8 lDc3+ [ 1 8 gxh3?? .f3 mate) 1 8 . . . �c7 19 lOxdS+ �c8 20 .e4 wins; I S . . . .i.xh2+ 16 �h2 �S+ 1 7 �gl .xgS 1 8 lOxa7+ �c7 19 lbbS+ �b6? [Black prefers to get mated rather than lose a rook by 19 . . . �d8 20 .xa8) 20 .d4+ �c6 2 1 .c3+ �b6 22 .c7+ �a6 23 lbd6+ �a7 24 .cS+ b6 25 .c7+ .i.b7 26 .xb7 mate,

Strautins-Svendsen, corr. 1998) 16 �h I �d7 ( l6 . . . .i.d7 17 .i.e3 .n 1 8 lOxa7+ �c7 19 .as+ and mate) 17 lOxd6+ 1Oec6 18 lbbS hS 19 .a3 IIffi 20 IIn 1-0, Strautini-Melchor, corr. I998 .

c) 12 . . . .i.cS? 1 3 .i.e3 +/-d) 12 . . .• ffi? 13 "'5+ g6 14 .i.gS

( 14 Wg5 is also convincing) 14 . . .• xf2+ 1 5 �f2 ()-()+ 16 .f3 .i.cS+ 1 7 .i.e3 .i.xe3+ 18 IIxe3 lhf3+ 19 �f3, White has an extra exchange, MalmstrOm-Niemand, corr. 1994195.

I3 lbsd6 .dl+ 14 �bl

14 •• ..*.14 The only reasonable try, as after

14 . . . �?!, Lambers-Kamitzky, corr. 1987, I S g3 .ffi White wins as per the next game, 14 . . .• ffi1! IS lOxc8 11xc8 16 Wg4 lbbc6 17 .i.gS Wd6 1 8 1Ie§ Wc7 19 J:lh6 ( l 9 lhe7 ' lOxe7 20 .e6+ is also convincing) 19 . . . lbg6 20 .i.xg6 bxg6 2 1 .e6+ 1-0, MalmstrOm-Melchor, corr. 1997.

IS Wdl! I S .i.e3?! .i.xdl 16 .i.xfl .i.xc2

17 .i.xc2 lhfl 18 lhe7 ( 1 8 .i.b3? lbbc6 ; Elburg-Heap, corr. 1989) 1 8 . . . lDc6 19 .i.xh7+ Wxh7 20 lhb7, Elburg-Melchor, corr. 1988, when by 20 . . . J:lafl followed by the dou­bling of rooks on the seventh, Black would draw comfortably.

IS ••• � The endgame thal follows

IS . . .• xd2 16 .i.xd2 lObc6 1 7 lOxb7! favours While: 1 7 . . . :abS IS iOcS Ilxb2 19 lOb3 (the rook is trapped!) 19 .. . .i.1'5 20 .i.c1 (or 20 .i.c3) 20 .. . llxc2 2 1 .i.xc2 .i.xc2 22 .i.a3 :eS 23 :acl , Black's compen­sation for the exchange is , insuffi- . cienl, Straulins-GrivainislHayward, corr. I999.

16 b4! Straulins' idea. While prepares to

bring his bishop 10 Ihe a l -hS diago­nal, whilsl al the same lime b4-bS can be a useful resource.

16 ltitb1?! is loo lime-consuming, 16 . . . lObc6 17 b3 (others: 17 .c3 d4 I S .c4+ �hS 19 .i.d2 J:lf2 20 lCld6 :afS Strautins-the while king is short of defenders; 1 7 lOcs J:lf6! I S c3 :afS [ IS . . . J:lh6!] 19 _e3 [ 1 9 �gl?? 112-+ MalmstrOm] 19 . . . lOf1 20 hf5 1lx1'5 2 1 .e6+ �hS 22 .i.e3 .xel+! 23 Ilxel J:lfl+ 24 .i.gl he6 25 1lxe6 J:llf6, unclear, Mel­chor) 1 7 . . . :abS (rather than 1 7 .. . J:lf6?! I S .i.a3 J:lh6 19 .i.d6 covering b2) I S lOcS, Evans­Stummer, corr. 1994195, and now Straulins suggests IS . . . lCld4! 19 �g I J:lb6, swinging the rook over 10 the kingside, with a powerful attack. I S lOas is similar: I S . . . lCld4! 19 �gl J:lb6 20 .i.b2 J:lh6 2 1 .xh6!

Leonhardt 's variation 4 fOc4 95

lOf3+ 22 gxf3 gxh6 23 fxg4 .xg4+ Ih-Ih Chorfi-Simmelink, corr. I99S.

16 ••• lObc6 16 . . .• f6? 17 .g�! . • ul IS

.xg4 lObc6 ( IS . . .• f6 19 .i.gS

.xd6 20 .i.xe7 J:leS [20 . . .• d7 2 1 . �] 2 1 .i.xh7+ wins) 19 .e6+ �hS 20 .i.gS .xa2 2 1 lOf7+ �gS 22 J:lfl lOd4 23 .i.xh7+ �h7 24 �3+ �g6 25 lOes+ �gS 26 .g4+ Wh6 27 � male, Strautins-Clarke, corr. 1997.

17 .i.bl J:l.d8 1 7 . . . J:lt2?! I S .c3 .gS 19 J:lfl

:afS 20 Ilxt2 Ilxt2 2 1 .e I ! .f4 22 bS .i.h3 23 .i.c I .16 24 bxc6 .xal 25 cxb7 1Oc6 26 gxh3 1-0, Strautins-Destrebecq, corr. 1997.

18 .cl _15 19 lOxb7 J:lbS

lO bS! Reaching the crilical posilion for

this varialio... Ruggerl Laderchi­Radovic, corr. 2000. Black has drawing chances after 20 lOas lOxb4 2 1 .xg7+ .xg7 22 .i.xg7 J:lt7 23 .i.c3 lOxd3 24 cxd3 .i.1'5 due to the <>ppOsiae-colour bishops.

lO •• Jbb7 11 bu6 lOxc6 11 .i.a3! 22 J:lfl is also annoying. 22 .. . J:lf6!

is the besl chance, (22 .. . llxfl +? 23 Ilxfl is losing, for instance 23 . . . lOb4 24 Aa3 as 25 .cS) 23 IlxI6 .xf6 24 .xf6 gxf6 25 j,xf6 lOb4 26 .i.c3 (26 J:lb I .i.cs 27 J:le I lOxd3 2S cxd3 .i.1'5) 26 . . . lOxd3 27 cxd3 .i.f5 2S d4 J:lb 1+, drawing.

96 Leonhardt 's variation 4 lDc4

11 ••• :C8 13 .i..6 ltlbB 14 .xc8+ .i.xc8 lS .i.xb7 .i.xb7 16 1te8+ Wf7 17 JbbB

27 Ae7+ Wg6 28 Axb7 IOc6 29 .i.f8 .f5 30 Lg7+ Wf6 also fa­vours White.

17 •• ..i.c6 27 . . . L6? 28 :el .d2 29 :ts+

Wg6 30 Ae6+ WhS 3 1 h3 threatens mate.

18 Ac8 .i.bS 19 lIel 29 Ac7+ Wg6 30 .i.b2 .d2. 19 •.•• dl 30 b+ Wg6 31 JIe6+

WbS 31 b3 Again Black is in serious trouble.

Black desperately needs an im­provement here.

Cl11 n ltlc4!?

This guarantees some advantage, although Black may be able to pick his way through the miMfteld, and draw.

1 1 ••• dxc4 There is no reason to decline this

piece and, anyway, neither 1 l . . ..i.c7? 12 .i.gS .i.e6, Tiemann­Criel, corr. 1984, when 13 .e2 h6 14 .xe6 .xe6 IS Axe6 bxgS 16 Aae I gives an advantage, nor 1 1 . . . .i.cS? 12 .i.e3 .i.b4 13 ltleS! with advantage, Kozlov-Nyman, corr. 1990192, are worth the bother.

11 .i.xc4 .p!?

After this, the best Black can hope for, given correct play, is a drawable ending, a pawn down.

12 ... .i.xh2+! leads to a safer equality. 1 3 Wxh2 .xc4 14 .i.gS 0-0

IS he7 ( I S Axe7?! is inferior, I S . . . h6 16 Ac4 .f7 1 7 .i.h4 .i.f5 18 Ae3 ltla6 and I prefer Black, loguJlov-Heap, corr. 1994/94) IS . . .• f4+! ( 1 S . . . Ae8? 16 ltlc4! .i.e6 17 .i.gS .i.f7 18 ltld6 Axe I 19 .xe I with advantage, Kozlov-Littleboy, corr. 1991 ) 16 Whl Ae8 17 L3 ( 17 �S! looks stronBer, 17 . . .• f7 [ l 7 . . . .i.d7! 1 8 :ad1 .f5 is safer] 18 .xf7+ Wxf7 19 .i.a3 and Black suf­fers from his lack of queenside de­velopment, Budovskis-Elburg, corr. I 994I9S) 17 . . . .i.e6! (must be the best move, 1 7 ...• f7?! 1 8 .0 [ 1 8 ltlc4 is also dangerous, 18 . . . b6 19 ltld6 Axd + 20 .xel �S+ 2 1 WgI , Black has problems with his undeveloped queenside, Budovskis­Kinnes, corr. 1980] 1 8 . . . .i.e6 19 ltlc4! [ 19 .xf7+ .i.xf7 20 Axe8+ .i.xe8 2 1 ltlc4 .i.g6 is equal, Jackson-Malmstr'Otn, corr. 1990] 19 .. .• xO 20 gxO .i.f7 [20 . . . .i.dS?? 2 1 ltlffi+] 2 1 ltld6 with advantage Budovskis-Krol!Jl'8f, corr. 1992; 17 . . . Axel + 18 .xel � 19 WgI .i.e6 20 Adl � Elburg-Nyffeler, corr. 1989) 1 8 ltlc4 (not 1 8 Axe6?

�4+; 18 WgI [to �rotect the f­�wn] 1 8 . . . 1086 19 :e3 .i.g4 20 Wd2 lhe3 2 1 fxe3 .eS, play is level, Adamczyk-Bonmann, COlT. 1997; 1 8 :c4 �6+ [ 1 8 . . .• xf2?? 19 .d6 wins the bishop, but IS . . .• f7 is strongest, keeping the white king on a worse square] 19 WgI � 20 .d4 .g6 [20 . . . :cdS!? 2 1 .eS :es 22 :ae I .i.f7 23 .xeS+ heS 24 IlxeS+ lhe8 2S lhe8+ � 26 1le7+ WgS 2\:eS+ drawing, Budovskis-Krantz, COlT. I 994I9S] 2 1 Ilael l£x.7 22 .i.d6 IC.)dS equalising completely, dUdovskis-Kozlov, COlT. I 994/9S) IS . . . 1Oa6 ( 18 ... �7 19 �6 1lf8 20 lhe6 �+ 21 WgI .xf2+ draw­ing, Viola-Downey, COlT. (999) 19 �6 1lad8?! (19 . . . 1lf8! is stronger, as 20 Ilxe6? �+ 2 1 WgI .xf2+ 22 Whl Ilf4 23 1le4 lhe4 24 lOxe4 �+ 2S WgI .xe4 wins a pawn) 20 lOxeS! (but not 20 .cl? .xf2 2I lOxeS Ild4!) 20 . . . llxdl 2 1 1laxdl . .i.ds 22 �6, Herbst-Magee, COlT. 1999, and now 22 . . . h6 23 c4 .i.xc4 24 :c4 .gS menaces .. . �S+, and keeps plar level.

Not 1 2 . . .• xc4?! 1 3 .xd6 .n 14 .i.gS 0-0 IS .i.xe7 .xf2+ 16 Wh l La 1 7 .i.h4 with advantage, Budovskis-Morgado, COlT. 1977179.

Leonhardt 's variation 4 li:k4 97

This leads to a strong attack. 1 3 .i.e6 poses less problems,

13 . . . .i.xh2+ ( 13 . . . .i.xe6? is clearly worse, 14 .xd6 Wf7 IS � h6 16 l£x.s :es, Svendsen-Clarlte, COlT. 19S9, when 17 lOxe6 �S IS �8+ gives him a good pawn more. 13 . . . .i.cS!? is an alternative, though, 14 .i.xcs .i.xf2+ IS Wxf2 0-0+ 16 WgI ltlxcs with a minimal white advantage.) 14 Wxh2 .i.xe6 IS .d6 0-0 ( IS . . . oo!? 16 .xe6+ [16 lhe6+?? Wf7 1 7 .c7+! Wxe6 IS .i.gS is unclear, but not 1 7 lhg6?? hxg6+] 16 . . .• xe6 1 7 lhe6+ Wt7 is completely equal.) 16 .xe6+ (16 .xe7 .i.d7 17 � .i.f5!? =; 16 lhe6 �+ 1 7 WgI ltlf5 I S .f4 .f7 19 :c4 Ih-Ih Grobe-Kozlov, COlT. (994196) 16 . . .• xe6 17 lhe6 ltlf5 (In this endgame White's bishop should allow him a very slight edge) I S .i.f4 (there is no lack of alternatives: I S .i.gS �?! [Black should take more care, IS . . . h6 19 .i.f4 �7 is level.] 19 Ile4 cS [ 19 . . . lOxc2? 20 Ilcl and the knight is lost] 20 .i.e7 IleS 2 1 �S and White is getting on top, Herbst-Maltez, COlT. 1999; I S 1le4 1086 19 b3 ltlcs 20 Ilc4 ltle6 21 ll)e4 IladS 22 g4?! Too aggressive, 22 . . . ltlfd4 23 .i.e3 ltlf3+ 24 Wh3 ll)es and the weaqess on f3 allows Black the better prospects, Kozlov-Downey, COlT. 1994/97; I S ll)e2 1086 with a level endgame, Dravieks-Downey, COlT. (992) IS ... 1Oa6 ( 1 S . . . �7) 19 Ildl lIMS 20 lhdS llxdS 2 1 AeS � 22 � and White has a tiny plus, Melchor­Malmstrom, COlT. 1994I9S.

13 •• ..i.c7 14 .i.g5 14 ltlgS?! is not so worrying,

14 . . . 1lf8 IS .i.d3 ( I S ltle6? .i.xe6 16 lhe6 Ilffi 17 .i.d3 .f7 led to an easy win for Black in Diepstraten­Rittenhouse, COlT. 1989; White has

98 Leonhardt 's variation 4 lDc4

almost no compensation for the �iece) IS . . . .i.f5?! ( I S . . . �6 16 g3 .f6 is stronger) and now, instead of 16 .e2? .i.d6 17 .i.xf5 Lf5, Black is consolidating, Lo�unov-Downey, corr. I 994I9S, 16 lOe6! 1Oa6 17 �xf8 WxfS IS .13 is far less clear.

14 ••• .i.f5! 15 .i.u7 I S .13!? �1? ( l S . . . .i.xc4! 16

Lc4 .xgS 17 .f7+ WdS I S Adl+ �7 19 .i.e6 and now Black has the resource 19 . . . -*.d6 because of the weak white back rank) 16 .i.xe7 Wxe7 17 �6+ 1-0, Krustkains­Bomer, corr. 1997.

15 ••• .i.u4

16 -*.g5! This move is very dangerous for

Black. 16 13 may lead to no more than a

draw 16 . . . Wxe7 17 J:lxc4+ WfS I S .e2 ( I S .d2!? .d6 19 .gS! �7 [ 19 . . .• xh2+?? 20 Wfl leaves Black without defence to the white mating threats] 20 llae I with a strong attack, Senecbaad-GUrd, corr. I 994/9S, although Black may be able to defend with 20 .. . bS 2 1 lle7 .d4+ 22 Whl bxc4) IS . . . �7 19 llel �f6 20 lle7 .i.b6+ 2 1 Whl �S 22 -*.xdS cxdS 23 lle6 �S 24 g4 .f7, De long-Downey, corr. 1994/9S, and White should take the draw here, by 2S lle7 .g6 26 lle6 .17 etc.

16 ••• �7 This is Black's only move. a) 16 . . .• xgS? 17 Lc4+ WfS IS

.0+ .f6 19 llael ! (mate on eS will be a recurrinB theme) 19 . . . �7 ( 19 . . .• x13?? 20 lleS mate) 20 .e3 .i.eS 2 1 f4 �b6 22 LeS �c4 23 .cS+ �6 24 lle6 and Black is crushed, Budovskis-Niemand, corr. I 994/9S.

b) 16 ... h6? 17 .g4 bxgS ( 1 7 . . . .i.xh2+ IS Wxh2 bxgS+ 19 Wgl WdS 20 Lc4 is also hopeless) IS Lc4+ 1-0, Pape-Logunov, corr. 1994I9S, as I S . . . WdS 19 Adl+ .i.d6 20 Ld6+ .xd6 2 1 Ad4 wins the queen plus some �wns.

c) 16 . . . Wf8? 17 .g4 .i.f5 IS .e2 .i.d1? ( I S . . . .i.e6 fends oft' the mate, but 19 .xe6 .xe6 20 Le6 is not very encouraging for Black) 19 .e7 mate, Budovskis-Clarke, corr. I 994/9S.

17 .g4! Others: a) 1 7 .i.f7+? .xf7 IS Lc4+

lLleS! 19 .i.f4 0-0 ; b) 1 7 .f3!? lDcs I S :adl? ( I S

b4! .i.eS!? 19 bxcS!? [ 19 .g4] 19 . . . .i.x13 20 LeS+ WfS 2 1 .i.e7+ drawing) IS . . . llf8 19 .e3 llf5 20 h4 J.f4! and the white attack soon ran out of ammunition, De long-Kozlov, corr. 1994/9S. \. . c) 1 7 .d4!? �f6 IS llx-"+! �e4 ( IS . . .• xc4?! 19 .i.xf6 .xd4

20 hd4 0-0-0 2 1 �xg7 is more pleasant for White, although Black has drawing chances) 19 lIel .xgS 20 llxc4+ � 2 1 .d7 .cl+ 22 �fl �b6 23 .e7+ (or 23 g3 !? when ' 23 . . . hS? 24 �g2! wins, so Black must try 23 . . . h6) 23 . . . �gS 24 .e6+ � 25 g3 hS 26 lieS (26 IIf4+ .xf4 27 gxf4 IIh6 2S .CS+ as per the Jlrevious note) 26 . . . lIh6 27 IICS+ 1If6 2S IIxf6+ gxf6 29 .xf6+ �eS Ih- 1h, Kozlov-Heap, corr. 1994197.

17 ••• � . 1 1 7 . . . �f6? I S .e6+ �f8 19 l1xc4

:es 20 .xeS+ .xeS (20 . . . �eS?? 2 1 �e7 mate) 2 1 llxeS+ �eS 22 lIel+ �dS and with the bishop pair, and an extra pawn, White is winning, Budovskis-Malmstrom, corr. 1994195.

18 b4!

18 ••• hS! I S ... �xh2+? 19 � ! (not 19

�h2 h6 [to capture the bishop with discovered check] 20 � bS! [20 . . .• xgS? 2 1 bxcS wins] 2 1 bxcS bxc4 22 f3 0-0 and the king es­C:!IPCS!) 19 . . . h6 20 bxcS hxgS 2 1 .xc4+ (2 1 llxc4+ �f8 22 lib I 1Ih4 23 .f3+ .f6 24 llxb7 is also strong) 2 1 . . .• xc4 22 llxc4+ �f8 23 lld I ! (White threatens both g3,

Leonhardt 's variation 4 liJc4 99

trapping the bishop, and IId7; the immediate 23 g3? allows 23 ... �xg3 24 £Xg3?? llhl+) 23 . . . llh4 24 llxh4 gxh4 25 g3! Anyway! 2S . . . hxg3 26 �g2 gxf2 27 chh2 and White will win, Budovskis-Downey, corr. 1994/97.

19 � 19 �3!? also promises some

advantage, 19 . . .. xgS 20 bxcS .eS 2 1 �d3 0-0 22 �xc4 �hS and Black has some problems with his king, but the opposite-colour bish­ops, and weakness of f2, make a draw possible, Budovskis-Strautins, corr. I99S.

19 ••• �d8! 19 . . . �d7? 20 lladl+ �cS 2 1 bxcS

lieS 22 �e2 llhS 23 �g4+ 1 -0, Budovskis-Heap, corr. 1994/96.

10 buS �KR5 11 .Ke4+ 2 I llxc4+? Wf8. 11 •••• Ke4 11 IIKe4+ �d7 White has an extra pawn, and the

more active position, but the opposite-colour bishops, and weak cS-pawn, make a win problematic.

13 b4 23 f4 �f6 24 llael Ih-Jh Sireta­

Doplmayr, corr. 1995, as 24 .. . llaeS 25 �e6+ �c7 .26 �f2 �e7.

13 ... �f6 14 IIbl �c7 lS �f7 llbd8 16 �KbS llb8 17 �g4 IIKb4

White has a slight plus, Strautins­Grivainis, corr. I �7.

Summing up: As about the best Black can hope for after 9 . . . �cS is a position with a pawn less, and 9 . . . �d6 is also difficult, unless Black finds improvements here, I think that 9 . . . �e6, as well as the line with S . . . �b4!? and 9 . . . � (or the other way round), are worth investigation.

6 3 ... li)c6 and other third move alternatives for Black

I e4 e5 1 �fJ f5 3 �](e5

Many Latvian players are not happy to accept the small positional disadvantage, or potential endgame, sometimes entailed by playing 3 . . .• a; (considered in the first five chapters of this book) and prefer to continue in a real gambit spirit. Black has several riskier third move options that appeal to these free spirits, of which the strongest, and most popular, is 3 . . . �6. Although many of the positions that arise are objectively better for White, they tend to be messy and not without practical prospects for the tactically minded second player, especially considering their undoubted surprise value in over-the-board games, and with the clock ticking away.

A 3 ••• /l)c6 B 3 .•.• e7 C 3 ••• d6 D 3 ••• �f6

A 3 ••• lOc6!? Introduced in 1 873 by both the

Scot G.B.Fraser and the Dane H.Moller. White races an important choice, and we further sub-divide:

AI 4 �u6 Al 4 -...s+ A3 4 d4 White has one other, rare, move in

4 lOc:4? fxc4 5 d4 exd3 6 -*.xd3 � 7 0-0 dS 8 Ilel+ -*.e7 9 �3 0-0 10 b3? -*.b4! I I .i.d2 -*.d6 12 c3 �5, when the white kingside beg!ns to look a bit exposed: 13 -*.n?! �fg4! 14 0 -eM 1 5 h3 lOxo+ 16 gxO .f2+ 0-1 , Bosbach-Hector, Dinard 1985.

AI 4 �u6 White damages the black pawn

structure. 4 ••• du6

So, Black is a pawn down\ what compensation does he have for it?

3 . . . �6 and other third move alternatives/or Black /01

Well, firstly, he will be able to develop quickly and easily, and sec­ondly, he will have pressure down the open central files, especially af­ter he castles queenside. White has a wide variety of potential fifth moves as Black has no threats as yet-he cannot even take the e4 pawn.

s lOc3 There are many alternatives: a) S d4 and, a l ) S.t.eb4, as White po longer

possesses his king's knigHt there is no reason why Black should not .,..ail himself of this aggressive post. Meanwhile, e4 is attacked. 6 eS (this leaves the white centre rather static, 6 exfS!? transposes to (e), below) 6 .. . .i.e6 (6 . . .• e4+?! ignores the first rule of development, 7 .i.e2!? [7 .i.e3 f4 8 ltld2 .dS 9 .i.xf4 .xd4 10 .i.gl .xb2 1 1 .i.d3 also favours White, Wehr-Schinner, corr. 1989] 7 . . .• xg2 8 .i.o �3 9 .i.gS .i.e7 [9 . . . .i.d7!?] 10 .d3 [menace: 1 1 .i.xc6+] 10 . . . f4 1 1 .i.xf4 .i.f5 12 .i.e4 .xd3 1 3 cxd3 and White enjoys a two pawn advantage, Lincistrom-Holm, corr. 19 10) 7 c3 (White 'adds protection to his d­pawn, 7 lOc3 .i.b4 8 .i.d3? .xd4 9 0-0 0-0-0 ;: Bengtsson-Hector, Swedish ch 1985; not 7 g3'!? .e4+; 7 .i.e3?!, Massong-Hayward, corr. 1986, allows 7 .. . f4 8 .i.d2 [8 g3?! fxgl 9 fxgl'!? .e4] 8 . . . 0-0-0 9 c3 gS and Black has good prospects) 7 . . . 0-0-0 8 ltld2! (8 .i.d3 cS 9 .i.e3 c4 10 .i.c2 gS? [ 1 0 . . . f4 I 1 gl !] I I 1Od2? [missing 1 1 g3 � 1 2 .d2 .i.e7 1 3 h4, winning a further pawn] I I . . .f4 1 2 lDO �S 1 3 .i.d2 g4 with a powerful initiative, Ruben­Sorensen, Copenhagen 1 876; 8 .i.e2 cS [there is also the hyper­aggressive 8 . . . f4?! 9 ltld2 gS 10 100 �6 1 1 h4 .i.e7, Bemard­Hector, Montpellier 1985, when 12

g3 is critical] 9 .a4 [9 g3?! .e4 10 o .c6 I 1 O-O? cxd4 12 cxd4 lhd4 + Gilnicki-Monneal, corr. 19S3; 9 .i.e3?! f4 10 gl fxg3 1 1 fxgl .e4 12 �f2 still favours White] 9 .. . �b8 10 .i.0 cxd4 1 1 �S?! [ 1 1 .i.e3! kccps the d-pawn] 1 1 . . . .i.c8 1 2 0-0 d3 13 .i.e3 a6 14 �3 lDh6 IS g3

. .e7 16 .i.xh6 gxh6 17 ltld2? .xeS +, Hingst-Stummer, corr. 1991) 8 . . . cS would then meet 9 lDO (9 .c2 h6 10 b3 li:k7 1 1 .i.b2 1Dc6 12 .i.c4 .i.xc4 1 3 .xfS+ �b8 14 bxc4 cxd4 with counterplay, Muller­Sinke, corr. 1982; 9 g3 is also good, 9 . . .• e7 10 lDO with the threat of .i.gS, 10 . . . h6 1 1 .i.g2 gS 12 0-0, Von Otte - Drill, Hessen 1991 , and White has almost consolidated) 9 .. .• e4+ 10 .e2 h6 1 I .i.e3 gS 1 2 dxcS! to use the d4-square, 1 2 . . . li:k7 1 3 lDd4 with advantage, Kastner­Drill, corr. 1989.

al) In the game Punt-Sinke, corr. 1986, S . . . cS !? 6 dS lDffi 7 .i.bS+ .i.d7 8 .e2 fxe4 9 1Dc3 .i.e7 10 .i.gS 0-0 worked out OK .

a3) but S . . . lDffi?! is wrong, 6 eS li:k4 (on this square Black will have to beware of the move 0, while on 6 . . . lDds White will hit Black with a later c4, 7 .i.e2 .i.e6 8 0-0 .d7 9 c4 lOb(; 10 b3 . White has managed to set-up a large centre, and the black knight is mispla�, De Glena­Smits, Hengelo 1995) 7 .i.e3 .i.e6 (7 .. . cS?! is not recommended, 8 gl?! [8 dS!] 8 . . . cxd4 9 .xd4 .i.d7 10 .i.c4, Duras-Neumann, Hilver­sum 1903, 1O . . . .i.c6 I 1 .xd8+ lhd8 1 2 0 lDcs and Black does not have much for his pawn) 8 g3 (8 O! is obviously critical, 8 . . . lDgS 9 .i.d3 with a great �) 8 . . . .i.dS 9 0 lDgS 10 .i.g2 (this is a good blockading square for a knight) 1 1 1Dc3 ( 1 1 .d3!?) 1 1 . . . .i.b4 1 2 0-0 .i.xc3 13 bxc3 f4!? and Black has

102 3 . . . llX6 and other third move alternatives for Black

kingside play, Bering-Henrichsen, Copenhagen 1993.

a4) S . . .• ffi!? This is important as it may transpose from what could become Black's best line against 4 d4 (4 . . .• ffi) 6 eS .f7 (6 . . .• g6 is no better, 7 .0 [7 liX:3!? .i.e6 {7 .. . .i.b4!?} 8 lDe2-f4 exploits the black queen' s position] 7 . . . .i.e6 8 .i.d3 lDe7 9 c4?! [9 liX:3] 9 . . . 0-0-0 10 liX:3?! [the d-pawn is a problem even after the superior 10 .i.e3 .f7 I l lDdl h6, intending .. . gS and . . . f4] 1O . . . :xd4 1 1 lik2, Kyroelae­Provoost, Finland-Holland corr. 1980, and now 1 1 . . . :d8 is fine) 7 f4 (this move has the merit of control­ling gS. 7 lDdl [the knight goes to o to add defence to the d-pawn] 7 . . . .i.e6 8 lLlO h6 9 .i.d3 0-0-0 10 0-0 gS and Black has a good posi­tion and can combine the kingside pawn advance with pressure on d4, Downey-Zemitis, corr. 1997) 7 . . . .i.e6 8 .i.e3 0-0-0 9 c3 gS!? was interesting, Nyman-Strautins, corr. 1990, although the straightforward 9 . . . lLle7 10 .i.d3 lOds is a good alternative.

b) S eS?! .i.e6 (the game Nicolaiczuk-From, Copenhagen 198 1 . took a different path here: S . . .• d4!? 6 .e2 [6 .i.e2?! .xeS 7 0-0 .i.e6?, Lauritsen-Henrichsen, Copenhagen 1993, placing too many pieces on the file! 8 :el and White wins material, 7 . . . .i.d6 8 g3 .i.d7 is perfectly good] 6 . . . .i.e6 7 liX:3 [7 c3 ma� stronger, 7 . . .• d7 8 d4 cS 9 .i.e3 with typical play, Van't Land-Van Bree, corr. 1982] 7 . . . lLlh6 8 .e3 .i.cs 9 .xd4 .i.xd4 10 f4 0-0-0 1 I d3 gS! 12 lDe2 .i.b6 13 h3 g4! intending to answer 14 h4 with 14 .. . g3 ! and pl�y on the king­side) 6 d4 .d7 (6 . . .• h4 transposes to (a) above) 7 .i.e3 0-0-0 8 .d2 lLle7 9 .as �b8 (9 .. . a6!? 10 .i.xa6

OO! is interesting,) 10 liX:3, Capablanca-Corzo, Havana 190 I, ( 10 dS? b6) 10 . . . cS!? II dxcS liX:6 with some play for the pawn . .

c) S .i.c4 � 6 d3 (the most solid choice, 6 .e2 fxe4 7 liX:3 .i.fS 8 .i.�g8?! Lg8 9 .c4 O-O-O! 10 d3 [ 1 0 .xg8? .i.cS 1 1 g3 .g4 12 .f7 :fS is strong, but White contrives to lose his queen �yway] 1O . . . .i.e7 1 1 lLlxe4 .i.xe4 12 .xe4? .i.b4+ 1 3 c3 :ge8 + Zagt-Sinke, corr. 1982; 6 O-O?! White should avoid commit­ting his king like this, 6 . . . fxe4 7 d3 .i.d6 [7 . . . .i.g4!?] 8 g3 �3 9 dxe4 .i.g4!? [9 . . . lLlffi] 10 0 hS? 1 1 .i.xg8? [Black's attack is not quite sound, and can be beaten off with 1 1 fxg4 h4 12 :fS+! vacating the fI square for the bishop: 1 2 . . . .f.xfS 13 .i.fI .i.cS+ 14 �h I hxg3 IS .i.xh3 Lh3 16 �g2] 1 l . . .h4! 1 2 fxg4 hxg3 13 .i.f7+ �d7 14 .e2 .i.cS+ 0-1 , Tschurgulia-Steinikow, USSR 1988) 6 ... lLlffi 7 liX:3 fxe4 (7 .. . �cS!?) 8 .e2 .i.g4 9 .e3 .i.d6?! (9 . . . .i.f5) 10 dxe4 .i.eS 1 I .i.d2 :f8 12 g3 �S 13 f4 .i.d6 14 eS 0-0-0 IS 0-0 (IS .xa7 is even stronger) I S . . . .i.e7 16 :£2 with ad­vantage, Hazai-Hector, Berlin 1985.

d) S �S+!? g6 6 .e2 .i.g7! (very dashing, 6 . . . lDe7) 7 exf5+ lDe7 8 fxg6 :i.fS 9 gxh7 .d7 10 �S+? .i.g6 I I �3 lLlf5, Black certainly has a large lead in devel­()pment for his four pawns, 1 2 .d3 .f7 1 3 .c4 .ffi 14 liX:3 0-0-0 I S d3 :de8+ 16 .i.e3? �e3+! 1 7 fxe3 lLlxe3 18 .a4 :fti 19 .i.e2? xc3+! 20 bxc3 .i.xc3 mate, N.N.-Keres. corr. 1940.

e) S exf5 .i.xf5 6 d4 (6 d3 .d7 7 .i.e3 lLlffi 8 liX:3 0-0-0 9 .i.e2 .i.d6 10 .d2 :he8 I I 0-0-0 is a sensible approach, Vajda-Pessi, Odorheiu Secuiesc 1993) 6 .. . � 7 .i.e3 (7 .i.d3!? .i.g4 8 .i.e2 lLlffi 9 h3 .i.e6 ;t,

3 . . . lOc6 and other third move alternatives for Black 103

Hermann-Thiede. Berlin 1988) 7 . . . 0-0-0 8 .i.d3 (8 � lIe8 9 lOc4 [9 M?1 lbe3+] 9 ... cS 10 "'e2 cxd4 1 1 .xd4 .xd4 12 "'xd4 "'xc2 1 3 �3 "'e4 = Jacobs-Hector, Seville 1986) 8 . . . %6!? 9 0-0 (9 hh6?! 1le8+) 9 . . .... g4 10 .d2 .i.d6 1 1 h3 and Black has some activity for the pawn, Prat-Melchor. corr. 1985.

f) S d3 will almost certainly trans­pose into the main line, once White plays '&3.

5 ...... CS As White has defended his e- '

pawn there would be no point play­ing S . • • 1Fh4 at this juncture. as 6 g3 and "'g2 is possible. The text is per­haps the most logical. controlling d4 and taking aim at Cl. but there is. again. a wide choice:

a) S . . • �fti a l ) 6 exf5 (it seems UMCCessary

to give back a pawn with 6 d4? fxe4 7 .A.gS "'f5 8 "'c4 "'e7 9 0-0 .d7 = Kinnes-Sprod. corr. 197 1172; but 6 d3 fxe4 [6 . . .... b4] 7 �e4 .i.e7 8 .i.e2 0-0 9 0-0 is also good.

. Grivainis-Atars. corr. 1979/80) 6 . . .... xf5 7 d4 "'b4 8 "'d3 .e7+! (8 . . .... g4?! 9 13 "'hS 10 .e2+ Wf7 1 1 "'c4+ 00 12 0-0 lIe8 13 � [or 13 .Cl "'xc3 14 bxc3 with ad­vantage Elburg-Goldt. corr. 1984] 13 . . . �g8 14 .Cl Wh8 I S "'gS "'e7

16 "'xe7 IIxe7 17 IIfe l and White has an extra pawn. and the better position. MalmstrOm-Jansson. corr. 1986) 9 "'e3. Heap-Stummer. corr. 1994/96. 9 . . . �! 10 .13 0-0 with threats.

a2) 6 eS �g4 7 d4 cS 8 dS!? (8 h3 cxd4 9 bxg4 dxc3 10 .xd8+ �d8 1 1 :iogS+ [ 1 1 "'d3 !?] 1 1 . . .... e7 12 0-0-0+ We8 White has a slight edge. Melchor-Bonte. corr. 1998) 8 . . . �xeS 9 .e2 .e7 10 "'f4 lOg6 (best, 1O . . . 1Of7?! 1 1 lObS gS 12 �c7+ Wd8 13 lOxa8 gxf4 14 0-0-0 .xe2 1 S "'xe2 "'d6 16 Ilhe I and Black has no easy method of attack­ing the a8-knight, Nilsson-Weber. corr. 1990) 1 1 "'xc7 Wd7 12 .xe7+ "'xe7 13 d6 "'xd6 14 "'xd6 �d6 I S 0-0-0+ Wc7 16 .i.c4 White has a slight plus. Melchor-Turian. corr. I99S.

b) S . . .• e7?!. popular with some. but it doesn't really fit in with Black's attacking ideals. 6 d3 (6 "'c4 lOfti [6 . . .... e6 7 .i.xe6 .xe6 8 0-0 0-0-0 9 .f3!? {9 exf5 .xf5 10 d3 ;t Kapitaniak-Andersen. corr. 1978} 9 . . . f4? 10 .xf4 with advan­tage. Sinke-van Eijk. corr. 1984] 7 O-O! fxe4 8 d3. Strelis-Harju. corr. 1984. 8 . . .... g4 9 .d2 0-0-0 10 lie 1 . White will win the e-pawn; 6 .e2 is always a good idea. White will try to play an endgame. 6 . . . lOfti 7 d3 [7 eS �S?! Jean-Destre�. corr. 198 1 . when White can play the bor­ing 8 �dS cxdS 9 d4 and the black position is deprived of its dyna­mism. but 7 . . . lLlg4 is better] 7 . . . fxe4 8 lOxe4 �e4 9 .xe4 "'e6. Elburg­Oren. corr. 1994/96. 10 "'e3 and Black has less than nothing for the pawn) 6 . . . lOfti 7 .13 (7 "'gS "'d7 8 f3?! [8 .e2] 8 . . . 0-0-0 9 "'e2 h6 10 "'e3 Wb8 1 1 0-0 gS! 12 .el f4 13 "' Cl hS gave Black a nice kingside pawn roller in Janikowski-

104 3 . . . 1iX6 and other third move alternatives/or Black

Kapitaniak. Poland 1979; 7 .e2 fxe4 is also fme, as above) 7 . . . fxe4 8 dxe4 .*.J4 9 Wf4 0-0-0 10 13 .*.e6 1 1 .*.d3 Wb4 1 2 a3 with advantage, Harper-Magee, corr. 1991 .

c) S . . . .*.b4 6 .*.c4 (6 eS?! .d4 7 .e2 .*.e6, Purins-Strelis, corr. 1974177, but 6 exfS!) 6 . . .• d4!? (6 . . . .*.xc3!? 7 bxc3 � is possible) 7 .*.xg8?! (7 -..0+ g6 8 .e2 is stronger) 7 . . . .*.xc3! 8 0-0 .*.xb2 9 c3 .*.xc3 10 �3 hal 1 1 Wti+ �d8 12 WfS+, drawing by perpetual a few moves later, Stayart-Bullockus.

6 d3 Once again, there are several

alternatives: a) 6 Wf3 .e7 (6 . . . lOe7!? worked

out well for Black in Dries-Sinke, corr. 1982: 7 .*.c4 llfS 8 O-O?! [8 d3 �g6] 8 . . .• d4! 9 .*.b3 [9 .e2 f4 lO d3 13 I I pf3 IClg6] 9 . . . fxe4 10 �S+ [ 10 _xe4 loses n, but was the better choice] 10 . . . 86 1 1 � .*.fS ;) 7 .*.e2! (7 d3 lOffi 8 .*.e3 .*.xe3 9 WJe3 0-0 10 O-O-O?! [ 10 .*.e2] 10 . . 6e4?! [ lO . . . �g4 1 1 _g3 f4 12 .13 �S provides genuine positional compensation] I I �e4 [ 1 1 f3! exploits the e-file _j)in] 1 1 . . .�g4 1 2 Wel ? [ 1 2 Wg3] 1 2 . . . �xf2! 1 3 �f2 .xel 14 llxel llxf2 I S lle2 llf8, draw, Bullockus­Stayart, corr. 1977) 7 . . . �ffi 8 exfS g6! 9 d3! .*.xfS lO .*.e3 .*.b4 I I

O-O!? .*.xc3 1 2 bxc3 .*.g4 13 Wg3 .*.xe2 14 IIfe l 00 1 5 c4 �e3 16 llxe2 0-0-0 1 7 llxe3 Wa3 is better for White, though soon drawn in Lamoureux-Kosten, Paris, 1994.

b) 6 �S+ g6 7 We2 �7 8 d3 (8 �4!? .*.d4 9 eS .dS 10 f4 a6 1 1 c3 h7 1 2 d4 .*.e6, White is better,

but the offside knight is awkward, Downey-Stamer, corr. 1994/97) 8 . . . .i.e6 (8 . . . fxe4?! 9 dxe4 .*.d4 10 .d3 .*.e6 I I �2 Ag7 12 �f4 with advantage Budovskis-Stummer, corr. 1990) 9 Ae3 .*.b4 (9 . . . Axe3?! lO .xe3 .d6 I I g3 0-0 12 f4 with advantage Budovskis-Morgado, corr. 1974/77) lO g3 0-0 1 1 .*.g2 fxe4 12 dxe4 bS, Grobe-Stamer, corr. 1994/95, 13 .d3, White is, not for the first time in this section, a whole pawn to the good.

c) 6 We2 is very similar to the above, except that Black does not have the dark-squared weakness on the kingside. 6 . . . lOe7 (6 . . .• e7?! Black doesn't really want to play an ending a pawn down, 7 d3 �fti [7 . . . Ad7 8 exfS 0-0-0 9 .xe7 .i.xe7 10 .*.e2 AxfS I I 0-0 �ffi 12 .*.e3 with advantage, Budovskis-Grobe, corr. 1977] 8 A8S .*.d4 9 exfS AxfS lO .xe7+ Wxe7 I I .*.e2 h6 1 2 Ad2 and White is a pawn up, Krustkains-Downey, corr. 1991) 7 �! (7 d3) 7 . . . .*.b6 (with the black pawn on g7, 7 . . . .*.d4?! is worse, 8 c3 Afti 9 eS and the bishop is 8ettting kicked around the board) 8 lOxb6 cxb6 9 d3 0-0 lO .*.d2 .i.c6 I I Ac3 and White has the bishops to add to his extra pawn, Daurclle­Felber, corr. 1997.

d) 6 .*.c4? Strangely, this is a recommendation of BeO, which assesses the position as ± after White's seventh move. I can only think that someone was looking at the wrong position!

3 . . . 1Oc6 and other third move alternatives/or Black 105

I

6 . . . .i.�t2+ 7 � (7 Wxf2 may represent the best chance, 7 . . .• d4+ , Wel [S Wg3? �lIy tempts fate, � . . . f4+! 9 WXf4 lDe7 10 .e2? { IO .i.f7+ allows White to struggle on a bit ) 10 . . . :f8+ 0-1 , Schuttrich­Grobe, corr. 1976] S . . .• xc4 9 d3 .e6 10 .e2 IOffi, Burk-Leisebein, corr. 19S I , and I really feel that White should take this opportunity to go into the equal endgame by 1 1 exfS .xe2+ 1 2 Wxe2 .i.xfS) 7 ... .i.d4! (the best square for the bishop. 7 . . . lDe7 is good too, S d3 .i.b6 [S ... .i.d4!] 9 .e2 f4?! [9 . . . fxe4] 10 1I'hS+ IOg6 Mathieu-From, corr.; 7 . . . .i.b6?! S .0 .ffi [perhap� best, S . . . lOffi 9 d3 IOg4 10 exfS .i.xfS, Marquez-Micheloud, corr. I 974nS, 1 1 .i.f7+ Wxf7 12 • xfS+ =; S . . .• d4 9 d3 .i.d7? {bet­ter 9 . . . lDe7) 10 .i.gS IOffi 1 1 :el 0-0-0 12 eS IOg4 1 3 e6 and White is on top, Jedlicka-Walther, corr. 1975] 9 d3 .i.d7 10 .i.f4 0-0-0 1 1 eS and White has avoided the worst, Hernandez-Gan:ia Camejo, corr. 19S9) S .i.xgS (S .f3!? lDe7 [or S . . . Wh4] 9 d3 :f8 ;; S d3 .ffi 9 eS What else? 9 . . .• xeS 10 .i.d2 .i.d7 and, exceptionally, it is Black who has the extra pawn, Randa-Payot, corr. 1997) S . . . LgS (S . . .• ffi!? is also possible) 9 -.bS+ g6 10 .xh7 .i.e6 1 1 d3 .ffi 12 We2 0-0-0 1 3

.i.f4 :d7 14 Wh3 .i.xc3 I S bxc3 fxe4 =i= Denaro-Fanoni, corr. 1992.

e) 6 .i.e2 IOffi 7 eS? (7 d3 trans­poses to the main line) 7 . . .• d4 S 0-0 .xeS, Pedersen-From, Den­mark 1974, is fairly level.

6 ••• lOf6

7 .i.el In the game Bertram-Stummer,

corr. 1991 , White played the infe­rior 7 .i.e3?! .i.xe3 S fxe3 fxe4 9 .i.e2?! (White can maintain a small pJus by 9 lDxe4 0-0 [9 .. . lDxe4? 10 WhS+ g6 1 1 .eS+ Wf7 12 dxe4 is strong for White] 10 1Ot2) 9 ... lDdS!? 10 Wd2 exd3 1 1 .xd3 lDxc3 ( 1 1 . . ..gS) 12 .i.hS+!? g6 13 .i.xg6+ bxg6 14 .xg6+ and had an attack that was probably sufficient to dnw� and ultimately won .

7 ••• 0-0 8 0-0 flle4 Alternatively, there is S . . . .i.d4 9

.i.f3 (a better way of defending e4 than 9 .i.gS .i.xc3 10 bxc3 fxe4 1 1 .i.xffi .xffi 12 dxe4 .xc3 1' .d3 .eS 14 .e3 gS!?, stopping f4, and limiting White's advantage, Incelli­Elburg, corr. 1992) 9 . . .• d6 (9 . . .• e7 10 lDe2 .i.eS 1 1 1Og3 .i.x� 12 bxg3 fxe4 13 dxe4 .i.c6 14 .e2 with a solid pawn more, Harper-Stummer, corr. 1991) 10 exfS .i.xfS 1 1 .i.e3 (1 1 �White should not be afraid of exchan�� I I . . .AeS 1 2 g3 Axc3 13 bxc3

/06 3 . . . 1Dc6 and other third move alternatives for Black

14 .d2 .i.h3, Vargha - Schultz Pedersen, corr. 1979, Black is close to equality.

9 �xe4! To exchange the last two knights.

9 dxe4 is not as good, 9 . . .• eS (or 9 . . .• e7 10 "'c4+ "'e6 1 1 .e2 b5! 12 .i.xe6+ .xe6 13 "'g5, Paul­Walther, East Germany 1975, Black should have no I'roblems, i.e. 13 ... :aeS 14 .i.xffi :,effi 15 )hel Ilets) 10 "'c4+ ( 10 "'0 .e5!? [ 1O . . .• g6] 1 1 g3 -*.h3 12 "'f4 .e6 13 Ilel ? :adS?! [ 1 3 . . . �g4! is more incisive, 14 .i.xg4 "'xB4 +] 14 .e2 [ 14 .cl �g4 15 lle2 defends] 14 . . . �g4 15 .i.xg4?? "'xg4 16 .CI .i.h3 1 7 .e2 :,ef4 I S gxf4 .g6+ 19 <ithl .g2 mate, Svendsen­Downey, corr. 19S9) 1O . . . <ithS 1 1 .i.e3, Schuetze-Bondick, corr. 19S3, when Black's simplest line was per­haps I I . . .-*.xe3 12 fxe3 b5 1 3 .i.d3 .e5 and the extra white pawn doesn't count for too much.

9 ••. �xe4 10 dxe4 "'4

11 -*.c4+ 1 1 .i.e3 !? is a simple solution,

I I . . ..i.xe3 12 .i.c4+ WhS 1 3 fxe3 .i.g4 14 .d4 and White exchanged both �s of rooks to exploit his passed e-pawn(s), Legemaat-Van de Velden, corr. 1996.

1 1 -*.0 is more passive, 1 1 . . .... e6 12 .e2 b5 13 b3 .i.d4 14 Ilbl and

Black's compensation for the pawn is minimal, Brinck Claussen-Schultz Pedersen, Copenhagen 1996.

1 I ••• <itb8 11 "'e3! Not 1 2 .e2? "'g4 ll .el IlaeS

with strong threats, nor 12 .e I ? "'g4 13 "'e2 (best, if 13 j,e3? j,0! 14 gxO? "'d6 IS f4 lhf4 mating; if 1 3 "'d3 IlaeS menaces both . . . .i.0, and ... lle5-h5) 13 . . .:xt2? (pretty, but not correct, 1 3 . . . .t.xe2! 14 .xe2 :aeS wins the e-pawn, with advan­tage) 14 :,efl illS I S j,f3 1lx0 16 gxO ( 16 "'e3! is nasty, 16 . . . j,xe3 17 .xe3 and Black's weak back rank scuppers him) 16 . . .... xO 17 "'e3? (if White wants to play for a win, 1 7 .d2! is the move, evacuat­ing el for the king) 1 7 . . .• g4+ I S <itCl -.u+ 19 <itgl .g4+ draw, Doncevic-Hermann, Bundesliga 19S5.

11··..t.14 12 . . .... xe3 13 fxe3 "'g4 14 .d4

transposes to the note to move eleven.

13 .d3 Il.d8 After 13 . . . j,d6 14 f4 IladS 15 e5

White reminds Black that he has a large kingside pawn majority: 1 5 . . .... �5 ( 1 5 . . .... xe5?? 16 fxe5 1lxd3?? 1 7 1lxlS mate) 16 .c3 and Black's case is pretty hopeless, Epstein-Walsh, corr. I99S.

14 .c3 1l0? This fails miserably, 14 . . .• e7 is

much better, although Black's dreams of attack are no more.

IS pO j,xO 16 "'e6 The problem, Black has no

checks. 16 •• ..t.d4 16 . . . j,d6 17 e5. 17 .85 Keeping an eye on g5, and

winning, Blaskowski-Stamer, corr. 19S5.

3 . . . 1Oc6 and other third move alternatives/or Black 107

Al 4 "'5+ g6 5 ��g6 White goes for the win of mate­

rial. which will involve his queen straying away from the defence of her ,king. although by now it is �robably too late to change tracks. S ll)xc6?! was played in a game of Keres' considered in A I ).

5 ••• �r6

I :� • . �B � 1 �� 1 � � 1 �� ... � lf� .

.

'��� � .. � �'"Z.J� 'Zi .> .� � :t� � '� � � . r'iti! �� � � � �2 � .!1. � � � � � � .:l- '/� .:l- r'� r'� .:l- 'Y�, � � � � � � � §{)�" : · " :il�§

6 -.u , It is far 'from certain that this is

stronger than 6 1i'h4. although this last does allow 6 . . . 11gS?! (in addi­tion to the 6 . . . hxg6!? 7 .xhS dealt with later, and 6 . . . d6?! transposing into C) with the following lines: I

a) 7 It)xf8 1l&4 and a I ) S 1i'h6 llxe4+ with a further

sub-division: a l l ) 9 .i.e2! .e7 10 �3 ( 10

�xh7!? llxe2+ [ 1O . . . fuh7? 1 1 �3 llOO 12 1i'hS+ �dS 13 �S .g7 14 .h4+ �7 IS �f4 llc6 16 c3 llh6 17 .g3 1i'hS IS d4. and White had consolidated in Kellerer-Stummer. corr. 199IJ 1 1 �dl �g4 [ 1 1 . . . �xh7? allows 12 1i'hS+J 12 _g6+ �dS 13 .gS+ forcing off queens, 1 3 . . .• eS 14 .xeS+ llxeS I S lift . Vita-Houthuijzen, corr. 1991 . when I S . . . llhS 16 h3 llxh7 17 bxg4 fxg4 offered the best chances) 1O . . . 11xe2+ ( IO . . . �4 1 1 �e4 .xe4 12 0-0 �e2+ 13 �h l �g4. Meshkov-Sukhin, USSR 1977. 14

1i'h4! Wxf8 IS 0 with extra mate­rial) 1 1 �e2 � 12 0-0 �e2+ 13 �hl d6 14 �h7 �xh7 I S 1i'hS+ �f8 16 d3 �c I 1 7 :.xc I and White has the better chances. Atars-Morgado. corr. 1976.

a 12) 9 �dl? �g4 10 1i'hS+ �f8 1 1 .xf3+ �g7 ( 1 1 . . .�gS 12 d3 dS 13 .0 �t2+ 14 .xt2 .i.g4+ I S .i.e2 [if I S �d2 then I S . . .• gS+ 1 6 �c3 .g7+ 17 �d2 llf8 I S .cS .gS+ 19 �c3 .eS+ is strongJ I S . . . .i.xe2+ 16 Wxe2 llxe2 17 �e2 .eS+ I S �d2 .g6 19 �c3 d4+ 20 �b3 lOas+ winning. Capablanca­Corzo. Havana 190 I ) White is in a bad way: 1 2 d3 ( 12 b3 dS 13 .i.b2+ �gS 14 WO Wh4 I S g3 1i'hs 16 .i.e2 �t2+! 0-1, Gubic-Kurucin •

Yugoslavia 1969. as 17 .xt2 llxe2 IS .xe2 .i.g4; 12 c3 dS 13 .0 �3+! 14 .xe3 llxe3 I S dxe3 Wh4 gives good chances, Racz-Motyvay. corr. I 967/6S; 12 .i.d3 dS 13 .0 �xfl+ 14 .xt2 .i.g4+ I S .i.e2 llxe2 16 .xe2 .i.xe2+ 17 �e2 �4+ I S �dl .gS 0-1 , Llorach­Herranz. Barcelona 19S I ; 12 .0 .h4 1 3 8}? �4! 14 .g2 �xt2+ I S .xfl .g4+ 16 .i.e2 llxe2 17 h3 .e4 0-1 . Nemo-Kleinerteufel. Sitzbad 1932) 12 . . . d6! 13 .0 �t2+! 14 .xfl .i.g4+ I S .i.e2 (forced with the pawn on d6 instead of dS. I S �d2? .gS+ 16 �c3 .as+ mates) I S . . . llxe2 16 .xe2 .i.xe2+ 17 �e2 1i'h4 ( H ." Krenzisty-Borgeson, Denmark 1964.

a2) S 1i'h3 is less active, S . . . llxe4+ with the same choices:

a2 1 ) 9 �dl �g4 10 1i'hS+? (tempting. but 10 Wg3 ! is stronger. then 1O . . .• e7 1 1 .i.d3 dS 1 2 .i.xe4 dxe4. Bomemann-Mackenzie. New York I S77. when 1 3 h3 is simplest) 1O . . . Wxf8 1 1 .xf3+ �g7 transpos­ing to S 1i'h6.

108 3 . . . �6 and other third move alternatives for Black

a22) and 9 �e2 � 10 1&3 ( 10 1Oxh7?! Mathe-Matwienko, corr. 10 . . . Le2+ 1 1 Wfl? [ 1 1 �dl ] I I . ..lDg4 [ 1 1 . . .:e 1+! 12 Wxel .e7+ 13 .e3 lOxc2+ wins the queen] 12 11bs+ �e7 13 .gS+ Wf7 when 14 .xd8 Lf2+ would have been a draw) 10 . . .• e7 1 1 .d3 ( 1 1 lOxc4!? .xc4 12 .d3, Grim­Sursock, Malta 1980, 12 . . . lOxc2+ unclear) I l . ..lOxe2 12 lOxe2 ( 12 lDxc4?! lOxcl 1 3 Lcl WxfS 14 0-0 fxc4 is very pleasant for Black, Sukhareva-Shtejnikov, Voronezh 2000) 12 . . . WxfS 13 f3 :eS 14 �f2?! .cS+ IS �f1 , messy, Pegoraro-Elstner, Marostica 1995;

b) 7 cS! lDxeS (this is not satisfac­tory, but 7 . . . Lg6 8 exf6 .xf6 [8 . . . Lf6 9 .i.e2 .e7 10 0-0 is clearly advantageous for White, Morgado-Grobe, corr. 1972175] 9 .xf6 [not 9 .xh7?? .e6+ 10 .A.e2 :h6 winning the queen, Weening­Goedhart, corr. 1982] 9 ... Lf6 10 c3 dS 1 1 d4 .i.d7 1 2 .i.e2 and Black has no discernible compensation for the pawn, Den Hertog-Goedhart, corr. 1982) 8 lDxeS .e7 9 .i.e2 (9 d4? :g4 10 Wit3 Ld4 1 1 f4 d6 12 .i.bS+ c6 13 0-0 dxeS 14 .i.e2 is very messy, Matwienko-Schuttrich, corr.; 9 f4 :g4 10 .12 d6 1 1 d4 dxeS 12 fxeS lDc4 1 3 .gl .i.e6 is not clear either, Black has a useful advance in development, Steuer­Stummer, corr. 1989) 9 . . .• xeS 10 d4 .e6 (1O . . .• aS+ 1 1 �dl ! � 12 :el .i.e7 13 .i.gS with advantage Miiller-Elburg, corr. 1990) I I .i.gS .i.b4+ ( 1 1 . . ..t.e7 1 2 1&3 �d8 1 3 0-0-0...&6 14 .t.d3 with advantage van der Sterren-van der Lijn, Apel­doom 1978) 12 1&3! (getting on with his development, 12 c3 �e7 13 lDd2 dS 14 O-O-O? LgS IS .xgS • xe2 16 :hel .a6 1 7 Le7+ Wxe7 1 8 .g7+ is unclear, Magee-

Stummer, corr. 1991) 12 ... 00 1 3 O-O-O! �xc3 14 bxc3 lOxc3 IS �hS+ � 16 :be1 lDc4 17 Lc4! fxc4 18 .f4+ �g7 19 dS :f8 20 dxe6 Lf4 21 e7 1-0, ChmiIewski­Stummer, corr. 1991 .

Otherwise, 6 .xf51 i s a mistake, 6 . . . d6 7 .f3 bxg6 8 �bS �7 and the white pawns do not provide sufficient compensation for the piece, Ludvigsen-Sorensen, Aalborg 1872.

6 ••• bKg6 a) 6 . . . :g8? does not have quite

the same eITect in this position, as there will be no ... :g4 move, fork­ing queen and e-pawn: 7 lOxf8 (not now 7 cS?! as 7 .. . bxg6 8 exfti .xf6 9 1&3 �cS I 0 �c4 .eS+ gives ex­cellent compensation for the pawn, Steuer-Stummer, corr. 1990) 7 . . .• e7 (7 . . . Lt8 8 exf5 � 9 .e3+ .e7 10 .xe7+ Wxe7 1 1 -*.d3 Steuer-Stummer, corr. 1989; Black will recapture the i3-pawn, but that still leaves him two down) 8 lOxb7!?, a waste of time, but good anyway, 8 . . . lOxc4?! (8 .. . lOxh7) 9 .e3 ! dS 10 d3 f4 I I .xf4 10141 when instead of 1 2 �dl 11 Lg2! with complications, Petersen­Sorensen, Denmark 1873, 12 dxc4 must win .

b) Objectively 6 .. . fxc4? is insuffi­cient as Black obtains positions

3 . . . 1Oc6 and other third move alternatives for Black 109

similar to the main line but without having captured the knight, but it's fun: 7 lOxh8 (7 �iS? :XiS 8 .i.e2 .e7 9 0-0 dS, Black has a strong cef\tre, and two open files for the attack, I 0 � .i.e6 I I d3 lOd4 1 2 .i.dl 0-0-0 1 3 .i.gS filf5, Masur­Stamer, corr. 1979, 14 .f4 h6 with good chances) 7 . . . dS 8 "g3 (White . impedes Black from playing . . . .i.g7, and taking the knight. Others: 8 "'3 .i.g7 [8 . . . .i.d7!? is also inter­esting, the threat is .. \.�, and 9 c3 {9 c:4?! doesn't look �ght, 9 . . . lOd4

1 10 "c3 cS 1 1 fila3 ( 1 1 d3 .e7 12 .i.e3 exd3, Champion-Foutelet, corr. 1982, 1 3 .i.xd3 and White must surely be better) I I .. ..e7 1 2 fllc:2 filxc2+ 1 3 .xc2 d4 14 b4 .i.g7 I S bxcS, GUrd-Maltez, corr. 1999, when I think I would play I S ... .i.c6, .. . 0-0-0, and . . . .i.xh8, with compen­sation for the exchange} 9 . . . flleS!? {9 . . . .i.g7} 10 d4 exd3 Laganes-van Mulder, Paris 1983, I I .xb7 is ran­dom but, assuming Black recaptures the knight fairly soon, he 81ay not have enough compensation for the exchange; 8 . . . .i.cS?! 9 .i.bS .i.d7 10 .i.xc6 .i.�c6 I I .g3 .e7 1 2 b4!? with advantage, Strautins-Gaard, corr. 1992] 9 .*.bS .*.xh8 10 fllc:3 .d6 I I 0-0 filg4 1 2 g3 .*.e6 with some play for the exchange, Melchor-Stummer, corr. 1990191 ; or 8 � lOd4 9 fila3 when 9 . . . .*.xa3 is probably to be avoided, Black re­captures a rook, but loses his dy­namic play, 10 bxa3 �c2+ I I �dl �al 12 .i.b2 d4 1 3 0, Kozlov­Alberts, corr. 1979, 1 3 . . . �e7!? 14 .*.xa I .xh8 I S fxe4 with excellent winning chances; 8 .e3 filg4 9 .f4 fllc:eS 10 h3 .*.d6 I 1 �dl �4 12 d4 exd3 1 3 .*.e3 �e3+ 14 .xe3 [ 14 fxe3 �S+] 14 . . . dxc2+ I S �c2 .*.f5+ 16 �c I 0-0-0 and Black will suc:c:eed in taking the

knight while retaining his attack, Woods-Pierce, corr. 1916) 8 . . .• e7 (nther than 8 . . . lOd4 9 fila3! [9 .eS+ f£le6 10 d3 .*.d6 I I .c3 .e7 1 2 fild2?, Steuer-Stummer, corr. 1989; objectively, Black should take the dnw by 1 2 . . . .*.b4 13 .eS .i.d6] 9 . . . .i.xa3 [once again, Black regains material, but loses the initiative, 9 . . . .*.d6!? keeps the tension, 10 .g7 .*.e6, Schultz. Pedersen-Sc:huttrich, corr. 1977, I I c3 filf5 12 .i.bS+!? c6 13 "xb7 cxbS 14 �bS with ad­vantage, is one possibility] 10 bxa3 lensen-Svendsen, corr. 1990, as 10 . . . filxc2+ I 1 �dl filxal 1 2 .*.b2 with advantage, solving all of White's problems) 9 .i.bS .*.g7 10 filg6 bxg6 1 1 .xg6+ �d7 12 a4 a6 1 3 .*.xc:6+ �c6, lensen-Stummer, corr. 1990; Black has a lead in de­veloement, but it is insufficient.

7 .xbS .e7

In the game Hindle-1.Littlewood, Hastings 1963, Black played the in­ac:c:unte 7 . . . fxe4? and was horribly crushed: 8 d4! �f7 9 fllc:3 �d4 (9 .. . dS 10 .*.e3 .*.f5 offers better chances, although Black has little for the exchange) 10 .*.gS filxc2+ I I �dl filxal 1 2 .*.c:4+ dS 1 3 �dS bS (otherwise, 1 3 . . . .*.g7 14 .xd8 .*.g4+ I S �cl :Xd8 16 filxf6+ .*.e6 17 .*.xe6+ �e6 1 8 filhS is also hopeless) 1 4 .*.xffi .d6

J J 0 3 . . . llX6 and other third move alternatives for Black.

I S �7+ �e6 16 _g8+ �d7 17 .i.xbS+ c6 18 _17+ .i.e7 19 .i.xe7 .e6 20 .xe6+ 1 -0.

8 d3! White must take great care, espe­

cially with his queen, who can find herself short of squares:

a) 8 � fxe4 (or 8 . . . llXl4 which is also good, 9 lDa3 fxe4 10 c3 �f5 1 1 �3 dS with plentiful compensa­tion, Elburg-Svendsen, corr. 1990, but 8 . . . dS?! 9 .i.bS Wf7 10 .i.xc6, Kurtovic-Atars, corr. 1974, is less effective) 9 .i.bS?! (9 .i.e2 llXl4 10 .i.d I dS 1 1 0-0 �f5 with typical play against the white queen, reminiscent of Svedenborg's line, Vargha-Hammar, corr. 1974, but 9 _g3! is a better bet) 9 . . . llXl4 10 �3 c6 1 1 .i.fl dS 12 .SS .i.fS 1 3 .e3 � 1 4 .i.e2 d4 I S .g3 d3 16 .i.dl .i.h6 1 7 �4 .g7 18 g4 �f4 19 .g3 dxc2 20 �xc2 �+ 2 1 �fl .i.f4, when the black pieces were crawling all over White's posi­tion, Diepstraten-Sinke, corr. 1984.

b) 8 llk3 defends the e-pawn, as per the main line, 8 . . . 1bb4 (8 .. . dS?! 9 d4! [.i.gS is coming. Other moves are less promising: 9 d3 �b4 10 �d I Ibg4 1 1 .i.e3 f4 1 2 h3 fxe3 13 bxg4 exf2 14 a3 dxe4!? I S axb4 exd3 exposing White's king, Prust­Brondick, corr. 1987; 9 .i.bS dxe4 10 .i.xc6+ bxc6 1 1 0-0 Wf7 12 d3 .i.g7 13 � .i.a6 14 lIel IIh8 and Black has attacking chances on the h-file, Steuer-Stummer, corr. 1989] 9 ... lbxd4 10 .i.gS! �xc2+ 1 1 �dl Ibxa I 1 2 �xdS and Black is lost, Rodriguez Capay - Acosta, Amici Sumus 1996) 9 d4!? 1bxe4 10 Ibxe4 .xe4+ I 1 .i.e3 �f7? (Black natu­rally avoids I l . . .lbxc2+? 1 2 �d2 �xal-1'3 .i.d3 [ 1 3 .i.h6?? .c2+] 13 . . .• e7 14 lIel , but I l . . .f4 !? is feasible, as 12 .i.d3 1bxd3+ 1 3 cxd3 We7 14 0-0-0 fxe3 I S IIhel [it

looks like the black pieces are going to be massacred before they move, but. . .] I S . . . e2! 16 Ild2 �f7 17 IIdxe2 _gS+ 18 �bl �S swap­ping queens with good chances) 12 .i.d3 Illxd3+ 13 cxd3 and White is well on top, Percz Cruz-Acosta, corr. I99S.

8 •.. fxe4

9 .i.e3 This is eminently sensible. a) 9 �dl ?! �g4 10 .i.e3 dS 1 1

.i.e2?! ( 1 1 h3) 1 1 .. .d4! ( I 1 . . .Wf7 threatening . . . .i.g7, 12 .i.xg4 [ 12 .c3? d4! 1 3 Wb3+ �e8 14 h3 { 14 .i.xg4? .i.xg4+ I S �c l dxe3 16 .xb7 looks promising, but 16 . . . �d7 17 _xa8 exf2 and White is busted) 14 . . . lbgeS I S .i.f4 .i.e6 16 c4 dxc3 1 7 .xc3 �xd3 1 8 .i.xd3 1Id8 with a venomous attack, Schumacher-Vetter, Porz-Wahn 1980] 1 2 . . . .i.xg4+ 1 3 �c l exd3, Chmilewski-Svendsen, corr. 1990, when the obvious 14 /be3 may keep White on t�) 1 2 h3 ( 12 .i.xd4? 1bxd4 1 3 .xd4 .i.g7 loses the hi-rook) 12 . . . dxe3 13 hxg4 exf2 14 � exd3 IS cxd3 .i.e6 16 1Ib7 .i.f7 17 lOe4 0-0-0 1 8 �xf2 llXl4 and the white major pieces are mere spectators to the action, Svendsen­Pape, corr. 1992.

b) 9 d4?! Although a tempo down on the Hindle-I.Littlewood game,

3 . . . �6 and other third move alternatives for Black J J J

this is still viable, 9 . . . �d4! (9 . . . Wf7?! , adding defence to the ffi knight, and preparing . . . .i.g7, evict­ing the white queen, 10 c3 [ 10 .i.gS!? .i.g7 I I � ltlxd4 12 �3?! { 1 2 .i.c:4+} 12 . . . �c2+ 13 Wdl �al 14 .i.c:4+ dS! { 14 . . . WfB? IS /[}dS .eS ( I S . . .• cS?? 16 .i.xffi .xc:4 17 .i.xg7+ Wxg7 1 8 .e7+ mates) 16 .i.xffi .i.xffi 1 7 .x:ffi+ .xffi 1 8 �ffi and the a I -knight will be lost} I S �dS .d6 16 .i.xffi .i.xffi draw, Jensen-Magee, COrT. 199 1 , as 1 7 .xffi+ .xftN 8 �ffi+ Wxffi 19 Wd2 =] 10 . . . dS 1 1 .i.e2 .i.g7 1 2 � .i.e6 13 .i.gS and Black is an exchange down for nothing, GHrd-Pape, COrT. 1994/9S) 10 .i.gS �c2+ 1 1 Wdl ltlxal 12 • xffi .xffi 13 .i.xffi Wf7 14 .i.c3 dS I S b3 .i.g7 16 h3 .i.e6 17 Wc:1 , Gaard-Svendsen, COrT. 1991 , when 17 . . . �b3+ 1 8 axb3 d4 19 .i.al .i.xb3 provides Black with an ar­mada of queenside passed pawns

c) 9 dxe4!? is a simple solution to White's problems: 9 . . .• xe4+ 10 .i.e3 Wf7 1 1 /[}d2 ( 1 1 1&3! .i.b4 12 0-0-0 is promising) 1 1 .. ..xc2 12 .i.c:4+ d� 13 .i.b3 .d3 14 .i.dl , Burk-Leisebein, COrT. 1982, 14 ... .i.g4 with some play for the material.

9 •.• d5

10 .i.c5! Both finding space for the queen,

and exchanging a piece, but there are other reasonable alternatives:

a) 10 .i.e2 exd3 (not 1O . . . .i.f5 1 1 d4 [ 1 1 dxe4!?] 1 1 . . .0-0-0 12 1&3 /[}hS!? 13 .i.xhS .i.g7 14 1th7 l%h8 I S .i.xg6 lhh7 16 .i.xf5+ Wb8 17 . .i.xh 7 :i.xd4 1 8 .i.xd4 �d4 19 0-0-0 cS 20 �dS .xh7, Burk-Leisebein, East Germany 1982, when 2 1 c4 must be good for White) 1 1 .i.xd3, Geisler-LOmer, COrT. 1992, and now Black should by 1 1 . . .d4!? 12 .i.xg6+ wd8 13 ltk3 dxe3 14 ()-()-()+ .i.d7 I S l:lhe l .g7 16 .xg7 .i.xg7 as White can­not capture on e3 with his rook be­cause of . . . .i.h6 .

b) 10 dxe4 is also promising; 10 . . . ltlxe4 1 1 .i.d3 !? ( 1 1 .i.e2 is a more staid approach, 1 1 . . . .i.d7?! [ l 1 . . . .i.e6 12 �3 { l2 O? O-O-O! 13 fxe4 .i.g7} 12 . . . 0-0-0 is better] 1 2 1&3 and White is soon consolidat­ing, Dupont-Redon, COrT. 1991) 1 1 . . .d4 (as this doesn't work, then Black should consider 1 1 . . ..i.(5) 12 .i.h6! (the discovered check is of no great use to Black) 12 . . . 1t1g3+ 13 Wd2 �hl 14 .i.xg6+ wd7 I S .xfB .xfB 16 .i.xfS and White's queenside passed pawns will win the endgame, Tait-Elburg, COrT. training game 1997.

c) 10 d4?! .i.f5 1 1 1th4?! 0-0-0 12 c3 .e6 13 /[}d2 .i.e7 14 0-0-0 Itlg4 I S .g3 gS 16 f4 .i.d6 17 l:lel �e3 0-1 , van Eijk-Sinke, COrT. 1984.

10 •••• xc5 1 1 .d6 .i.f5 Il dxe4 /[}d4

If 1 2 . . . dxe4 then 1 3 1&3 ( 1 3 c3?! e3 is less clear) 13 ... e3 14 0-0-0 must surely be winning as Black is playing without his one and only rook.

1 12 3 . . . �6 and other third move alternatives/or Black

13 e:d5! IOxcl+ 14 Wdl lOxal 15 .t.d3 .d6 16 :e1+ Wd7 17 .17+ .t.e7 18 1Ie6 winning, Femmel­ViUarreal. COrT. 1997.

A3 4 d4

Recommended by Nunn and NCO. so this move is likely to be­come White's most popular.

4 ••• �? Apart from this. my previous

recommendation. which is doubtful, Black has a wide choice:

a) 4 . . .• f6!? 5 lOe3 (5 IOxc6 dxc6 transposes to 4 1Oxc6) 5 . . . .t.b4. best. (the game Morcno-Padula. COrT. 1980. terminated abruptly: 5 . . . lOxe5? 6 lOds .c6 7 dxe5 d6?? [but Black is losing anyway] 8 .i.b5 1 -0) 6 IOxc6?! (6 exf5! lbxe5 7 • e2) 6 . . . .t.xc3+ 7 bxc3 .xc6 8 exf5 .xc3+ 9 .t.d2 .xd4 10 .i.d3 1Oc7 1 1 0-0 0-0 and Black has a rea­sonable position, Van de Velden­Harper. COrT. 1999.

b) 4 . . . lOxe5 5 dxe5 d6 (5 . . . fxc4 6 .d5 1Wh4 7 .i.c4 IOh6 8 .t.xh6 .xh6 9 0-0 and Black will experi­ence severe problems completing his development, Van Forccst-Bruin COrT. 1 88 1 ) 6 exf5 .i.xf5 7 .0 .c8 8 .t.d3 .t.xd3 9 .xd3 .d7 (9 . . . dxe5 10 1Wb5+) 1O .c4 and White has a pawn more. Ruggeri Laderchi­Stcphen. COrT. 1999.

c) 4 . . .• e7?! 5 lOe3! (5 lOxc6 .xc4+ 6 .t.e2 .xc6 7 0-0 dS 8 lOe3 .t.d7 9 .t.f4. Canfell-Flitncy •

Canberra 1998. when 9 . . . a6 10 .t.h5+ Wd8 is uncomfortable for Black, if playable) 5 . . . lOxe5 6 lOds 1Wh4!? 7 dxe5! (7 lbxc7+ Wd8 8 IOxa8 .xc4+ 9 l.e3 f4 and Black will recover the a8-knight in due course. when he will have two pieces for a rook) 7 . . .• xc4+ 8 .t.e2 .t.b4+ 9 c3 .t.a.s 10 0-0 and White has a significant lead in develop­ment, Vinogradnik-Panchcnko. Ukraine ch. Scvastopol. 2000.

d) 4 . . . d6? 5 lOxc6! now that Black cannot recapture with the d-pawn, (5 1Wh5+!? g6 6 IOxg6 1Of6 7 1Wh4 IIg8 8 lOxfS IIg4 9 � 1lxc4+ 10 .i.e3 IOg4 is messy. Ornstein-Tall. simul. Malmo 1973. but neverthe­less. after 1 1 .xh7 White is better) 5 . . . bxc6 6 .t.d3 �f6 7 exf5 .i.e7 8 0-0 0-0 9 lOe3 Black is lost, K1ein­Soclter. Bad Meinbcrg 1986.

e) 4 . . . �f6?! 5 .i.c4?! (5 lOxc6 dxc6 6 e5 transposes to 4 lOxc6. and 5 exf5!? to 3 . . . /l)f6) 5 . . . lOxe5 6 dxeS IOxe4 7 0-0. Oll-Tasc RJO. The Hague 1997. 7 . . .• e7 with a reason­able position. Black cannot castle kingside. but can try the other side .

5 �f3! White returns the pawn for a lead

in development. S .t.d3?! fxc4 6 g3 1Wh3?!

(6 . . .• f6 7 .t.xc4 �xeS) 7 .t.xc4 �f6 8 lOxc6 dxc6. Melchor­Svcndscn, COrT. 1992. continuing 9 .O?! and now. instead of 9 . . . lOxc4?! 10 .xc4+ .i.c6 1 1 .i.gS ;t, 9 . . . .t.g4! 10 .d3 0-0-0 is strong.

S lOe3 .t.b4 6 �O is similar to the main line. excg>t that Black can cas­tle casilr: 6 . . .• xc4+ 7 .i.e2 .t.xc3+ 8 bxc3 /l)f6 9 0-0 0-0 10 lie I Wh8 1 1 .t.c4 .g4 12 h3 1Whs and White

3 . . . 1Oc6 and other third move alternatives for Black J J 3

has an edge, Leite-Pinto, Lisbon ch 1997.

5 •••• xe4+ 6 .te1 NCO evaluates this as +1-, and,

indeed, Black does have severe problems with his delayed develop­ment, and exposed queen.

6 .• .Ab4+ So that the white knight is denied

the c3-square. 6 . . .It�b4?! 7 1Oa3 IDds 8 /.[}bS 1-0,

Ruggeri Laderchi-<>Wens, corr. 1999, as, for instance, 8 . . . /.[}gf6 9 0-0 .tb4 10 c3 .te7 1 1 /.[}gS �4

i l2 g3 �6 13 1lle6. / 6 . . . dS 7 0-0 .td7 8 1llc3 We6 9

/.[}gS .f6 10 IllxdS and Black's position has fallen apart, Popovit­Kalinski, Zadar I 99S.

7 cl .te7 8 0-0 /.[}r6 9 c4 9 lIel .dS 10 c4 .t7 1 1 dS is

also promising. 9 ••• /.[}b4 9 . . . 0-0 10 1llc3 .g4 1 1 dS /.[}b4

12 /.[}gS will win material. 10 a3 /.[}.6 1 1 l[}cJ .e6 11 lIel

d6 13 c5 The black position is creaking,

Canfell-Van Mil, Utrecht 1988.

B 3 •••• e7?!

4 "'5+ Once again this is the critical re­

ply, although 4 d4 /.[}f6 (4 . . . 1llc6

transposes to 3 . . . 1llc6) S Ac4 is also strong, and less complicated for White. (S 1llc3?! offers White little, S . . . d6 6 Illc4 /.[}xe4! [6 . . . fxe4 7 .tgS] 7 1llxe4 .xe4+ 8 .l.e2 1llc6 9 c3, Wijker-Diepstraten, corr. 1982, when Black can try 9 .. .• xg2!? 10 .to � 1 1 We2+ Wet8) S . . . fxe4 ·(Black can also play S . . . d6 which is analysed in Chapter 9, e.g. 6 .tt7+! [6 fi)t7 dS with: 7 Illxh8 {or 7 .txdS?! IllxdS 8 /.[}xh8 Wxe4+ 9 .e2 .xe2+ 10 �e2 g6 1 1 lie I .tg7 1 2 Wfl + �f8 13 �xg6+ hxg6 14 c3 b6, Diepstraten-Elburg, corr. 1987} 7 . . . dxc4 8 eS IlldS 9 Whs+ g6 10 /.[}xg6 hxg6 1 1 Wxg6+ .t7 12 .xt7+ �t7 13 c3 .i.e6 and light-squared control, de Hoer­Elburg, corr. 1986) 6 .. . �d8 7 exf5? [for 7 .tb3! see Chapter 9] 7 . . . .txf5 8 o-o? dxeS 9 dxeS+ /.[}fd7 � Meszaros-Elburg, corr. 1988; S ... 1llc6 is considered in Chapter 9) 6 .tt7+t? (displacing the king, but wasting time, 6 0-0 dS 7 .tb3 .i.f5 8 c4 /.[}bd7 9 .tf4 slightly favours White, Rubel-Sorensen, Copenhagen 1879; 6 .tgS! c6? [6 . . . 1llc6] 7 1llc3 transposing to Chapter 12, i.e. 7 . . . dS 8 .txdS! win­ning, De Jong-Stamer, corr. 1994/94; 6 1llc3! also transposes to Chapter 1 2) 6 . . . �d8 . 7 Ab3 dS 8 .tgS c6 9 0-0 /.[}bd7 10 f4 exf3 1 1 /.[}xf3 h6 12 lIel Wt7 13 Ah4 .td6 14 IlleS? lllxeS I S dxeS .tcS+ 16 �hl gS 17 exf6 gxh4 18 c4?! h3 19 cxdS?? ( 19 g3) 19 . . . hxg2+ 20 �g2 .g6+ 2 1 Whl .tg4 0-1 , Podlesnik­Mohr, Ljubljana 1989.

4 ••• 16 5 �xg6 .xe4+ 6 .te1 White must resign himself to los­

ing his g-pawn, 6 Wd I ? /.[}f6 7 Wh4 (7 �3?, NN-Greco, Italy 1620, [what is a book without at least one reference to the seventeenth century?] 7 . . . /.[}g4 winning a piece)

J 14 3 . . . lOc6 a�d other third move alternatives for Black

7 . . .• g4+ (7 . . . lOg4!? may be even better, 8 .t.e2? [8 f3 bxg6 9 .xg4! is s"perior] 8 . . . bxg6 9 .xh8 lOxi2+ 10 Wel lOxhl 1 1 d3, Slous-Bone, London 183S, and now any sensible queen move like 1 1 .. .• e7 should be g� for Black) 8 f3?? (of course, 8 Wxg4 lOxg4 9 lOxh8 lOxf2+ had to be played) 8 . . .• xg6 � Jonkers­Aartsen, corr. 1 982.

6 • • • lOf6 7 W'h3 The g-pawn needs defence. 7 ••• bxg6 8 .xb8 .xgl 9 :0 W17 Better to let out of the pin imme-

diately, 9 . . . �.s4 10 d3 ( 10 .t.xg4 .xg4 1 1 WeS+, Liepins-Silavs, Latvia, is possible) 10 . . . d6? 1 1 .t.xg4 .xg4 1 2 .t.h6 with advantage, Littleboy-Gonsalves, corr. I990.

10 ..... I'm not sure that this is com­

pletely necessary, 10 d4 or 10 d3 are also sensible, aiming to complete White's development, 10 . . . lOc6 1 1 Wh4 dS 12 .g3?! .xg3 1 3 bxg3 lOd4 with active play for the lost exchange, Coriell-Hayward, corr. 1988.

10 ••• lOc6 1 1 cl b5! The best way to complete the

queenside development and bring the rook to the e-file. 1 1 . . . .t.d6 1 2 d4 lOd8?! 1 3 .t.gS .t.e7 was far too slow, Hegedus-Kurucsai, corr. 1984.

1l d4 .t.b7 13 .t.g5 1 3 lOd2 lte8 14 Wdl lOd8 I S

.g3 .t.d6! with reasonable play for the exchange, Parool-Prins, Holland 19S3.

13 • • • .t.g7 14 �d6 .t.d6 15 .g3 .xg3 16 bxg3 1le8 17 Wdl b4

Again with some compensation, although objectively not quite enough, Koetsier-Bertoni, corr. 1999.

C 3 ••• d6?

Apparently, this is an idea of Bronstein's, but seems to offer only disadvantages by comparison with 3 . . . lOc6.

4 "'5+ This is the only way to refute

3 . . . d6. 4 lOc4? fxe4 S lOc3?!I(S d3 is better) s . . . lOf6 6 d3 (6 .t.e2 dS 7 lOes d4! left White facing big prob­lems in Lehman-Diemer, Offenberg 1982, the game continuing 8 .t.c4!? dxc3 9 .t.f7+ We7 10 bxc3 .d6 1 1 d4 exd3 12 .t.f4 d2+! 1 3 WfI .t.e6! 14 lOg6+ bxg6 I S .t.xd6+ Wxf7 16 .t.xc7?! lOa6 17 .t.f4 .t.c4+ 18 Wgl 1Id8 -+.) 6 . . . dS 7 lOes .t.d6 8 d4 .e7 9 .t.gS c6 10 .t.xf6 .xf6 with advantage to Black, Landgraf­Kozlov, corr. 1991 .

4 ••• g6 5 lOxg6 lOf6 6 .....

3 . . . lOc6 and other third move alternatives for Black 1 15

6 •• .:cs Sacrificing the exchange is also

insufficient: a) 6 . . . hxg6?! 7 Wxh8 We7 8

.i.b5+!? c6 9 .te2 fxe4 10 0-0 .tfS 1 1 b and Black has little compensa­tion, Cordoba-Montero, COrT. 1 975.

b) 6 ... �6!? is interesting as· the: white queen sometimes has diffi- . culty finding shelter from Black's minor �ieces, but a rook is a rook! 7 lOxh8 IOJ4 8 lDa3 (8 Wdl !? fxe4 9 .i.e2 lOfS 10 .th5+ We7 1 1 Wh3 lOh6 12 We3 lOfS 1 3 .e2 .te6 14 JP lDd4 1 5 We3 lOfS 1 6 1Ih3 [ 16 .e I ! lOxh5 17 .tg5+ 1Of6 b .xe4 should win] 16 . . . lOh6 Ih-Ih (?), Eglitis-Morgado, COrT. 1979. White should certainly continue the fight; 8 .tc4!? lOxc2+ 9 Wdl lOxal 10 :el fxe4??, Viaggio-Morgado, COrT. 1978, when White missed the evi­dent 1 1 :xe4+ lDxe4 12 �5+ Wd7 1 3 WfS+ and mate next move) 8 . . . fxe4 9 .tc4? (9 d3 or 9 .83 are

far more testing) 9 . . . d5 (9 . . .... g4?! 10 lOf7 .e7 1 1 �6 +/­Destrebecq-Viaggo, COrT. 1979) 10 .txd5 .xd5 1 I Wxf6 .txa3 12 bxa3 lDx,,2+ 13 Wd I lDxa I 14 .tb2 ( 14 .xal .tg4+ is unclear) 14 . . .• xa2 1 5 .txa l Wbl+ 16 We2 .d3+ !h-'h, Downey-Destrebecq, cOrT. I996.

7 10xfl 7 e5?! is over-elaborate, 7 . . . dxe5!

8 lOxe5 We7 9 f4, Berg-Aartsen, COrT. 1982, 9 . . . �6 10 ... b5 .td7 and Black will castle long with a p0-tential attack on the g-file.

7".l:tg4 Or 7 . . . l:txfS 8 d3 �6 (8 . . . fxe4 is

no beller, 9 dxe4 We7 10 �3 with advantage, Houten-de Neef, COrT. 1982) 9 �3!? (9 .tg5) 9 . . . lOd4 10 .i.g5! lDxc2+ 1 1 �dl .te6?? ( 1 l . . .lOxal 12 lDd5 Wd7 13 lDxf6+ l:txf6 14 .txf6 Wa4+ 15 b3 is also

hopeless, but at least White has something to think about) 12 Wxc2 1-0, Morgado-Gabrans, COrT. 1976.

8 "'61lxe4+ In Bullockus-Grobe, COrT.

1973n5, 8 . . .• e7 was pferred with the follow-up 9 f3 :g7 10 lDxh7 l:txh7 1 1 .e3?! (but 1 1 Wg6+ Wd8 1 2 d3 seems more convincing) 1 1 . . . fxe4 with slight compensation.

9 .te1 And certainly not 9 Wdl ?? lOg4 . 9 .. .lIe7 Without the knight on c6, there is

nothing for Black here; 9 . . .• e7 10 �3 and 0-0 with advantage.

10 0-0 1-0 As 10,..he1 l l lDc3 :e5 11 d4 .. .Ieaves Black material down with

a bad position, Sadeghi-SCn«haud, COrT. 1991/92.

D 3,..1Of6?!

Lowenthal's move. This time there is no need for Black to worry about Wh5+, but on the other hand, he loses a whole pawn without embroiling White in complex calculations. , . 4 ed5 -V

Simple and good. 4 .tc4!? .e7 5 d4 transposes into B, whilst 4 lDc4? fxe4 5 1&3 d5 6 lDe5 d4 (6 . . .• e7 -+) 7 .tc4 dxc3 8 bxc3, Sinke-

1 16 3 . . . 1tX6 and other third move alternatives/or Black

Elburg, COlT. 1987, 8 . . .• e7 and White can almost resign. 4 �3? d6 5 � fxe4 transposes into C.

4 .... e7 4 . . . d6 5 lOn Axf5 6 d4 Ae7 7

Ad3 Ag4 8 c3 is similar; Black plays a position resembling the French Exchange, without his f­pawn, Budovskis-Gabrans, COlT. 1969.

4 . . . �6!? goes for quick develop­ment; 5 d4!? (5 lLlxc6 dxc6 6 .n, Gelman-Schuyler, Concord 1995, and now 6 . . .• d5! and Black will recover one of the pawns, with some pJay) 5 . . .• e7 6 .A.e2 lLlxe5 7 dxe5 Wxe5 8 0-0 Ad6 9 g3 0-0 (9 . . .• xf5 10 Ad3 �3 1 1 lIel+ i..e7 12 �3 lOg4? 13 lIxe7+! is devastating) 10 Af4! .xb2 1 1 .A.c4+ �h8 12 Axd6 cxd6 1 3 .xd6 lIe8 14 lLla3! IOc4 ( 14 . . . b6 is no better, I S Afl 1Id8 16 � .xc2 17 lLle5 .c5 1 8 .xc5 bxc5 19 Ac4 lIiS 20 lOfl+ �g8 2 1 lLld6+ �h8 22 IIfe l as [22 . . . lIb8?! 23 llabl 1Ib6 24 IIxb6 axb6 25 a4 is almost Zugzwang] 23 lieS) d 5 .d5 .xa3 16 llae 1 .f8 17 1Ixe4 1Ixe4 18 .xe4 d5 19 Axd5 Axf5 20 .e3 • c8 2 1 lie I Ad7 22 .d4! 1 -0,

Kosten-Tiemann, email 2001 , as 22 . . . .i.c6 23 1le7· .iS 24 IIfl _g8 25 .i.b3 1le8 26 f4 is completely crushing.

S .el 5 d4?! sacrifices a piece, 5 ... d6 6

.e2 (6 f4!? dxe5 7 fxe5 Axf5 is very speculative, Punt-Elburg, COlT. 1986) 6 .. . dxe5 (6 . . . Axf5?! 7 00 lLlc6 8 c3 0-0-0 allows White to es­

cape the loss of the piece, Rogalski­Diepstraten, COlT. 1985) 7 dxe5 lLlds 8 "'5+ .fl 9 .xf7+ �f7 10 Ac4 c6 1 1 �3?! ( 1 1 e6+) I I .. .Ab4 12 Ad2 Axc3 1 3 AxdS+ cxdS 14 e6+ �ft) I S Axc3+ �f5. White's p�d advanced pawns have disintegrated, Kroonen-Diepstraten, Amsterdam 1989.

S ••• d6 6 �4 6 lOo is perfecdy reasonable, as

well; 6 . . . Axf5 7 lLld4 Ad7 8 �3 �6 9 lLlxc6 Axc6 10 .xe7+ .i.xe7 1 1 .i.b5 and Black has no real com­pensation for his pawn, Muller­Diepstraten, COlT. 1985.

6 •• ..i.xf5 7 .xe7+ Axe7 8 lOe3 Ae6 9 d4 � 10 cl

With advantage, Tener-Denny, COlT. 1972 .

7 3 i.c4 fxe4 4 lbxeS 1i'gS (Poisoned g2 pawn variation)

This is probably one of the sharp­est and most extensively analysed opening variations of all . I suppose for this reason alone, the practically }Ilinded player may consider that it 'is not worthwhile making the �ffort to memorize the complicated varia­tions involved. In addition to that, Black often finds himself subjected to a venomous attack. On the other hand, a variation that might be dubi­ous in correspondence chess (nowa­days players often give preference to 4 . . . dS, chapter 8) may well gain untold 'gift' points ·ove .... the­board'. Black would be most unfo .... tunate to find himself playing against an opponent versed in the theory, and finding a way through the maze of variations with only two hours (orliess) on the clock? Forget � -

1 e4 e5 1 �f3 IS 3 .i.c4

Aiming straight down the a2-g8 diagonal, weakened by Black's

second move. This move was the favourite of Paul Keres, and is one that appeals to players of the white pieces who are intent on winning rapidly, and who are not afraid of giving up a few pieces to force mate. The slightest inaccuracy from White, however, and he may rand himself material behind with no compensation.

3 ... fxe4

4 �se5 Neither 4 .i.xg8?! Axg8 (4 . . . exf3

S .i.dS c6 6 .i.xf3 dS 7 0-0 .i.d6, Oren-Diepstraten, con. 1992, is not bad either) S �eS WgS 6 �g4 dS 7 h3 IOc6 8 d4 -86 9 .i.f4 .i.d6 10 .i.xd6 exd6 1 1 1&3 .i.e6 12 ed2 0-0-0 � Fischer-Leisebein, East Ger­many 1987, nor 4 �gl? �a; S 1&3 lOc6 6 0 dS :J: Druke-Heap, con. 1991 , are serious alternatives. Fur­ther, 4 d4?, a speciality of Breiden­bach in the early 80s, is guite unsound, 4 .. . exf3 S dxeS (S .xO ea; 6 Whs+ g6 7 exeS+ .xeS+ 8

J J 8 3 "'c4 fxe4 4 �e5 .g5 (Poisoned g2 Pawn Jlar;al;on)

dxeS liX:6 is completely hopeless, Dogiel-LOffier, corr. 199 1 ) S . . . fxgl 6 lIgl dS (probably the simplest) 7 -'.xdS .i.b4+! (of course not the im­mediate 7 . . . 1Oe7?? 8 "'17+) 8 c3 1Oe7 9 .a4+, any bishop retreat allows the exchange of queens, 9 . . . "'d7 10 .i.f7+ Wxf7 1 1 .xb4 .i.c6 12 .f4+ �e8 1 3 .g4 .d3 14 -'.e3 hS! IS .xg7 lIg8 16 �6 "'0 0- 1 , Breidenbach-Leeners, corr. 1982.

4 •••• g5 This was first given by Bilguer as

the strongest answer against 3 "'c4, Black simultaneously counters White's projected 'A'hS+ and threat­ens the important gl pawn. In the next chapter we consider the key al­ternative, Svedenborg's 4 . . . dS! , which is likewise not without its problems for both players.

Whilst Blackburne's 4 . . . �f6?! also sees the light of day occasion­ally, often with positive results for the second player: S �f7 (Critical, although White has a choice: S .i.f7+ �e7 6 "'b3 dS is simpler, and causes Black a certain amount of in­convenience, 7 d4 cS !? 8 "'gS �6 9 'A'hS?! .e8 10 'A'h4?! cxd4 1 1 .i.xdS �xeS 12 .xe4 �d6 1 3 "'xf6 gxf6 14 f4 .i.fS I S fxeS+ .xeS 16 WxeS+ fxeS and Black has reached a very favourable endgame, not

least because of his active king! Gamman-Blackbume, London 1 869; but S d4?! dS 6 .i.e2 [6 "'b3! transposes to Chapter 1 1 ] 6 . . . "'d6, Hughen-ortiz, corr. 1968, is too in­sipid; however S �3 is also good, see Chapter 12) S . . .• e7 6 �h8 dS 7 "'e2 (the alert reader may have noticed that Black is a rook down, but, bearing in mind that the knight on h8 will inevitably be losl, this should only be a temporary state of affairs. Meanwhile, Black has a good development and some potent threats, 7 "'xdS? for instance, being met by 7 . . . "'g4 8 "'xb7! [8 .xg4? �xg4 9 "'xb7, Strelis-Atars, corr. 1977, is worse, 9 . . . � 10 g3 .f6 1 1 0-0 "'cS and White cannot even capture the a8-rook, 12 "'xa8? .0 winning; 8 O?! exO+ 9 �f2 .cS+ 10 d4 WxdS 1 1 �3 'A'hs 12 gxO 'A'h4+ and a strong attack, WolI­Amilibia, corr. 1974177] 8 . . .• eS!? 9 �3 "'xd I 10 �xd I �S 1 1 "'xa8 c6 12 c3?! , Purins-Amilibia, corr. 197 1 , when the obvious 12 . . . �b6 13 "'b7 .e7 is strong, Black wins the b7-bishop, and later the h8-knight. i.e. 14 0-0 gS I S lie I "'g7 16 d3 .xb7 17 l1xe4+ Wf8 1 8 "'xgS "'xh8 19 lIe7 �8d7 20 lIxh7 "'g7, unclear; 7 -'.b3?? "'g4 8 0 exO+ 9 �f2 �+ I 0 �e3 .gS+ 1 1 �d3 �f2+ 12 �c3 d4+ 13 �xd4 .cS mate, Carapelli-Young, Greece 1 874; 7 "'bS+!? is a suggestion of Lein, after 7 . . . c6 8 "'e2 play is simi­lar to the main line, except that Black can no longer play his knight to c6) 7 . . . �6 with the following possibilities:

a) 8 d4 exd3 9 cxd3 -'.84?! (9 . . . "'f5) 10 �3 0-0-0 1 1 0-0 .i.e6 12 "'f4?! d4 13 �bS �S 14 .i.g3 a6, Gamman-Blackburne, London 1869, White certainly stands better here, but went on to lose.

3 �c4 fxe4 4 �e5 .g5 (poisoned g2 Pawn Variation) / / 9

b) 8 �hS+? is popular, aiming to get a couple of pawns in return for the knight, but it is very time con­suming: 8 . . . g6 9 ltlxg6 (9 �xg6+? hxg6 10 IOxg6 "g7 1 1 lOil4? [bet­ter 1 1 IOf4, but I I . . .�d6! 1 2 d3 �xf4 1 3 �xf4 "xg2 14 Itfl �g4 is also promising for Black] Gemi.snani-Pastor, corr. 197 1172, I I . . .Wh7! 1 2 g3 �g4 1 3 0 exO 14 ltlxo "e7+ wins) 9 . . . hxg6 10 �xg6+ �d8 1 1 �hS IOeS!? ( l 1 . . ..g7!?) 1 2 o-O?! (the king is none too safe here, 1 2 �e2) 12 . . . Wh7 1 3 �e2? ( 1 3 d4! is neces­'sary, 1 3 . . . lOeg4!? [ 1 3 . . . exd3 14 �gS] 14 �xg4 �xg4 IS "d2! and White has defensive chances, 1\ O?! .i.d6) 1 3 . . . �d6 14 &3 ( 14 h3 .i.xh3) 14 . . . Wh3 I S 0 lDeg4! 16 fxg4 .i.xg3 17 1tf2 .i.xf2+ 1 8 �f2 "xh2+ 19 �e3 "g3+. chasing the king up the board to its demise, Morillo-Ortiz, corr. 1970.

c) 8 0-0 �f3 9 d4 0-0-0 10 c3 .e6 I I .i.f4 �e7 1 2 b4 Itxh8 1 3 a4 gS !? Objectively White is better, but Black has practical chances, Keast-Borrmann, corr. 1974.

d) 8 d3 �f3 9 dxe4! IOxe4?! 10 .xdS!? (greed worthy of a com­puter! 10 0-0 0-0-0 I I �g4 "ffi 12 �xf3+ "xf3 1 3 "O?! is a simpler method, Verolme-Ouwerkerk, corr. 1978, 1 3 . . ... xO 14 gxO 1Od6 I� ltk3 and White should win) 1O . . . ltd8! this is not sound, but is the best prospect, 1 1 "xf3!? ( 1 1 "f7+ is more circumspect, I I . . . .. xf7 12 IOxf7 �f7 1 3 '*.e3 1Od4 14 �d3 �cS I S 0-0 Black is lost, Borrmann-LOmer, corr. 1991) I l . . .i0d4 1 2 �hS+! ( 1 2 Whs+?! g6 13 ltlxg6? [ 1 3 �bS+ is the only possibility, 13 . . . c6 14 IOxg6 ltlxc2+ I S �e2 ltk3+ { l s . . . IOg3+ 16 �O} 16 �O 1Od4+ looks very nasty for White, who is forced to wander

along the wrong side of his pawns, but may be playable] 1 3 . . . lOxc2+ 14 � Itdl+! A superb refutation! IS -*.xdl IOg3+ 0- 1 , as it is mate next move, Petersson-ortiz, corr. 1970; 1 2 -.f7+? -.xf7 1 3 ltlxf7 IOxc2+ 14 � �f7 IS O IOffi 16 ltk3 ltlxa I 1 7 -*.f4, Madrid-Ribe, corr. 1983, is quite equal after 1 7 . . . c6) 12 . . . g6 1 3 .bg6+ hxg6 14 -.xg6+ �d7 IS 0-0 1Oe2+ 16 �h I should be winning, nevertheless, in Madrid-Ribe, corr. 1984, there followed: 16 . . . Wh4 17 IOf7?? (White has a number of good defences, for instance 17 .13+ �c6 1 8 -.0 [ 1 8 IOg6? 102g3+ 1 or 1 7 ltk3) 1 7 . . . ltlxf2+ 1 8 1txf2 .xf2 19 -.d3+ -*.d6 20 -.d I (20 Wh3+ �e8 2 1 ltlxd6+ Itxd6 forces White to take a perpetual check) 20 . . . b 21 1Oc3 and now 2 1 . . .IOg3+! 22 hxg3 .xg3 23 -*.f4 -.xf4 24 Whs 1txf7 is at least equal.

5 d4 Almost forced, as the two

• obvious' alternatives both give White a hard time:

a) S IOf7?! "xg2 6 Itfl dS! and: al ) 7 ltlxh8 dxc4 (7 . . . 1Of6?! 8

-*.xdS! [8 �e2 -*.h3 9 d4 .xh2 10 -*.e3, Pemberthy-Pierce, corr. 1915, 1O • . • g6 1 1 -.d2 -*.xfl 12 -*.xfl -*.g7 is good for Black] 8 . . . -*.g4 [8 ... -*.h3 is considered below] 9 0 .txO 10 Itxo exO 1 1 -.xO [ 1 1 -*.xO?! -'gl+ 12 �e2 -.xh2+ 13 �fl 1Oc6

120 3 �c4 fxe4 4 �e5 .g5 (Poisoned g2 Pawn Variation)

14 d4 (H)-() IS c3 �S is VCI)' dan­gerous for White, NN-Holwell. corr. 1986) 1 1 . . .• g l+ 1 2 �e2 .xh2+ 1 3 �d3? [ 1 3 �fl ! c6 14 �b3 ltlbd7 IS .e2+ has to be good for White] 1 3 . . . ltlbd7 14 ltlf7 ltlxdS I S .xclS .g3+ 16 �e2 .g4+ 17 �el �e7 with a powerful attack, Vossebeld­Olof, LeusdenlUtrecht 1987) 8 ebS+ g6 9 .xh7 �h3! 10 .f7+ �d8 I 1 .xf8+ �d7 12 .g7+ (forc­ing the black king to impede the rook, 12 .f7+?! �7 1 3 .xc4 ltlbc6. the knight wishes to come to the O-square, via eS. 14 ltlf7 1If8 IS ltlh6 �S � Kirchner-Hacker, Switzerland I96S) 1 2 . . . �7 13 .d4+ �c8 14 .xc4 ltlbc6 I S d4 (but now IS ltlf7? is weaker, as IS . . . ltld4 is possible, in the other line White could cajlture this with check) I S . . . bS ! 16 _xbS lilxd4 17 • a6+ �d7 18 lilxg6 lilxg6 19 .a4+ �e7 20 lild2 It)0+ and the white king did not last too long, Rosso-Ruggeri Laderchi, corr. 1998.

a2) 7 AxdS lilf6 (the most popu­lar move, by far, but maybe not best, 7 . . . Ag4! 8 0 Ae7 9 .e2, forced, 9 . . . �h4+ 10 �dl and now 1O . . .• xe2+ [ 1O . . .• xfl+? is flash, but faulty, 1 1 .xfl �xO+ 12 .xO exO 13 lilxh8 f2 14 �e2 1Oc6 I S Axc6+ bxc6 16 d3 Perez-Vidal, Spain 1997, and even after best play, 16 . . . �d7 17 Ae3 lilh6 1 8 lild2 J:lxh8 19 lilO Black is worse be­cause of his shabby pawn SbuCture) 1 1 �e2 exO+ 1 2 J:lx0 ltlffi 13 Axb7 �f7 14 Axa8 l:le8+ I S �d3 ltlbd7 and Black has an enormous lead in development and must win; 7 . . . Ah3!? 8 .e2 lilffi 9 Axb7 �f7 10 d3?? [White has near equality af­ter 10 �xa8 �d6) 10 . . . exd3 1 1 �xg2 dxe2 1 2 �xh3 exfl=.+ 1 3 Axfl Ironically, it is White who finds himself a rook light!

Marshall-Steen. corr. I99S) 8 ltlxh8 �h3 (8 . . . ltlxclS? 9 ebS+; 8 . . . �g4 transposes to 7 . . . ltlffi) 9 �f7+ (9 �c4? 1Oc6 10 c3 �S 1 1 d4 (H)-() 12 .l.e2? lilO+ 1 3 �xO .xfl+ 14 �d2 .xt2+ with a dec:isive attack in Gudju-Betim. Paris 1924) 9 . . . �e7 10 �c4 (10 .e2? 1Oc6 1 1 c3 �S 12 �c4 ltlO+ with the at­tack. Ziedens-Vitols. Latvia) 1O . . . lOc6 1 1 d4 lId8 12 �e3 !? ( 12 b3 ! looks promising, e.g. 12 . . . ltlxd4 13 .i.a3+ �e8 14 �f7+ �d7 IS .xd4+ �c8 16 .c4) 12 . . . �S 1 3 dxeS!? J:lxdl+ 14 �dl ltlg4 I S lie I ltlxe3+ 16 fxe3 .xh2 and White's awkward king position and trapped knight make Black favour­ite, Crimp-Melchor. corr. training game 1997.

b) S �f7+?! �e7 6 ebS!? (6 d4 .xg2 7 :n [7 ebS? .xhl + 8 �e2 lilffi � Thielen-Varga. corr . 197 1/721 7 . . . d6 8 �xg8 [8 �S? dxeS 9 gS+ ltlffi 10 dxeS �g4 1 1 exffi+ gxffi 12 �xffi+ �ffi 13 .clS 1Oc6 and White is material down with a bad position, NN-Sorensen, 1879) 8 . . . J:lxg8 9 1Oc4 Ah3 10 �3, Waller-Quinones. National Open, USA 1996. when 10 . . .• 0 would be simplest, with a pleasant ending; 6 �xg8 .xg2!? [6 . . . J:lxg8 7 ltlg4 clS is reasonable] 7 IIfl J:lxg8 8 -..S. Sanpera-Melchor, La Pobla de Lillet 1997. 8 . . . g6 9 .xh7+ IIg7 10 eb4+ �e8 unclear) 6 . . .• xg2!? (6 . . .• xhS 7 �xhS. Alias-Martin Echeandia, Euskadi 2000, 7 . . . ltlffi 8 �e2 d6 avoids any complications) 7 bg8 (7 IIfl ltlffi is terrible. and 7 �e8? .xhl + 8 �e2 lilffi 9 .f7+ �d8 10 1Oc3, Probst-Lowig, Oeynhausen 1922. 1O . . . ltlc6 loses vast amounts of material) 7 . . .• xhl + 8 �2 J:lxg8 9 1Oc3! g6 (9 . . . 1Oc6!) 10 00+ �e8 I 1 .gS �d6 1 2 ltlffi+ the queen and two knights give the black king

3 �c4 fxe4 4 �e5 .g5 (Poisoned g2 Pawn Variation) J 2 J

a hard time, Rodholm-Hemtrom, corr. I996.

5 ...• xgl

6 Wb5+ The only corre<:t plan, and one

that will involve White in the sacri­fice of half of his army.

a) The timid alternative 6 :fI ? is perhaps the sort of reac�ion one might expect in an 'om' game, 6 .. :ltlffi 7 lilc3 (7 �f7+ We7 S �h5?! g6 9 .i.g4 �g4 10 .xg4 .xg4 1 1 �xg4 of Broquen-Purins, corr. I96SI70, is also to Black's ad­vantage� 7 �f7? d5 S �xhS dxc4 9 .i.f4 .i.g4 to .c I , Pruess-Vinerts, USA 1996, and now Black is in p0-sition to ",in the game, without even botherint:� pick up the hS-knight, by 10. . . ! 1 1 .i.e3 .i.h3 1 2 00 �e7 and White has no defence against . . . �e3, and then . . . �h4+) 7 . . . .i.b4 S �f4 d6 9 �f7 �g4 (9 . . . :f1U to �g5 d5 1 1 .i.e2 h6 wins a piece) 10 .d2 :f8 1 1 13 .xd2+ 12 i.xd2 �xc3 1 3 �xc3 �xf3 14 �g5, Arasola-Susi, Naantali 1993, and with two extra pawns, Black should certainly be able to score the full point.

b) 6 .i.fI? .xhl 7 �5+ g6 S �g6 bxg6 9 .xhS Wf7 10 �e3 d5! ( 10 . . . Ag7 1 1 � �ffi?! led to a curious repetition in Douwes­Tissink, Utrecht 1991 : 1 2 � dS

1 3 0-0-0 �g4 14 �e2 .g2 15 .i.fI �I Ih-Ih) 1 1 � .i.g4 just in time to stC>p White castling, 1 2 .eS �ffi 1 3 .xc7+ �7 14 .i.f4 :eS I S .xb7 e3 16 fxe3 .gl 17 �3 �d6! 0-1 , Milnes-Mlotkowski, Philadelphia 19 1 1 .

6 ••• 16 7 �f7+

7 ••• Wd8 The black king must try to make

himself scarce. 7 .. .r�e7? loses, S �g5+! (S .gS+ is also strong S . . .• xgS 9 hgS+ �f6 to lilc3 c6 I 1 �c4 �g7 1 2 .i.xg6! [ 1 2 .i.b3 dS 13 �ffi hffi 14 f4 is nothing special for White, Hayes-Downey, corr. 1991 ] 1 2 . . . bx� 1 3 �g6+ We6 [ 1 3 . . . �!? 14 �hS+ .i.xhS IS 1lXl6+ We6 16 �cS �6 17 �a7 :Xa7 certainly favours White, but Black has some chances] 14

, �xffi .i.xffi 15 �ffi [15 �xhS?! �xhS 16 ().()-() d5 1 7 �g5+ Wffi Malmstr6m-Downey, corr. 1992, is less clear] IS . . . :b6 [there is little choice, I S . . . Wxffi 16 �hS Wg7 17 Wd2 WxhS I S :ae I will win a piece] 16 � �6 17 :gl and Black soon had to resign, Hage­Svendsen, e-mail 1999) s . . . lOf6 9 � .xhl+ 10 Wd2 (or to We2 �g7 when White should avoid 1 1 lilc3? when, as Grivainis and Elburg point out, 1 1 . . .• f3+! is possible, e.g. 1 2 �f3 [ 12 Wel? should be

J 22 3 Ac4 fte4 4 �e5 .g5 (Poisoned g2 Pawn Variation)

answered by 1 2 . . .• 15 {rather than 12 . . . �1+? which gives White the opportunity to change his mind by 1 3 �d2 transposins.. into 10 �d2! } 1 3 IOIs+ �f8 14 l&ffi IOc6 with good chances] 1 2 . . . exO+ 13 �d2 [superior to 1 3 �O?! �17 14 .f4 d6 I S IOIs Af5 16 Axffi Axffi 17 fuc7 1&6 I S lOxaS fud4+ Jackson-Oowney, corr. 1992, the three pieces being more than a match for the queen] 1 3 .. . �17 14 _f4 d6 IS IOds Af5 [rather than I S . . . �7? 16 J:lel ! and Black is lost] 16 Axffi Axffi 17 IOxc7 1&6 I S fuaS J:lxaS and both sides have chances, Vadot-Sireta, corr. 1994. Instead of all this, 1 1 1Ol2! in­creases control of 0, and avoids the black defence, i.e. 1 I . . .• g2 [if l l . . ..xal ? 12 lDxe4 and mate will be swift] 12 fue4 �f8 13 lDxffi 1&6 14 AdS etc.) 10 . . . e3+ (trying to confuse matters, 1O . . . Ag7? 1 1 1Dc3 .xa l 12 Axffi+ Axffi 1 3 ltldS+ gives a quick mate) 1 1 �e2 (capturing with the pawn leaves Black an important lateral check: 1 1 fxe3? .g2+ 1 2 �d3 .fI+ 1 3 �d2 1Dc6 14 �3 .g2+ I S �d3 lDxeS+ 16 dxeS �17 1 7 Axffi gS I S "ehS+ �gS 19 AxhS b6! 20 Affi Aa6+ and, remarkably, Black wins by direct attack, Elburg-Oowney, corr. 1991 , but 1 1 �e3 also wins, I I .. ..i.h6 [ I I ...Ag7 1 2 1&3] 1 2 1&3 .xal 1 3 .xh6 .el+ 14 �O "ehl+ IS �e2 1&6 16 .g7 fud4+ 17 �d2 and the attack decides, Pape-Melchor, corr. 1994) I I . . .Ag7 1 2 1&3 _g2 1 3 lOe4! (White must protect f2, 1 3 IOIs+? � 14 Axffi [ 14 fue3? e4 IS f4 d6 -/+ Alloin-

\..E?wney, corr. 1991] 14 . . . Axffi IS ... xffi .xf2+ 16 .xf2 exf2 1 7 lDxc7 d6 I S AdS dxeS 19 dxeS �g7 20 �xf2 lidS 2 1 c4 1&6 22 lDXaS lDxeS 23 1&7 and the game

is quite level, Stdmmer-Sireta, corr. 1994/9S) 13 . . . �f8 14 Axffi gS I S �S! .xe4 (if I S . . . exf2 16 lOxf2 1&6 17 .*.xg7+ �g7 IS AdS! wins) 16 Axg7+ �g7 17 -x.gS+ � I S .dS+ �g7 19 �gl + _g6 20 Jlxg6+ hxg6 2 1 WsS 1 -0,

. Kozlov-Svendsen, corr. 1991 .

8 .*.xg6! Boldly leaving the h I rook to its

fate, and with check too! The alter­natives, though, are either insipid or losing:

a) S .gS+?! .xgS 9 .*.xgS+ Ae7 10 .*.b3 (neither 10 .*.f4?! d6 1 1 lDc4 lDffi -/+ Hughen-Dreibergs, corr. I96S, nor 10 h4?! d6 1 1 lik4 1&6 ; Neukirch-Dukurs, corr. 1965, offer any compensation for the g2 pawn, whilst 10 .*.ds .*.xgS 1 I 1017+ �e7 12 lDxgS lDffi 1 3 1&3 c6 14 Axe4 lDxe4 IS lDgxe4 dS, Grob-Tartakower, Meran 1926, favours Black because of the pawn structure) 1O ... .*.xgS 1 1 lDn+ �e7 1 2 fugS ( 1 2 lDxhS?! lDh6 [ 1 2 . . . lDffi!? is interesting; 13 fi)n?, Thorn-De Snaijer. corr. 19S2, al­lows 13 . . . .*.c 1 with advantage. and the superior 13 1&3 Ah4 14 0-0 c6 keeps the knight in the cage, whilst preparing the defence of the e4-pawn] 13 J:l.s1 [ 1 3 1&3?! c6 14 lDxe4 Affi 1 S l&ffi {there is no es­cape for the knight: I S lDn?! lDxn

3 s.c4 fxe4 4 �e5 .g5 (poisoned g2 Pawn Variation) J 23

16 s.xf7 �f7 17 �6+ �e6 I S lOxcs lOa6 White is lost) I S . . . �f6 16 �d2 dS 17 Zlael s.e6 I S lIe2 { I S J:txe6+ �e6 19 lIe1+ �f6 20 lieS still doesn't save the knight. 20 .. . bS 2 1 a4 a6 22 c3 �g7 23 .i.c2 lbgS) I S . . . lOa6 19 lIhel lbc7 and the knight is finally captured, when Black's two pieces should make short work of White's rook and fee­ble pawn, Zirnis-Vitols, Latvia 1972] 1 3 . . . s.f6 14 lbxg6+ hxg6 I S lIxg6 lbfS 16 c3, Zirnis-Gubats, corr. 1972, 16 . . . c6 and . . . dS, with advantage) 1 2 . . . ffi 1 3 lbc3 c6 ( 1 3 . . . lIdS! is a more accurate way to prepare .. . dS: 14 lbgxe4 [ 14 dS d6 I S 0-0-0 s.fS 16 lIhel lbbd7 is fine for Black, as White has obstructed his bishop] 14 . . . lOxe4 I S lOxe4 dS with good prospects) 14 dS! (White prevents . . . dS; 14 lbgxe4?!, Birkenfelds<ensonis, corr. 1976, 14 . . . lOxe4 IS lOxe4 dS with an edge) 14 . . . d6 I S o-o-o s.fS 16 lIhel cS 17 lbcxe4 and White has broken through on the e-file before Black can finish his queenside develop­ment, Weijers - Van de Casteele, Haarlem 1996.

b) S �gS+? i..e7 9 s.xg6 (9 i..xe7+? l/:Jxe7 10 � .xh1+ 1 1 �d2 [ 1 1 �e2 d6 12 lbc3 s.g4+ 13 .xg4 .xa l ] l l . . .e3+, Gillis­Pupols, USA 1966, is even worse) 9 . . . s.xgS! (9 . . .• xh 1+? loses, 10 �e2 c6 I I lbf7+ �c7 1 2 lbd2! [ 1 2 s.xe7 .0+ is less clear-cut, Leiva-Amilibia, corr. 1970] 1 2 . . .• xal 1 3 s.xe7 b6 14 s.dS+ �b7 I S �6+ �a6 16 s.xe4 forcing mate, Hazlett-Nakamura, Hawaii ch 1990) 10 lbf7+ �e7 1 1 lOxgS lbf6 12 Wh6 lIgS!? ( 12 . . .• xh 1+ 13 �e2 lIgS 14 s.xh7 llxgS I S .xgS .0+ should also be sufficient) 1 3 lbc3 c6! (safety first. although 13 . . . J:txg6 14 �S+ lbxdS

I S .xh7+ �f6 16 .f7+ �gS 17 0-0-0 �h6 IS .f8+ �hS is_proba­bly winning. too) 14 lOxh7 Wxg6?! ( I4 . . .• xh l+! I S �d2 .g2 16 l&f6 .xt2+ is clear) 1 S lOxf6 .xf6 beat­ing off the attack, and remaining a solid piece to the good, for a couple of pawns, Crivosh-Apsenieks, Hun­gary 1925.

8 ••.• xbl+ There is little to be gained from

not taking the rook, as the alterna­tives all lose! The analysis of the other moves, some of which ema­nates from Keres, is as follows:

a) S . . . hxg6? 9 s.gS+ i..e7 10 .xhS .xh I + 1 1 �e2 i..xgS 12 .xgS+ �e7 13 .17+ �d6 14 .xg6+ 1-0, Heap-Druke, corr. 1991 .

b) S . . . d6? 9 lbf7+! (9 .gS+ .xgS , 10 s.xgS+ s.e7 1 1 �f7+ �eS 12

IOxhS+ [ 12 s.xe7? lOxe7 13 lOxhS+ hxg6 14 lbd2 dS I S h4 s.g4 keeps the knight encaged, Scott­Bullockus, corr. 1970] 12 .. . hxg6 1 3 lIgl s.fS 14 s.e3 extracts the knight and wins the exchange) 9 .. . �e7 (9 . . . �eS? 10 IOxd6+ 1 -0, Robins­Bullockus, corr. 1970) 10 s.gS+ lbf6 1 1 lbxhS .xh 1+ 12 �d2 e3+ 13 �e2 gives White his standard at­tack but without being a rook down!

c) S . . . s.b4+? Tartakower's idea, but 9 �e2! .xhl 10 .i.gS+ s.e7 I 1

J 24 3 �c4 fte4 4 1Oxe5 .g5 (Poisoned g2 Pawn Variation)

\

�n+ �eS 1 2 itWtS+ is devastat­ing.

d) S . . . �ffi? 9 �+ �eS 10 itWtS+ ( 10 .eS+ �e7 1 1 lOd6+ �dS 12 he4 .g7 1 3 �h6 is also convincing) 10 . . . bxg6 1 1 .xg6+ exchanging queens, with advantage.

e) S . . . �g1? 9 �f7+ �e7 10 .cS+! ( 10 �gS+? changes the entire situation, 10 . . . Wf8 1 1 �hS .xhl+ 12 �d2 .O! 1 3 .xh7? .xf2+ 14 �c3 .xd4+ 0-1 , Tallman-Smiley, COrT. 1984, excep­tionally, it is Black who gives mate!) 1O . . . �ffi 1 1 .5+ �e7 1 2 �gS+ �f6 1 3 .eS+ �fS 14 �xffi is crushing.

9 �el

Now we have A 9 •••• xcl B 9 ••• c6 The alternatives are hopeless: a) 9 . . . d6? 10 lOc3! (threatening

�gS+) 1O . . .• g2 (l0 . . . �e7 1 1 �f7+ �eS 1 2 �xe4) 1 1 �xe4 .g7 12 �n+ �d7 1 3 �f5+ �eS 14 1Oxd6+ 1-0, Morgado-Toro, COrT. I96S.

b) 9 ... bxg6? 10 �gS+ �e7 1 1 .xhS �xgS 1 2 .xgS+ losing ignominiously.

c) 9 . . . e3? 10 �f7+ �eS 1 1 �hS+ bxg6 1 2 .xg6+ �dS 1 3 .xgS c6 14 .xf8+ 1-0, Purcell­Summons, COrT. 1997.

d) 9 . . . �g7? may be the best of the bunch, but even· so, I 0 �gS+ ( 10 �f7+?! �e7 1 1 .cS+ �eS 1 2 �hS?! [ 1 2 lOd6+ draws] 12 . . .• 0+! 13 �xO exO+ 14 �O �f7 IS .xc7 lOc6 16 c3 �ffi 17 �gS )IfS IS �g2 �gS and Black's pieces should outgun the queen, Stummer­Hunstock, COrT. 19S9) 1O . . . �ffi ( 10 . . . �ffi loses, but White has to be careful: 1 1 iOd2! [ 1 1 lOc3 is also winning; but on 1 1 �xffi+?! �ffi 12 �f7+ �7 13 .eS+ { 1 3 .cS+ d6 14 .xc7+ �fS IS .xcS+ �g7 is far from clear} 1 3 . . . Wf8 14 lDd2 { 14 �gS �g7 I S �xe4 lOc6! } 14 . . .• g2 IS �hS there is I S . . .• g4+ 16 � .e6 beating off the attack; 1 1 �f7+?! �e7 1 2 �xffi+ �ffi trans�es to this last line] 1 1 .. ..g2 [ 1 1 . . ..xal 12 �xffi+ �xffi 1 3 �n+ �e7 14 .eS+ Wf8 I S .xffi leads to mate] 12 �f7+ �e7 1 3 �S .xgS 14 .xh7+ winning.) 1 1 �6! A wonderful tac­tical shot! I 1 .. .lOc6 ( 1 1 . . .• g2? 12 .xg7 1-0 Stummer-De Jong, COrT. I 994/9S, 1 I . . . �xh6?? 12 �xffi mate) 1 2 .xg7 �d4+ 13 � e3+ 14 �xe3 cS I S .xffi+ �7 16 lOc3 and mate follows rapidly, Ruggeri Laderchi-Lonsdale, COrT. training game I99S.

A 9 •••• xcl

3 J..c4 fxe4 4 ll$e5 .g5 (Poisoned g2 Pawn Variation) J 25

For a long time this was considered Black's best, as White's dark-squared bishop, which seems to fonn an integral part of White's attacking force, is thereby eliminated. But then Keres showed that White still has a strong attack and it is unlikely that Black ' can

survive as his king is pennanently exposed.

10 li)f7+ �e8

l l li)xb8+ The strongest continuation, but

there are others: a) I I .eH!? J..e1? (this is risky,

after 1 1 . . .lt)e7! White should take a draw by 12 lLld6+ �dS 1 3 li)f7+ as 12 lt)xhS+? bxg6 13 1Dc3?! Wxc2+ 14 �e l .xb2 is catastrophic, and Black garners a lot of material for his queen after 12 lLld2?! bxg6 13 Axc l llhS 14 li)gS dS) 12 lLld2! ( 12 1Dc3? .xc2+ 1 3 �el bxg6 14 It)xhs Wxb2 0-1 , Miller-Keres, corr. 1934, 12 .xhS bxg6 1 3 .xgS+ J..fS 14 Wxg6 �e7 I S It)es Wh6 de­fends) 12 . . .• xb2 ( l 2 . . .• xal ? 1 3 .xhS bxg6 14 .xgS+ J..fS I S It)es wins) 1 3 .xhS bxg6 14 .xgS+ J..fS I S .xg6 ( I S It)eS!? Keres, IS . . .• a3! 16 .f7+ �dS 17 It)xg6 1Dc6 seems OK for Black) I S . . . �e7 16 It)hs Wxd4 17 Wf7+ �d6 IS li)b3 ( I S .g6+ �e7 �ts) I S . . .• xhS 19 :d I + �c6 20 .dS+

�b6 2 1 .as+ �c6 22 It)d4+ forces Black to part with his queen.

b) 1 1 lLld2!? aims to brin-B another piece into the fray, I I . . ..xb2! the queen must return to defend her king, ( l 1 . . ..xal?! 1 2 .eS+!? [ 1 2 lOxhS+ bxg6 1 3 .xg6+ �dS 14 .xgS is sim.,ler] 1 2 . . .I&1?! [better 12 . . . J..e7 1 3 .xhS bxg6 14 .xgS+ J..fS I S .xg6 �e7 with some chances] 1 3 �hS+ bxg6 14 lOxe4 [if queen and knight is a potent at­tacking force, then how about queen and two knights!] 14 . . . J..h6 [Black is unable to survive by 14 . . .• xa2 either, I S li)ffi+ �dS 16 dS .a6+ 17 � .xffi+ IS Wxf6] I S li)ffi+ �fS 16 lOxg6+ �f7?! [amazingly, after 16 . . . li)xg6 17 .eS+ �g? I S It)hS+ �h7 1 9 Wf7+ �hS 20 .xg6 Wc l the attack continues by 2 1 f4!] 1 7 %S+ �fS I S -.atS and White forces mate, Robins-Re� corr. 1970; I l . . .bxg6?! 12 .xg6 .xd2+? [ I 2 . . .• xc2 1 3 lOxhS+ �e7 14 .xgS J..h6 is a better chance] 1 3 Wxd2 J..b4+ 14 �e2 It)e7 IS lLld6+ �dS 16 WeS+ AxeS 17 li)f7 mate, Stummer-Logunov, corr. 1994I9S is very pretty!) 1 2 It)xhS+ bxg6 13 .xg6+ �dS 14 .xgS Wbs+ I S �e3 Wf5 16 :gl ( 16 �g6? �eS 17 lOxe4?! dS I S It)xfS? allowing Black to wrap up the game quickly, I S . . .• xe4+ etc., Polleschi-Walther, corr. 197 1 ) 16 . . . cS! as originally analysed by 8etinl, 17 li)f7+ ( 17 :gS?! cxd4+ I S �e2 .f4! [ I S . . . d3+!? 19 cxd3 exd3+ 20 �fl .f4 2 1 li)f7+ �c7 22 .xfS, Kozlov-Smek, corr. 19S0, is also playable for Black, providin� he continues with 22 . . . 1Dc6] 19 :cS? [ 19 li)f7+ �c7 20 .xfS lDc6 fa­vours Black, anyway] 19 . . . lDc6 20 li)f7+ �e7 2 1 li)gS e3, Edwards­Gonsalves, corr. 1986, Black is win­ning; 1 7 dS?! �c7 I S li)f7 .xclS 19

J 26 3 �c4 fte4 4 �e5 .g5 (poisoned g2 Pawn Variation)

.xm .d4+ 20 We2 ltlc:6 2 1 Jlg6 b6 is very easy for Black, Reza­Bullockus, corr. 197 1 ) 1 7 . . . We7 I S JIgS �3+! ( I S . . . cxd4+ 19 We2 d3+ 20 cxd3 exd3+ 2 1 We I .xf7 22 JleS+ .e6 23 /tl0! ltlc:6 is un­clear) 19 We2 ltlc:6 20 d5 d6 (20 . . . lLl<i4+ 2 1 Wdl d6 22 /tlxd6 transposes, 22 /tlhS �d7 23 /tlg6+ WdS is playable) 2 1 lOxd6!? (2 1 lbhS) 2 l . . .�+ 22 Wdl J.e6! 23 dxe6 Wxd6 24 .17 J.e7 2S .f4+ (2S lOxe4+? Wc6 26 .xe7? .0+ mates) 2S . . . Wc6 26 .xe4+ Wb6 27 Jlg3 �S+ 2S We I Jlm 29 lbc4+ Wc7 and Black has finally fended off the attack, and is now wiMing, Magee-Jensen, corr. 19S9.

c) It is perhaps important to note that White can force a draw here, if he is so inclined, with 1 1 1tld6+ Wd8 12 lbf7+.

1 l ... hxg6 12 .xg6+ Wd8 13 lbl7+

13 .xgS? .f4 14 lbg6 .0+ IS We I WeS 16 Itles .f5 17 ltlc:3 c6 and White is unable to bring his rook into the game, and is losing, Schmitz-Packroff, East Germany 1968; for 13 llXl2!? see the analysis for 1 1 llXl2.

13 • • • We7 14 lOc3! .xc2+ 14 . . .• xa l ? I S ItldS+. 15 Wel

15 ... d6!

Best, if insufficient. Others: a) I S . . . c6? 16 1tld6! (White must

threaten mate, any hesitiation is �unished: 16 .xgS? .xb2 17 lOxe4 .xal+ I S We2 .xa2+ 19 Wel �1+ 0-1 , Scott-Rondon, corr. 197 1 ) 16 . . . /tlfti (I6 . . .• xb2??, Maly­DoMY, corr. 1970, allows mate by 17 _gS+ lbfti I S .e5+ Wd8 19 lbf7 mate; 16 .. . WdS? 17 .xgS .xb2? permits another sharp mat­ing attack: I S .xf8+ Wc7 19 lbeS+ WdS 20 lbfti+ Wc7 2 1 lbfdS+! cxdS 22 lOxd5+ 1-0, Boisvert-Fellows, Canada 1969) and now White has a choice of wins:

a l ) There is some nice analysis in NCO: 1 7 ltlc:xe4! ! lbxe4 ( I 7 . . .• xe4+ 1 8 lOxe4 lOxe4 19 "xe4+ WdS is forced, but must be wiMing for White, as the black pieces are undeveloped) IS .eS+ the queen persecutes the poor black king just with the help of !he knight, 1 8 . . . Wfti (and on Its own: I S . . . Wxd6?? 19 .eS mate) 19 .xm+ WgS ( I9 . . . Wg6 20 .17+ WgS 2 1 .5+ transposes) 20 .f5+ Wh6 2 1 lOxe4 Wg7 22 "fti+ Wg8 23 /tlgS and even giving up his queen will not stop Black from be­ing mated! 23 . . . -.b7 24 lOxh7 Wxh7 25 Wd2 etc.

a2) 1 7 lbf5+ WdS 1 8 .xfti+ WeS 19 .g6+ WdS 20 .&.5+ WeS 2 1 Jld .xb2 22 lixI5! 1-0, MalmstrOm-Ruggeri Laderchi, corr. I 99S, to avoid mate Black must concede his queen by 22 . . .• xc l+ 23 .xc I but continuing is hopeless.

al) 1 7 .g5 �g7 ( 1 7 .. . WdS? 1 8 .xfti+ J.e7 19 WbS+; 1 7 . . . lOa6?? I S .e5+ WdS 19 lbf7 mate, Scott­Richter, corr. 1977/S) 1 8 .xg7+! (this is the only move to win, 1 8 ltlc:xe4?! J.hS 19 lbf5+ WdS 20 /tlxfti J.xfti 2 1 .xfti+ Wc7 22 .d6+ only draws, Heinrich-Kunath,

3 �c4 fxe4 4 lfue5 .g5 (Poisoned g2 Pawn Variation) J 27

corr. 198 1 ; 1 8 1Of5+?! likewise, 1 8 . . .'�f7! [ 1 8 . . . �e6? 19 dS+ lOxdS 20 lOxg7+ �fT 2 1 lofS 11M3 22 IlcH �3 23 lOxdS, forcing mate, and a very neat one, too: 23 . . . cxdS 24 .g7+ �e6 2S .e7+ �f5 26 IlxdS+ �f4 27 .gS+ �f3 28 1lf5+ .xfS 29 .g3 mate, Hertnec:k­Hergert, corr. 1988] 19 .XB7+ �e6 20 �h6 .xb2 2 1 dS+ �xdS 22 .g6+ �e7 23 lofS+ �d8 24 .g8+ �c7 2S .g3+ �d8 26 .g8+ 'h-'h, Elburg-Ruggeri Laderchi, corr. 1998) 1 8 . . . Wxd6 19 .xf6+ �c7 20 .f4+ �b6 2 1 Ilcl .d3 22 .d6 ltla6 23 b4! lOxb4 24 .xb4+ �a6 2S ltldS! cxdS 1 -0, Ruggeri-Elburg, corr. training game 1998, because of 26 1lc3.

b) I S ... lOf6? 16 ltles c6 ( I 6 ... �e6? 17 .fT+ �f5 1 8 ltlds �b4+ 19 �f1 e3 20 lOxe3+ 1 -0, Oyarzu-Macmon, corr. 198 1 ) 1 7 "f7+ �d8 1 8 .xflH �c7 19 lOfT, menacing mate on d8, 19 . . . �b6 ( 19 . . . bS 20 .d8+ �b7 2 1 ltld6+ �a6 22 a4! forcing mate, 1 -0 Motivay-Barbosa, corr. 1970) 20 .cS+ �a6 (20 .. . �c7? 2 1 lObS mate), Padula-Seri, corr. 1976, when the simplest kill is 2 1 .a3+ �b6 22 ltla4+ �a6 23 b4!.

16 ltlciS+ 16 lOgS?! lOh6! 1 7 ltldS+ �d8

( I7 . . . �d7!?) 1 8 .f6+ �e8 19 ltlh7, Strelis-Donny, corr. 197 1 , 19 . . . lOfT 20 lOxfS �fS when White must force a repetition by 2 1 .e7+ �g7 22 .f6+.

16 ••• �d7 17 .xg8 (;t NCO), 1 7 .f5+ �e8 1 8 .g6

ltla6? ( 1 8 . . .• xb2!? is tempting, and 18 .. . �d7 offers to repeat), Lescot­Nilsson, corr. 1990, when White can win by 19 ltleS+! (rather than the 19 .xg8? played, 19 . . . �e6 20 lOxd6+ �d7 winning more bits) 19 . . . �d8 20 .xg8 dxeS 2 1 .xf8+ �d7 22

.e7+ �c6 23 ltle3, gaining time on the black queen to continue the attack.

17 •••• xb2 As well as this, 1 7 . . . e3?! aims to

bring the queen back to defend, with three �ibilities:

a) 1 8 lOxe3!? .e4!? ( 1 8 . . .• xb2 19 Ild l �+ 20 �f1 �e7 2 1 ltleS+ dxeS 22 dxeS+ �c6 23 .xc8 is difficult for Black, who can't use his queenside pieces, Clarke-PItpe, corr. 1990) 19 .xfS IOc6 20 Ildl lOxd4 2 1 .d8+?! (2 1 .g7!) 2 1 . . . �c6 22 lOgS (the point of Black's play is that 22 Ilcl + �bS 23 .xc7 is met by 23 . . . 1Of3+ 24 �e2 lOd4+ and White cannot avoid the perpet­ual) 22 . . . lOc2+ 23 �d2 �+ 24 �c I lOxe3 2S .e8+ J.d7 26 .xe3 lle8 27 .d2 and Black is over the worst, Clarke-Pape, corr. 1994I9S.

b) 1 8 ltleS+!? dxeS 19 .fT+ �d6 (sacrificing the queen by 19 ... �c6 20 .xc7+ �dS 2 1 .xc2 is hope­less, the black king is too exposed) 20 lOxe3 .e4 still looks unc:lear to me, seven years on, although White can force a draw if he desires, 2 1 .xf8+ �d7 22 1ld1 exd4 23 .fT+ �d8 (23 ... �c6? 24 .c4+ Wb6 2S 1lxd4 wins) 24 .f6+ .e7 2S 1lxd4+ J.d7 26 �8+ (26 .xe7+? Wxe7 27 lOds+ �d8 28 h4 loses, to 28 ... aS 29 hS :'6 for example)

128 3 J..c4 fte4 4 lOxe5 Wg5 (Poisoned g2 Pawn Variation)

26 .. .• eS 27 Wf6+ We7 (27 .. . �cS?? 2S lIh4) 2S WitS+ etc.

c) IS fxe3 ! is strongest, I S . . .• xh2 (the idea is that from here the queen can help defend the king, and also threaten annoying lateral checks.) 19 .g4+! ( 19 WxlS Witl+ 20 �e2 .g2+ 2 1 �d3 .xdS 22 .dS+ �c6 23 Jlc I + �b6 24 .xc7+ �a6 led to a draw in Nolden-Pape, corr. 1990; 19 lLlgS .gl+ 20 �d2 .t2+ 2 1 �d3 1&6 22 .e6+ �c6 23 Jlc l+ �bS 24 lLlxc7+, Kozlov-Pape, corr. 1993, also seems to lead to a draw on 24 . . . lLlxc7 2S .c4+ �b6 26 .xc7+ �a6 27 .c4+ etc.) 19 ... �eS 20 .xcS+ �f7 2 1 Wf5+ is win­ning, for example 2 1 . . .�eS 22 lLlxc7+ �dS 23 Jlcl , winning the exchange.

18 Jldl

18 ••• lLla6 The key to Black's defences in

this line is that he may be able to force a perpetual check when the dS-knight no longer covers the c3-square. So he has to avoid losing material with checks, and the text, an idea of Bartsch, defends the cS-bishop. However, this loses just like the alternatives:

a) I S . . . c6? loses to a series of forcing checks: 19 lOes+! dxeS 20 .17+ �d6 2 1 dxeS+ �cS 22 .xlS+

�bS 23 .xcS, Trobatto-Szura, corr. 1974.

b) I S ... J..e7?! is also hopeless: 19 .g4+ �e8 20 .xcS+?! (20 .g6! �d7 2 1 .f5+ �eS 22 lLlxc7+ � 23 .xcS+ chases the king to its doom, 23 . . . �f7 24 .e8+ � 2S lLlds+ �e6 26 lLlf4+ � 27 .17+ J..f6 2S /OdS) 20 .. . �t7 2 1 .f5+ �g7 22 lLlxe7 lOc6? (Black must play first 22 .. .• c3+! 23 WO , then 23 .. . lOc6 so that he has the resource ...• c4+) 23 .g6+ �f8 24 .f6+ �eS, Maly-Bullockus, corr. 197 1 , 2S lLlds! which wins the rook, for instance, 2S .. . JlbS 26 .g7 (threat­ening mate in two).

c) I S .. . e3?! 19 fxe3 .xh2 (I9 . . .• a3!?) 20 .g4+ �eS 21 .xcS+ �t7 22 .xc7+ /Od7 (the best try, for if 22 ... �e6, then 23 lLlf4+ �f6 24 .xb7 wins comforta­bly) 23 .xd7+ �gS 24 .xb7 with a dominating position, and an extra pawn, White will win, Elburg­Ruggeri Laderchi, corr. training game I99S.

19 .94+! In the initial game, 19 .xlS?! was

played instead, and there followed, 19 ... �c6 20 lOe7+ �bS 2 1 lLlxcs and Black can force a �tual check: 2 1 ....c3+ 22 JId2 Wc 1+ 23 �e2 .c4+ 24 �e3 .c3+ 2S �e2 .c4+ 26 �el .cl+ Ih-1h, Benatar­Bartsch, Sweden I99S.

3 �c4 fte4 4 ltlx-e5 .g5 (Poisoned g2 Pawn Variation) 129

19 ••• �ea Unfortunately for Black. Ruggeri

Laderchi has analysed this to a forced win for White, and 1 9 . . . �c6 likewise, 20 1&18+ (20 .xe4?! �f5! [20 . . .• a3? 2 1 1&18+ �bS 22 lIbl+ �aS 23 .c2 menacing lIb3, and winning] 2 1 .xf5 [Black might be able to bring his king to safety after 2 1 &De7+ �b6 22 lOxf5 .c3+ 23 �fl �b4 as White has no checks] 2 l . . .lIe8+ [Black has few prospects in the ending that arises after 2 l . . .�!? 22 lIbl lIe8+ 23 &DeS+ lIxeS+ 24 dxeS .xeS+ 2S .xeS 1&13+ 26 �e2 �xeS] 22 &De3 b6 [the black king is one move from safety on b7] 23 �S!? �+ 24 �fl Wbs+ 2S .xbS+ WxbS, reaching an endgame where Black has chances.) 20 . . . �bS 2 1 .xe4 �aS 22 lib I .xa2 23 .d3 and Black's only defence against the threats of mate is to give his queen.

20 .g6 �d7 21 �r6+! Others lead to a draw: 2 1 .f5+ �e8 22 M+ (22 .g6

�d7 repeats) 22 . . . �f7 23 1&17+ �e8 24 .xf8+ Wxd7 2S .g7+ �e8 26 .g8+ draws.

2 1 .xe4 cS! 22 dxcS �xcS obliges White to take his perpetual check again, 23 .f5+ �e8 24 �7+ �e7 2S I&IS+ etc.

21 .•. �c6 22 .xe4+

22 ••• d5! 23 .xd5+ Black can even win after 23

lOxdS?! �bS 24 lib I (24 .d3+ �aS 2S �3? .xc3+ 26 .xc3+ �b4) 24 . . . �b4+ 2S lOxb4 .c3+ 26 �dl lOxb4 and White quickly runs out of checks after 23 1&18+?! , 23 . . . �b6 24 lOxdS+ �aS.

23 ••• �b6 24 � 1I'b4+ 24 . . .• c3+ 2S �e2 cS 26 lld3

.c2+ 27 lld2 .c3 28 � 1I'b4 29 lDc4+ �c7 30 .eS+ �d8 3 1 dxcS+ �d7 32 lIxd7+! �xd7 33 .dS+ wins.

25 �e2 .e7 26 1I'b3+ �a5 27 lIbl c6 28 � �b4 29 � b5 30 .xb4+ �b6 31 .cl .ea 32 llg1

Black has managed to beat off the initial threats, but he remains mate­rial down, and undeveloped, Rug­geri Laderchi-Elburg, emaiI 2001 .

B 9 ••• c6

Diemer's move opens a route to the queenside, and possible safety, for the black king.

10 �! White brings up the reserves, and

menaces the deadly �gS+. Alterna­tive moves have been discarded:

a) 10 �f4?! d6 I I �f7+ �7 12 .gS ( 1 2 1I'h4 �ffi, Kotek­Budovskis, corr. 1970171 , amounts to pretty much the same thing; 1 2 OO!? Wxal !? [this appears to lead

130 3 s'c4 fte� 4 lfire5 .g5 (POisoned g2 Pawn Variation)

to a draw, perhaps 1 2 . . . hxg6 is a better try, 1 3 .xg6, Rondon­Berthelsen, corr. 1970, and now 13 . . . s'g4+ 14 .xg4 .xa l stalls the white attack] 1 3 lOxd6! [ 1 3 .cS? fails to 1 3 . . . s'g4+ 14 13 .l.xf3+ I S lOxI3 exf3+ 16 WO , Berthelsen­Hochmuth, corr. 197 1 , when the straightforward 16 . . . hxg6 1 7 lOxhS { 1 7 s'xd6+ *d7 I S .i.xfB lhh2+ forcing mate ) 1 7 . . . lOffi should bring the game to its conclusion] 13 . . . .*.xd6 ( 1 3 . . . s'e6!?) 14 s'xd6+ *xd6 I S .eS+ *d7, Sprod­Bondick, East Germany 1984, and now, rather than losing, White should take a �tual with 16 .eS+ *c7 1 7 WeS+) 1 2 . . . lOffi! ( l 2 . . . .*.e7?? 1 3 s'xd6+ s'xd6? 1 4 • dS mate, Krustkains-Gubats, corr. 1972) 1 3 .*.hS (obviously 1 3 .xf6?! .*.g4+ 14 *d2 .dl + I S *c3 .13+ 16 *d2 .xt2+ 1 7 *c3 .13+ I S *d2 e3+ 19 *c3 e2+ 20 s'd3 lild7 2 1 s'xd6+ s'xd6 22 .xd6+ *cS 23 lOd2 .xf7 should win easily) 1 3 . . . lObd7 14 lOxhS b6 I S c4 s'a6 winning quickly, Zernzars-Dreibergs, corr. 1960.

b) to lOI7+?! *c7 1 1 s'f4+ ( 1 1 lOxhS hxg6 1 2 s'f4+ d6 1 3 lOxg6 lOffi 14 .as+ b6 I S .d2 s'a6+ 0- 1 , Prietz-Atars, corr. 1965 or 1 1 � .13+ 12 *el s'e7 1 3 s'f4+ *b6 + Bauemdistel-Paperle, corr. 1973) 1 1 ... d6, the only move, but ade­quate, as above (but not 1 1 . . . *b6??, Scott-Pupols, corr. I 967/6S, 1 2 lOd2! threatening both mate in two, and the black queen).

c) to .*.gS+ *c7 I I �4 b6 1 2 .*.xe4, Carlsson-Gyuricza, corr. 1977nS, 1 2 . . . .*.a6+ 1 3 c4 .gl and now that his king has found safety on the queenside he will win.

Now there is another divide:

BI 10 ••• lOf6 B1 10 ••• *c7 B3 10 ••• e3

White appears to have a winning attack after all of these moves, with precise play, but (b) below merits more attention:

a) 10 . . .• g2?? loses swiftly, I I s'gS+ *c7 ( l l . . .s'e7 1 2 he4 .xgS 13 lOI7+ winning, Szilagyi­Alberts, corr. I 979/S I ) 1 2 be4 lOffi 1 3 hffi .gS 14 lOI7 b6 ( l4 . . . *b6 IS s'dS+ *a6 16 .as mate, Moyano-Castro, Barcelona 19S0) IS lObS+! 1 -0, Melchor­Jackson, corr. 19S7/SS, as I S . . . *b7 16 lilds+ *a6 1 7 tOc7 mate .

b) 10 . . . hxg6!?

.. .is very rare, but no immediate refutation is evident: I I .xhS *c7 12 .xgS, Perron-Wehmeier, corr. 1993, 1 2 . . . d6, forced, 13 .xfB dxeS 14 .17+ lild7 IS dxeS .13+, reach­ing a drawable endgame: 16 .xf3 exf3+ 1 7 *xf3 lOxeS+ IS *g3. Black must si�y avoid I S . . . -*.f5? 19 S,f4 lieS 20 lIel , by I S . . . *b6!?, say.

3 J.c4 fxe4 4 �e5 .g5 (Poisoned g2 Pawn Variation) /3/

Bl 10 ••• lOr6

I I .g5! Pupol '5 move seems to be the

most efficient. a) Although 1 1 1rh4 is not bad,

1 1 . . .J.e7 ( 1 I .. .lIgS? allows a choice of wins: 1 2 J.gS [ 1 2 J.xe4! is simple, 1 2 . . .• gl 1 3 J.gS .xal 14 J.xf6+ J.e7 1 5 J.xe7+ Wc7, Brieger-Nickel, Washington 197 1 , when 16 lOc4 decides immediately] 1 2 . . .• xal 1 3 J.xf6+ J.e7 14 J.xe7+ Wc7 1 5 lOc4 bS 16 lOxe4! [ 16 lOxbS+! cxbS, Gabrans-Strelis, COlT. 1974, is also effective as 1 7 _gS! mates] 16 . . . lOa6 [ 16 . . . bxc4 1 7 .g3+ Wb7 I S lOd6+ forces mate, 16 . . . hxg6 1 7 J.d6+ Wb7 I S liX:s mate and 16 . . . lhg6 1 7 J.dS+ Wb7 I S liX:s mate, likewise, but quicker] 17 .g3+ Wb7 IS lOed6+ WbS 19 lOes+ with mate to follow shortly. I I . . .J.g7 will probably transpose, 12 J.gS .xal 1 3 J.xfti+ Wc7!? 14 lOc4! J.xfti? [ 14 . . . d6 1 5 J.xg7 lldS has to be tried] 1 5 .f4+ winning, Zagata-lkarts, COlT. 1993) 1 2 J.gS .xal 1 3 J.xfti hfti (Black might consider 1 3 . . . b6 14 J.xe7+ Wc7 1 5 1Of7! [ 1 5 J.xe4 J.a6+ 16 Wd2 lIgS 17 IOn and now, in Kermeen-Clear, COlT. 19S7, Black has to move his a6-bishop so that the king can use

the a6-square, but the lack of de­fence on the dark squares doesn't

bode well] IS ... dS 16 .g3+ Wb7 1 7 lOd6+ Wa6 I S IOxd5 J.g4+ 1 9 .xg4 cxd5 20 J.xe4! 1Oc6, Dravnieks-Budovskis, COlT. 1970, but 2 1 .d7! is crushing; 1 3 . . . Wc7 14 lOc4?! [ 14 J.xe7 b6 IS IOf7 wins as per the previous note] 14 . . . bS 1 5 .g3+ d6 16 J.xe7 J.g4+ 1 7 .xg4 hxg6 I S .e6 bxc4 19 J.xd6+ Wb7 20 .n+ forcing a perpetual check, Gunderam-Pupols, COlT. 1970) 14 .xf6+ Wc7 1 5 lOc4 (if 1 5 lOxe4 lIgS? allows mate, 16 .d6+ WdS, Reichardt-Geisler, COlT. 1990, when White overlooked the obvious 1 7 IOn+ WeS I S IOfti mate, but Black can play I S . . . b6) IS . . . b6 ( IS . . . bS? 16 I&bS+ cxbS 1 7 .eS+ WdS I S .xhS+ Wc7 19 .eS+ WdS 20 lOd6 forces a quick mate, Burk-Rebber, COlT. 19S5) 16 .eS+ WdS ( 16 . . . d6? lost on the spot in Purins-Eglitis, COlT. 197 1 , 1 7 IObS+! as 1 7 . . . cxbS I S .xd6+ Wb7 19 J.xe4+; and 16 . . . Wb7?! is hardly any better, 1 7 J.xe4 1I,S? [but 1 7 . . . dS I S lOd6+ Wa6 19 fudS also wins] I S lOd6+ Wc7 19 lOdbS+ 1 -0, Bronislawa­Blaszczak, COlT. 1992) 1 7 .xhS+ Wc7 I S .eS+ WdS 19 lOd6 J.a6+ 20 Wd2 .gl?! 2 1 1Of7+ WcS 22 lOxe4 1-0, Gunderam-Budovskis, COlT. 1970.

b) I I lOi7+?! Wc7 1 2 .as+ ( 1 2 .eS+?! d6 13 .xfti J.g4+ 14 Wd2 lOd7 =1= Eglitis-Dille, COlT. I96S) 12 . . . b6 13 IObS+! cxbS 14 .c3+ IOc6 1 5 J.f4+ d6 16 lIxhl lIgS! 1 7 J.xh7! ( 1 7 lOeS? dxeS I S JlxeS+ Wb7 19 J.xfti J.g4+ 20 Wfl lhg6, garnering more than enough com­pensation for the queen, Herbert­Pupols, USA 1966) 1 7 . . . lIg7? ( 1 7 . . . b4 I S .e3 still looks unclear) I S d5 lOxd5? 19 J.xd6+ Wb7 20 lOds+ lOxdS 2 1 J.xe4, winning, Kirwald-Bondick, COlT. 19S5.

132 3 �c4 fxe4 4 ltlre5 .g5 (Poisoned g2 Pawn Variation)

c) 1 1 �gS?! _probably only leads to a draw, I I . . ..xal 1 2 �xf6+ Wc7 1 3 �f7 ( I 3 1&4!? �b4 14 �eS+ d6 I S �xd6 b6!? 16 �xcS+ Wb7 1 7 �fS?! [ 1 7 �xc4! IIxcS I S .f7+ Wa6 19 .c4+ keeps an edge] 1 7 . . .• xb2 and eventually it was the white king that found itself mated, Keller-Peperle, corr. 1965) 1 3 . . . �b4! ( I 3 . . . b6!? 14 .eS+ d6 I S �S+ cxbS 16 .i.dS+ Wb7 1 7 �xc4+, Djellouli-I1gart, France 1976, 1 7 . . . Wa6 unclear) 14 .eS+ ( 14 .i.xhS hxg6 I S .gS b6 16 �xc4?, given by Milic as unclear, despite the fact that 16 . . . �a6+ 1 7 Wf3 1I'h I + I S We3 .c I + exchanges queens, and wins! Instead 16 .dS+ �b7 17 �c4 .el+ I S WO and Black may have no better than a

draw; but not 14 .xh7?! .i.xc3 I S bxc3 1If8 16 1I'h4 bS =1= Levy­Strobel, Ybbs 1965) 14 . . . d6 I S .e7+ �7 16 �xhS .i.xc3?! ( 16 . . .• xb2!) 1 7 .dS+?! ( 1 7 bxc3 !) 17 . . . WbS IS bxc3 .xc3 and now, instead of Gunderam's 19 �S? 'Ibc2+ 20 WfI bS 2 1 �d7+ Wb7 and, perhaps surprisingly, White cannot avoid losing material, De­strebecq suggests 19 �xd6 when Black must be content with a per­petual: 19 . . .• xc2+ 20 WfI .dl+ 2 1 Wg2 .g4+ etc. as 2 1 . . ..f3+ 22 Wgl .f8? 23 �eS .xd6 allows 24 .i.eS! which is most embarrassing!

1 1 •• ':_ a) If 1 1 . ..�e7? then 12 �f7+

(stronger than 1 2 �f4 (preferred by NCO] when Black can struggle on a few moves, by 1 2 . . .• f3+ [ 1 2 . . .• xal? allowins. a forced mate, 1 3 �f7+ WeS 14 l&I6+ WdS I S �b7+ hb7 16 .as+ WcS 1 7 .c7 mate, Baer-Reiscr, corr. 1975] 1 3 �f3 exf3+ 14 Wxf3 hxg6 IS lie I with advantage, the black pieces stuck on the queenside are no help: I S . . . d6 16 J:xd6! JIh3+ 1 7 Wg2 �xd6 I S .xf6+ Wc7 19 :CS 1Ixh2+ 20 Wgl bS 2 1 � l.o, Mulleady-Champion) 1 2 . . . WeS 13 �hS+ ( 1 3 lOcs+ WdS 14 �f4 transposes to 1 2 �f4) 13 . . . hxg6 14 .xg6+ WdS I S �f7+ WeS 16 �S+ WdS 1 7 �f4! .ul IS �f7+ WeS 19 �6+ WdS 20 .eS+! ! forc­ing mate in a few moves, Borik­Novak, Czechoslovakia 1969, as 20 . . . �eS 2 1 �f7 mate.

b) I I . . .�g7? is met by 12 �xc4 .xh2 1 3 .xg7 1I'hS+ 14 �f3 .eS I S .xffi+ winning, Kozlov­Melchor, corr. 19S919 1 .

c) And I l .. .d6? by 1 2 .xffi+ Wc7 13 �xc4! (both 1 3 .xhS? hxg6 14 �f4 �g4+ I S Wd2 e3+, Atars­Tomson. corr. 19741S and 1 3 .i.gS?! .i.g4+, Padula-Tomson, corr., are less emphatic) 1 3 . . .• xh2 ( 1 3 . . .• gl 14 .i.gS ! .i.g4+ I S Wd3 with enor­mous advantage Kozlov-Melchor, corr. 19S9191 ) 14 �f3 ( 14 .xhS dxeS I S .xlS .i.g4+ 16 Wd3 is also more than sufficient, Heap-Melchor, corr. 19S9191 ) 14 . . . .i.g4 I S .xhS �7 16 .xh7 and Black has no compensation for the piece and pawn, Beoto-Bcdevia, Cuba 1975.

Il .xf6+ 1 2 .i.f4? is not so good:

1 2 . . .• f3+! and instead of 1 3 t&f3 exf3+ and 14 . . . lIxg6 =1=, White can play 1 3 Wel !? as pointed out by

3 �c4 fte4 4 lOx-e5 .g5 (Poisoned g2 Pawn Variation) J 33

Destrebecq, but 1 3 . . Jlxg6! 14 �g6 �g7 I S �S h6! 16 .xg7 • xf4 still looks fine for Black.

1 2 lOf7+? "'eS 1 3 lOd6+ 'h-Ih, Schendel-Haas, corr. 19S5.

11 • ...i.e7 13 .f7! White would be foolish to take a

draw here with 13 lOf7+ "'eS 14 lOd6+.

13 •• ':XI6 After 13 . . . 1If8 14 �f4! is very

strong, < 14 .xf8+ also wins, but in a more pedestrian manner: 14 . . . �xf8 I S �gS+ [± NCO] I S . . . "'c7 [ I S . . . �e7 16 .i.xe7+ �e7 17 Ilxh l bxg6 I S �e4 dS 19 �g6+! leaves White with a two­pawn advantage; 19 lOgS?! .i.fS al­lowed Black to hang on in Scott­Gebuhr, corr. 1972174] 16 Ilxhl bxg6, Grava-Budovskis, corr. 1970, was easily winning for White fol­lowing 1 7 �g6!) 14 . . .• xa l I S .eS+ "'c7, Padula-oliveira, corr. 197 1 , and now, instead of 16 1&4+? when 16 . . . lIxf4 17 .xe7 b6 makes the win _problematical, 1 6 lOn+! "'b6 17 .xe7 d6 I S �d6 "'a6 19 lOdxe4 wins by a mile.

14 �IS

This is winning. 14._.fJ+ Unfortunately for Black,

14 . . .• 01? allows a forced mate: I S .xe7+ "'c7 16 .dS+ "'d6 1 7

lOc4+ "'e6 IS .e7+ "'fS, Atars­Strat, corr. 1970, 19 �3 mate .

IS lOxfJ exfJ+ 16 �fJ 16 "'d3! .i.xgS 1 7 .xh7 is even

better. 16 •• ..i.xgS 16 . . . llxgS 1 7 lie I +/-. l7 .xb7 Or 1 7 lOc4 .i.e7 I S .xh7 11e6 19

IIg1 bS 20 IIgS+ "'c7, Pupols­Dreibe!8S, corr. 19S7, 2 1 I1eS.

17 •. ':f6+ 18 "'14 .i.d1 19 "'8+ 1-0

Purins-Eglitis, corr. 197on4.

81 IO •.. "'�7

1 1 .i.f4! Or: a) I I .i.xe4!? lOf6 12 ..... �e4

13 .i.gS! (White's attack soon pe­tered out after 1 3 lOxe4? .i.e7 14 .f4 d6 I S lOn 1If8?! , Siegers­Harju, corr. 197 1 , but I S . . . lIgS! 16 lOfxd6 .i.g4+ 1 7 "'d2 .d I + I S "'c3 .f.H would have left Black playing an endgame with an extra rook. 1 3 lOn?! �c3+ 14 bxc3, as in the game Ulschmid-Dohner, Mondorf 1962, 14 . . . �d6?? should have lost on the spot: I S �gS 1le8+ 16 "'d2; instead, Black must flee, 14 . . . "'b6 I S lOxhS "'a6 with advan­tage) now Black has the choice be­tween 13 . . . �gS <and 13 . . . lOg3+!? when in both cases, it is a question of whether Black can get his pieces

134 3 Ac4 fxe4 4 lIDe5 .g5 (Poisoned g2 Pawn Variation)

coordinated, 14 *d3 [otherwise the queen escapes. 14 fxg3? .�+ I S We3 Ag7; 14 .xg3?! .xal ] 14 . . . b6 I S lbbl �hl which is not too clear) 14 Lhl h6? ( 14 . . . :gS!?) I S f4 � 16 dS with a strong ini­tiative. Tomson-Harju. corr. 197 1 .

b) I I fOf7? allows Black his dream: a queen exchange! 1 l . . ..f.H 12 *fl .xhS 1 3 .bM fOf6 14 .i.gS Ag7 I S .*.f4+ *b6 ( I S . . . d6 16 Axd6+ *d7 is simple enough) 16 Ae2 as 1 7 fOa4+ *a7 I S �c7 rods 19 Ad6 bS which should bring White's attack to a halt, Padula-Harju. corr. 197 1 .

1 l .•. hXI6! I l . . .• xal? loses. 1 2 fOxd7+! with

1 2 . . . *dS ( 1 2 . . . *xd7 is quickest. 1 3 .f5+ *dS 14 .xfB+ *d7 I S .eS mate. Atars-Tomson. corr. 1973 or 12 . . . Ad6 13 .cS! [ 1 3 fObS+? *xd7 { 1 3 . . . cxbS?? 1 4 .cS+ *dS I S .xd6 It)xd7 16 AgS+ 1 -0. Vitols­Scott. corr. 1970} 14 AeS+ *dS I S �d6 fOffi 16 AgS :fB 1 7 � *c7 and White's attack is grinding to a stop. Uhlen-Malmstrom. corr. 19S5] 1 3 . . . hxg6 [there is nothing better. if 1 3 . . . b6 14 fObS+ mates. 14 . . . *"7 I S �d6+ *c7 16 fOf7+ *b7 17 rods+ *a6 ) S .a3+ *bS 19 c4+ *xc4 20 �3+ *xd4 2 1 fOe6 mate] 1 4 .xd6+ *dS I S

AgS+ 1-0 Downey-Svendsen. corr. 19S9) 1 3 .eS hxg6 14 .c7+ *eS I S .xcS+ *e7 16 AgS+ fOffi 17 �ffi *f7 I S .eS+ Wg7 19 fOg4 1 -0 as mate follows. Ruggeri Laderchi-Malmstrom. corr. 1995.

12 .15 The generally favoured move. but

Black has recently discovered a use­ful resource.

1 2 .xhS! is certainly playable. and is probably stronger: 12 . . .• xa I 1 3 .xgS d6 ( 1 3 . . . gS is no help. 14 .xfB gxf4 I S fijf7 0+ 16 *d2) 14 .xfB .xb2!? (l4 . . .• gl? Vitols­Budovskis. corr. 1972. I S �e4! wins. I S . . . dxeS? 16 AxeS+ *b6 17 �+ *a6 I S roes mate) I S fOrT .xc2+ 16 Ad2 .A.g4+ 1 7 *el � I S .xd6 fOa6 (Elburg's I S . . . aS is better. but 19 fOes e3 20 fOc4+ *-6 2 1 �e3 is winning) 19 fOes. winning. Gubats-Vitomskis. corr. 1970.

However. not 12 fOc4+? d6 13 fObS+ ( 13 .xhS .0+) 13 . . . cxbS 14 .cS+ *dS I S AgS+ Ae7 16 .xd6+ rod7 1 7 Axe7+ �e7 IS Lhl bxc4 -/+ OIen-Downey. corr. 1992.

12 •• ..i.e7 Asain 12 . . .• xa l? is impossible:

1 3 1Of7+ *b6 14 .A.c7+! 1-0. Ortiz­Amilibia. corr. 1965/69.

13 fOxc6+ d6 14 Axd6+!

3 �c4 fte4 4 �e5 .g5 (Poisonedg2 Pawn Variation) 135

14 ••• �d6! Neither 14 . . . *d7? I S lOes+!

*xd6 16 �bS+ *e6 17 .g4+ 1-0, Grava-Kotek, COlT. 1970, nor 14 . . . �xd6? I S .d8+ *xc6 16 .xc8+, Downey-Jackson, COlT. 1989, are playable.

IS .cS+! *e6! If I S . . . *d7 16 lOes+ *d8 17

.as+! is best, White must find a way to remove his queen from at­tack whilst giving check, ( 1 7 �f7+ *d7 18 .dS+ [ 1 8 lOes+! *d8 19 .as+ transposes into 1 7 .as+] 18 . . . *e8 19 :xitl �ffi 20 .c4 �g4+ 2 1 *el l%h7 which does not look entirely clear to me, Heap­Downey, COlT. 1992) and after 17 . . . *e8 1 8 Ilxhl , NCO indicates some advantage to White. 1 8 . . . �6 ( 1 8 . . . �6? 19 �S+ 1-0, Behnnann-MacNab, COlT. 1995; White maintains a decisive initiative after 1 8 . . . �ffi 19 �! �d8 20 �ffi+ �xffi 2 1 .dS, MalmstrOm­Elburg, COlT. 1998) 19 .a4! ( 19 �c6 bxc6 20 .c7 �g4+ 2 1 *e I :lc8 doesn't appear that clear, Black does have two bishops and a rook for the queen and pawns) 19 . . . �ffi ( 19 . . . :lb8 20 �g6 �g4+ 2 1 *el l%hS 22 dS 1-0, Koetsier-Rossell, COlT. 1999) 20 dS bS 2 1 .xbS �a6 22 .xa6 �xeS 23 .e6 �f3 24 :d1 �hS 2S .xg6+ *d8 26 .xe4 is devastating, Rosso-Rosscll, COlT. 1999.

16 .eS+ cH7 17 :lxbl �xc6 1 7 . . . �&4+ 1 8 *fl �c6 19 .xh8

.*.£5 20 Wh7+ *f8 2 1 � :d8 22 �e7 �gxe7 23 c3 and the queen and two pawns are stronger than the three pieces, especially as the black king is exposed, Owens­Sakellarakis, COlT. 1999.

18 .xb8 �r6

19 *dl 19 *e3?! �f8 20 h3 �g7 2 1 �

Elburg-Owens, COlT. 2000, and now Black played the tactic 2 l . . .�d4!? (but 2l .. .�£5 was also reasonable) 22 *xd4 gS 23 .xgS �h7+ 24 .xg7+ *xg7 2S �e4 with a piece for three pawns.

19_i18!? 19 .. . �d4 20 *cl is also

playable. lO dS �b4 Unclear, Tiemann-Elburg, email

2001 .

B3 IO .•• e3

This move, the so called 'blockade attack', was fust suggested by Dreibergs.

1 1 �f7+ It is now clear that this is White's

best move �, either immediately or one move later in (a). The others:

136 3 �c4 fte4 4 �e5 .g5 (Poisoned g2 Pawn Variation)

a) 1 1 �xe3 exal , the position resembles a game between two be­ginners, onc who has developed all his pieces but lost two rooks, the other who has only moved his king and queen! 12 .gS+?! (White does better to return to the main line here, with 12 lOf7+! Wc7 13 .gS, but not 1 2 �gS+? as 12 . . . Wc7 1 3 lOfT b6 14 �dS+ Wb7 I S .eS lOa6 16 lObS ZlbS and Black was out of the woods in Murray-Pupols, Sokane Wasch 1966) 12 . . . .*.e7 1 3 lOfT+ WeS! ( l 3 . . . Wc7, transposing into the main variation, is inferior) 14 lOxhS+ WdS I S lOf7+ WeS 16 lOcs+ (White took a risk with 16 .eS!? in Jimcnez Alonso-Miguel, Spain 1995, but the game finished in a draw anyway: 16 . . . hxg6 1 7 eg7 lOffi I S 1&4 lOxc4 19 .gS+ �f8 20 Ah6 d6 2 1 .xf8+ Wd7 22 o 1th1 23 edS+ Wc6 24 eeS+ Wffi In-In) 16 . . . WdS 17 lOfT+ WeS I S lOci6+ WdS In-In, Magee­Mclchor, corr. 19S9/9O.

b) I 1 �c4?! only leads to a draw with best play, I I . . ..gl 1 2 .*.xe3 exal (Black gains nothing from keeping his queen in the game, 12 . . .• g7?! 1 3 lOfT+ Wc7 14 .as+ b6 IS lObS+! cxbS 16 .c3+ lOc6 17 .*.f4+ Wb7 I S lOcis+ Wa6 19 a4, trying to bring the rook on a I into play with decisive effect, 19 . . . lOxd4+ 20 Wd2 b4? 2 1 .c4+ WaS 22 .dS+ AcS 23 .xcS+ bxcS 24 Ac7+ 1 -0, Amilibia-Alberts, corr. I 96S) 13 lOfT+ Wc7 14 • as+!? (a critical position. 14 lOxhS? .xb2 I S Af4+ d6 16 .f7+ lOci7 17 .c4 lOgffi I S .*.d3 bS 0-1 , Ciprian-Sneiders. corr. 197214. is no good; and 14 .gS?! �e7 I S Af4+ d6 16 �xd6+ Wd7 17 �xe7 lOxe7 I S lOcs+ Wc7 19 .xe7+ �d7 is OK for Black; but Hergert's sugges­tion 14 �gS!? is dangerous,

14 . . . lOa6 [not 14 . . . b6? which pro­vokes a king-hunt: I S lObs+! Wb7 16 lOcis+ Wa6 1 7 lOc7+ WaS fol­lowing Melchor's analysis. and now I S lOxc6+! Wa4 19 .f7 .xb2 20 .c4+ �b4 2 1 .*.c I threatening mate by 1tb3+, and winning.] I S .*.dS+ WbS 16 .eS+ d6 17 lDxd6 �g4+ I S 0 Axd6 19 .xd6+ WcS 20 fxg4 lOh6 and now White can force a draw with 2 1 ec6+) 14 . . . b6 I S lObS+ cxbS 16 Af4+! (as Mel­chor points out 16 .c3+? lOc6 17 .*.f4+ Wb7 I S lOcis+ Wa6 19 dS �b4 20 .xhS .e I + is winning for Black) 16 . . . d6 1 7 .c3+ Wd7 IS 1th3+ Wc7 (Black has little choice but to acquiesce to the draw as I S . . . We7? 19 .*.xd6+ WxfT 20 .*.dS+ [20 1ths+?! Wc6 is only a perpetual] 20 . . . Wf6 2 1 �+ Wg6 22 .g3+ Wf6 23 .f4+ Af5 24 h4! forces mate; and I S . . . WeS? 19 .xcS+ Wxf7 20 �dS+ likewise) 19 ec3+ Wd7 20 �3+ Wc7 In-In. Stibal-Croker. corr. 1992.

c) I I fxe3? is just what Black is hoping for: 1 1 . . ..g2+ 12 Wel .gl+! 1 3 We2 d6 (so as to have the annoying possibility of . . . A84+, later) 14 lOf7+ Wc7 I S �fS .i.xfS 16 .xfS .xh2+ and . . . lOh6 -1+, analysis of Alejandro Melchor (or first 16 . . . lOh6!).

d) . . . as is I I Wxe3? .*.h6+. 1 I •.• Wc7 Now there is a further imp()rtant

choice. between B3 1 12 egS! , which wins easily, and B32 12 Axe3. the traditional main line . otherwise:

a) 12 .eS+? loses, 1 2 . . . d6 13 .xhS exf2 14 Wxf2 exh2+ I S WO �3+ 16 Wf2 �+ 17 We3 eg3+ 0-1 . Plath-Phillips, corr. 1992 .

b) 12 eM! amounts to the same as 12 .gS, although NCO prefers this particular move order, i.e.

3 s'c4 fxe4 4 lih-e5 .g5 (Poisoned g2 Pawn Variation) J 3 7

This wins by force. Il •• ..i.e7 1 2 . . . b6?! 1 3 .dS+ �b7 14 liJe4

.xh2 1 5 .xfS removes Black's dark-squared defence and wins.

13 -'g3+ d6 14 s'e4 -*.g4+ Else the queen is trapped. IS .xg4 liJf6 15 . . .• xh2 amounts to the same,

16 he3 liJ10 1 7 .e6 liJxe4 IS 'ihe7+ transposes below.

16 .e6! 16 -*.xhl liJxg4 1 7 liJxhS isn't

bad, either, Black has insufficient compensation for the piece, still, that did not stop him from drawing in Lescot-Melchor, corr. 1995: 17 ... liJa6 IS liJf1? ( I S 0) IS ... AfS 19 0 J:txf1 20 fxg4 Af2+ 21 �e3 An 22 s'e4 S,g5+.

16 ••• liJxe4 16 . . .• xh2 changes nothing, 1 7

-*.xe3 ! liJxe4 IS .xe7+ liJd7 19 /.Oxe4 and White wins easily, 19 . . . AheS 20 .xd6+ .xd6 2 1 /.Ofxd6 Ae6 22 Ah I IUS 23 J:th6

J:txh6 24 s'xh6 1-0, Oren­Malmstrom, corr. 1994/95.

17 .xe7+ 1OcI7 18 1Oxe4 :ad 19 .xd6+ �c8 10 .f4 winning, Falkowski-Leisebein, corr. 19S5.

831 Il s'u3

This is very unpleasant for Black. as his king has to go for a walk, but it is not the strongest.

11 •••• 181 13 .g5 Threatening mate on dS. In one of

the first games in this line, Benner­Dreibergs, corr. 1965, White contin­ued 1 3 s'f4+? d6 14 .g5? ( 14 liJxd6!? is an alternative, 14 . . . s'xd6 leads to a draw, as does 14 . . . s'g4+ 1 5 .xg4 s'xd6 16 s'xd6+ �d6 17 .f4+ We7 I S .f1+ �dS 19 .eS+ and so on, Boil-Van Dieren, 1975, but 14 . . . hxg6!? may be better, for example 1 5 liJde4+ �dS 16 .xhS s'e6 and Black defends) 1 5 s'xd6+ �d6 16 .e5+ �d7 17 .eS+ �c7 I S .e5+ �dS 19 .eS+ In-In, Eckenfels-Alberts, corr. 1967) 14 . . . S,g4+! ( l4 . . . 1Oe7 was also good in Neukirch-Dravnieks, corr. 1965) 1 5 .xg4 hxg6 16 lOxbs ( 16 .e6 liJd7 1 7 liJxd6 AlIoin-Keller, corr. 19S5, and now Black should win by using the deflection 1 7 . . . S,xd6 I S .xd6+ �dS 19 S,g5+ 1Oe7!) 16 . . . liJd7 17 .xg6 /.Og1O IS Wd2 AeS 0- 1 .

138 3 �c4 fxe� 4 lOxe5 .g5 (Poisoned g2 Pawn Variation)

13 ••• �e7 Black has some choice: Not 13 . . . b6? 14 .dS+ Wb7 I S

.xfS hxg6?! (allowing mate, but I S . . . /.l)e7 is also hopeless, 16 �S+ Wa6 17 �d3+ bS I S .xhS) 16 �S+ 1-0, Elburg-Alderden, corr. 19S5.

However, 13 ... bS!? is possible; 14 .dS+ Wb7 I S � lOffi? ( 1 S . . . lOh6!? 16 �xh6 dS is not at all clear) 16 .as! lOa6 17 llXiS+ WbS I S �f4+ d6 19 lOxc6+ Wb7 20 llXis+ WbS 2 1 .xbS+, stripping the black king bare, Grava-Alberts, corr. 1965.

14 �f4+ It is difficult to believe the sur­

prising 14 dS? of Whittemore­Dreibergs, Detroit 1966: 14 . . . d6 ( l4 . . . �xgS?? I S d6 mate) I S .g3 hxg6 16 �cS �f5 17 lOxd6 ( 1 7 �xd6+ WcS) 17 . . . Wd7! I S .eS lOa6 19 lOce4 �g4+, winning.

14 ••• Wb6 Ugly, but essential as 14 . . . d6? I S

�xd6+ Wd7 16 .13+ forces mate, van Loon-de Vries, corr. 19S2.

lS lOa4+ Wa6

16 lOcs+ The black king may be out on a

limb, but he is only a move from safety, and Black does have two extra rooks, so White cannot afford to hang around:

a) 16 J.c7? b6! (somewhat supe­rior to 16 . . . bS?? 1 7 .xe7! 1 -0, Braczko-Holiman, corr. 1991) 1 7 J.d3+ ( 17 .eS lOffi?! [ 1 7 . . .• xa2 or 1 7 . . . hxg6!?] I S J.d3+ Wb7 19 J.d6 .xa2 20 b3 J.dS?! [20 . . . 1118 2 1 .xe7 1lxf7 22 .xf7 lOds should win] 2 1 lOxhS .al?? 22 1Of7 as 23 lOxdS+ Wa7 24 .xaS+ bxaS 2S J.cS mate, Melchor-Svendsen, corr. 19S7ISS) 1 7 . . . Wb7 I S .g3 ( I S �dS? lOb6! 0-1 , Elburg­Rittenhouse, corr. 1990) IS . . .• xa2 19 b3 � 20 lOxhS dS 2 1 .eS .a3 22 .f4 �7 23 .xh6 Wxc7 and Black has almost consolidated, Stummer-Malmstrom, corr. 1994/ 9S.

b) 16 J.d3+? bS 17 lOcs+ ( 17 .eS .xa2! [ 1 7 . . . d6? is worse, IS lOxd6 J.g4+ 19 f3 J.xf3+ 20 Wf2! J.b4+ 2 1 J.g3 J.xg3+ 22 hxg3 �dS and now White forces mate: 23 J.xbS+ WaS 24 �xc6 /.l)e7, Patron­Ciprian, corr. 1972, 2S lOc4+ Wb4 26 .xe7+ Wxc4 27 .cS mate] I S b3 lOffi 19 lOxhS lOds defending e7, and c7, 20 J.d2 d6 2 1 -..S �7 22 .xh7 107ffi 23 .g7 J.g4+ 24 f3 IIgS -/+ Schutz-Leisebein, corr. 19S5; 17 J.c7 Wb7 IS J.ds lOa6 19 J.xe7 lOxe7 20 J.xbS lOc7 0-1 , Turian-Melchor, corr. 1995) 17 . . . Wb6! (17 . . . �xcS transposes to the main line) I S �! lOffi! ( l S . . . lOa6?? is a blunder, allowing mate in five 19 lOc4+! bxc4 20 lOa4+ Wb7 2 1 �S+ cxbS 22 J.e4+ 1 -0, Repp-Paschitta, corr. 1991) 19 lOa4+ ( 19 lOxbS �S!? 20 J.c7+ lOxc7 2 1 lOxc7 Wxc7! [2 1 . . . J.xgS? 22 lOxaS+ WaS 23 lOb3+ Wa4 24 lOc:S+ draws] 22 .xe7 .xb2-+ analysis of Miguel) 19 . . . Wc7 20 lOxbS+ Wb7 2 1 .xffi lieS! 0-1 , Smith-Phillips, corr. 1995.

16_.Wb6

3 �c4 fxe4 4 �e5 .g5 (Poisoned g2 Pawn Variation) 139

Trying to hold on to the dark­squares. 16 . . . �xcS? loses, 1 7 �d3+ bS 1 8 .xcS .xb2 (this is an at­tempt to improve on both 1 8 . . . d6 19 �xbS+! cxbS 20 .xd6+ llX6 2 1 .xc6+ WaS 22 �d2+ b4 23 �xb4+ Wxb4 24 c3+ 1 -0, Tums-Dreibergs, corr. 1966, and 1 8 . . .• xa2 19 �6 .xb2 [how else to defend against the threat of �xbS+ ?] 20 �c I � I 2 1 �d2 �2 22 �c3 .xc3 23 �xbS+ WaS 24 .xc3+ forcing mate, Harding-Comley, corr. 1986, although 24 llX4+ would have saved a few moves) 19 a4 itlf6 20 axbS+ (20 lLxi6! is quicker) 20 . . . Wb7 2 1 lLxi6+ Wc7 22 llX4+ Wd8 23 Itlxb2, winning the queen and the game, Elburg-Rosso, corr. 1998, so still not sufficient.

! • .I.� ... �a '+'" t � 1 'fA\'. �� /."'\ � 1 .� � �'U�

� t � � � � ,W! � �dlL.� � ;r.v � � It? � � \G'J

�. ' ��. � �. � r� � �) � � � � 4). �. '�� 4). �('r+!�� f$ . ..ua ��� .oa �'g'�� ��

. ./ '� . "� . '�,'

17 .&3 As there is no sure win after this,

White should prefer Tiemann's 1 7 .eS! , i.e. 1 7 . . . d6 1 8 itlxd6 �g4+ 19 f3 .xa2 20 liX:8+! 1 -0, Nuutilainen-Rhodes, BPCF v Finland, corr. 1998, as 20 . . . �xc8 2 1 .c7+ WbS 22 �d3+ forces 22 . . .• c4. Parry's 1 7 lLxi6!? is at­tractive, but only seems to lead to a draw: 1 7 . . . hxg6! ( 1 7 . . . �xgS? 1 8

Itlxc8+ tbrces mate in 7) 1 8 a4 ( 1 8 Itlxc8+ WbS 19 1tld6+ equal, Turian-Melchor, COrr. 1995) 1 8 ... Wc7! 19 itlf7+ Wb6 20 itld6.

17 ••• d6 Miguel has analysed 17 . . . aS!? and

it is a good alternative, 1 8 �xb8 Jhb8 19 .xb8 hxg6 ( l9 . . . itlffi? 20 itla4+ WbS 2 1 �3+ Wb6 22 �d3 and 23 itla4 mate is difficult to stop) 20 .xc8 �xcS 2 1 dxcS+ WxcS 22 .f8+ WbS 23 ll)d6+ WcS 24 itlt7+ repeats.

18 �xd6! After 1 8 1tla4+?! the reply

1 8 . . . Wa6 (forced, as the alternatives allow mate, 1 8 . . . Wc7? 19 �xd6+ Wd7 20 .g4+, 1 8 . . . WaS? just takes a few moves more, 19 �d2+ Wa6 20 .a3 etc.) 19 �d3+ ( 19 liX:S+ dxcS 20 .a3+ Wb6 draws) 19 . . . bS and there is no obvious win.

18 .. �xd6?! This appears to be a fault, as

1 8 . . . a6! 19 �xb8 �g4+ 20 f3 �xf3+ is unclear.

19 itlxd6 �14+ There is a threat to c8, but

19 . . . �e6? is too cooperative, 20 itlxe6 .xa2 2 1 liX:8+ WbS 22 .d3+, mating.

10 .xI4 1Of6 1I llX8+! 21 .e6! also wins, 2 l . . .hxg6 22

liX:4+ WbS Schmelz-Schirmer, corr. 1997, (22 . . . Wc7 23 .eS+ leads to mate) 23 a4+ .xa4 24 ll)d6+ WaS 2S ltldxb7+ WbS 26 �a4 is overwhelming.

1I ••• We7 11 .f4+ Wxe8 De Jong-MalmstrOm, corr. 1994,

and now the most efficient win is 13 �f5+ �bd7 14 �xd7+ �xd7 15 1tle6.

8 Svedenborg's variation 4 . • . d5

I e4 e5 1 IllfJ f5 3 .*.e4 rxe4 4 Illxe5 d5

Clearly, this is the move Black would most like to play, expanding in the centre and at the same time shutting out White's king's bishop. White is obliged to play 5 W'hs+ and enter into the ensuing complica­tions. As is customary, I have named this variation after Sveden­borg but, in fact. 4 . . . dS was already mentioned by Leonardo and Polerio (circa 1 575), who gave the con­tinuation 5 W'hS+ g6 6 Illxg6 Illa; 7 .eS+ .*.e7 8 1llxe7 .xe7. The line was again quoted by Bilguer ( 1 843), before G.B.Fraser of Dundee ana­lysed the variation 5 1rhS+ g6 6 lOxg6 hxg6 7 .xh8 �t7! 8 .*.e2 .*.g7, unclear, in 1 875. It was only in 1970 that the Norwegian Sveden­borg re-introduced this line with the improvement 7 .xg6+ �d7! instead of Fraser's 7 . . . �e7? 8 d4! .

5 "'5+ Obligato!y, 5 .*.b3? .gS 6 d4

.xg2 7 Itn .*.h3 :J: Banbanov­Kozlov, corr. 1980/82.

5 ••• 16 6 1llxg6 Which leads us to Black's princi­

pal decision . . .

In A Black offers his king's rook or prepares to defend against an at­tack, and in B he chooses to play an endgame a pawn down, but with compensation.

A 6 ••• bxg6

There is now a further sub-division:

AI 7 .xg6+ Al 7 .xb8 Al 7 .xg6+ �d7

As mentioned in thc introduction, this is the best square for thc king: 7 . . . Wc7?! 8 d4 (but not 8 d3?! .i.h6 9 .i.gS+ [9 .i.xh6 lOxh6 10 .gS+ Wc6 I I .g6+ draws] 9 . . . .i.xgS 10 .xgS+ Wc8 [ 1O . . . lllffi!? I I .l.xdS cxd3 is possiblc] 1 1 .xdS [White must refuse thc rook, I I .cS+? .c7 1 2 .xh8?? cxd3+ 1 3 Wd2 .c2+ winning] 1 1 . . .• xdS 1 2 .i.xdS cxd3 1 3 cxd3 �ffi with chances to both players, Radoszta-Mayo, Bue­nos Aires 1992) and:

a) 8 . . . .i.h6 9 .i.xh6 Jbh6 (9 . . . lOxh6 10 .gS+ Wc6 I I .cS+ Wf7 12 .i.xdS+ Wg6, Thorn-Dc long, corr. 1983, and now 13 .i.xe4+ .i.f5 14 .g3+ Wh7 IS .i.xb7 should be sufficicnt) 10 .gS+ Illffi ( IO . . . llffi?! I I .i.xdS lDh6 12 lDc:3 �f1 13 �, Grosar­lacko, Torrcmolinos 1986, Whitc will find thc timc to castlc long whilst maintaining thc attack) I I .xh6 dxc:4 1 2 c3 .l.c6 1 3 lDd2 with advantage, Mclchor-Oren, corr. 1990.

b) 8 . . . lDffi?! 9 .i.gS lDbd7 10 .i.xdS .c8 I I .xe4+ Wd8, Alloin­Stcinbach, corr. 1985, 12 .xc8+ Wxc8 13 .i.f3 and thc four pawns arc likcly to outweigh thc piecc,

c) 8 . . .• d6?! 9 .i.gS+ Wd7 (9 . . . lDffi? 10 lDc:3! c6 I I lOxe4 Black is crushed, Traudt-Vossclman

Svedenborg 's variation 4 . . . d5 141

corr. 2000, and othcrs) 10 .fS+ Wc6 I I .xc8 dxc:4 12 .c8+, Bet­Malazs. 1 2 . . . Wb6 13 lDd2 .dS 14 .xfS is strong.

d) 8 . . . dxc:4? 9 .i.gS+ Wd7 10 .i.xd8 Wxd8 I I lDd2 and, although Black has three minor picc:cs for his qucen, his pieces are uncoordinated, his king is cxposed, and Whitc has sevcral pawns more. Thc game Hcap-Orcn. corr. 1991 , is an cxam­plc of thc problcms hc faces: I I . . .lDc7 12 .xe4 c6?! ( l 2 . . . aS and .. . :'6 is a possiblc plan) 13 lOxc:4 .i.f5 14 .cS llh6 I S .as+ Wc8 16 lDb6 llc6+ 17 WfI. 1-0.

l .i.xd5 " Picking off a third pawn for thc

picc:c. Although thc black king is awkwardly placed. thc white queen­sidc is dormant at present, and thc black pieces will dcvclop quickly. White should thcrefore avoid wast­ing moves unncc:cssarily, i.c. 8 .f5+? Wc6 9 .i.xdS+ .xdS 10 .xfS lDd7 (or 1O . . . .i.f5 I I lDc:3 .cS �, Ballon-Lcisebein, corr. 198 1 ) I I .g7 .cS 12 .17 b6 �, Langhcld-Sncidcr, corr. 1985. Like­wise 8 .f7+?! .c7 9 .xdS+ We8 10 lDc:3 Illffi I I .gS .i.e6 ;, lac:quart-Monnard, corr. 1979 .

1 ... lDr6

9 W For thc above reasons 9 .i.xe4?! is

bad: 9 . . .• c7 10 d3 ( 10 lDc:3?! is

142 Svedenborg 's variation 4 . . . d5

best countered by 10 . . . :gS I I .f5+ WdS 12 .0 �6 � see (b), note to move 10 below, rather than 1O . . . WdS I I d3 :gS 12 lOds .xe4+ 13 .xe4 lOxe4 14 dxe4 :xg2 I S �f4 lDa6 16 �g3, trapping the rook with a slight plus, Trobatto-Kozlov, corr. 19S I ) 10 . . . :gS I I .(5+ WdS 12 .0 ( 1 2 .f4 lOxe4 13 .xe4 .xe4+ 14 dxe4 l:lxg2 I S �e3 llX:6 �, Hickl-Spiegel 19S6) 1 2 . . . llX:6 and:

a) 13 lDa3! White needs to guard c2; 13 . . . lOes ( 1 3 . . . lOxe4 14 dxe4 �S I S .c3 lLki7 is a way of play­ing for more than a draw, 16 .0 lLlf6 winning the e:pawn, while keeping control) 14 We3 lLlxe4 I S .xe4 l:lxg2! 16 .xg2 ( 16 �e3? �+ 1 7 Wxb4 �xb4+ I S c3 �xa3 =!= Tiemann-Rasmussen, corr. 1974/ 7S) 16 . . . lOxd3+ 17 Wd2 ( 1 7 Wdl? allows Black's relatively undevel­oped pieces to start a king hunt: 1 7 . . . lLlxf2+! I S .xf2 �g4+ 19 Wd2 .d6+ 20 Wc3 �g7+ 2 1 Wb3 �e6+ 22 lLlc4 [22 c4 .d3+ amounts to the same] 22 .. . �xc4+ 23 Wxc4 bS+, forcing mate, Garcia-Kozlov, corr. 1979) 1 7 . . . .i.h6+ I S Wxd3 �f5+ 19 Wc3 .f6+ 20 Wb3 �+ 2 1 Wc4 .a6+ 22 Wc3 .f6+ 23 Wc4 .a6+ 24 lLlbS .a4+ 2S b4 .xc2+ 26 Wd4 �g7+ 27 .xg7 .e4+ In-In, Miiller-Gllrd, corr. 1990.

b) 13 c3 �S 14 .dl (so that, if Black captures on e4, the reply dxe4 is check) 14 .. . WeS! I S :gl ( I S O-O? lOxe4 16 dxe4 �g4 [good, but 16 . . . Wh4! is immediately crushing, as . . JlhS is coming] 1 7 0 [ 1 7 .dS?, Fiorito-Tiemann, corr. 1975, 17 ... lLlo+! IS Whl Wh4 19 �f4 :hS threatens . . .• xh2 mating, and there is little White can do about this; 17 .a4+ c6 I S �f4 �e2 19 lLki2 �xfl 20 lOxfl lLlo+ 2 1 Wh I :g4 0-1 , Jove-Baumann, corr.

1995] 17 . . .• cS+ [ 1 7 . . . �h3 ! I S :f2 .g7 19 �f4 .tcS is a simple method of winning an exchange] I S Wh l :dS 19 Wb3? OO!! 20 fxg4 lLlt2+ 2 1 l:lxt2 .xt2 22 .e6+ �e7 23 .xgS+ Wd7 0-1 , Haller-Webe, corr. 1993) I S . . . �g4 16 .14+ c6 17 �e3 lOxe4 IS dxe4 :dS 19 lLld2, Top�Leisebein, Pokaltumier 19S7, 19 ... 00+ 20 Wfl a6, Black is better.

c) The extent of White's troubles is amply demonstrated by the fol­lowing games, first Schubert­Leisebein, 19S6, 1 3 O-O? IOd4 14 .dl �h3 I S Whl? xe4! -/+.

d) And 13 �e3? is no better, 13 . . . �g4 14 .f4 ( 14 �cS .eS, Monsalvo-Grobe, corr. 1975, changes nothing) 14 . . . lLlhs and the queen is trapped in mid-board! IS -*.cS .xcS 16 .d2 �g7-two pieces is really too much for four pawns, Edwards-Lonsdale, corr. 19S5.

e) 1 3 lLlc3? lLki4 14 .dl l:lxg2 I S �e3 �g4 16 .d2 lLlO+ 17 �xO �xO, the O-bishop cuts White in two, I S lOe2 lLkis and the end is nigh, Siggens-Faldon, corr. 19S 1 .

9 •••• e7 a) Stronger than 9 . . . lOxdS?! 10

lOxdS :h6?! (l0 . . . c6! , although rare, is the best move, I I .17+ �e7 12 lLlf6+ Wc7 13 lLlxe4 :f8 14 .g7 .tf5 IS d3 lLki7 16 �f4+ WcS 17

.i.d6, Destrebccq-Carlsson, corr. 1979, and now 17 . . . .i.ffi 1 8 lOxffi Jbffi 19 .i.g3 �S threatens the wicked 20 . . . �, trapping the white queen, with go4d chances) 1 1 .17+ �c6 12 1tlxc7! .xc7 (Black has Iit­de choice, as 12 . . . Jlffi? 13 .c4+ .i.cS 14 lOxa8 .i.c6 1 S .c3 Jbf2 16 b4 +/- Schinner-Hansson, corr. 1993; and 12 . . . AcS?? 1 3 d4 [ 1 3 lOxa8 Jlffi 14 .c4 is also effective, as in the p�vious line] 1 3 . . . Jlffi 14 dS+ �b6 13 lOxa8+ also winning easily, Roszner-Lcisebein, corr. 1992) 13 .xf8 :h7? ( 1 3 . . . Jlc6 is better) 14 .e8+! ( 14 .ffi+ .d6 I S .xd6+ Wxd6 1 6 d3 exd3 1 7 .i.f4+ �dS 1 8 cxd3 �6 and .Black has good drawing chances with the o�itc-colour bishops. and extra piece, for the four pawns, Nolden­Mage, corr. 1991) 14 . . . �dS ( 14 . . . �b6 I S d3 is also very un­p'lcasant for Black) I S d3 Jle7 16 WbS+ 1 -0, Kozlov-Zschom, corr. 1989, as 16 . . . �c6 1 7 dxc4 leaves White· with four pawns and an at­tack for the piece.

b} 9 . . . 1OC6!? is an alternative, though, 10 d3 I&S? ( 1O . . . exd3 seems fine) 1 1 .f5+ �d6 12 lOxc4+ lOxc4 13 .xc4 Jlh4 14 f4 Itlg4 I S g3 with advantage, Farinas Scijas-Acosta, Lascurain memorial 1995.

IO d3!

Svedenborg 's variation 4 . . . d5 /43

White plays for the attack, open­ing up the central files and attempt­ing to complete his development as quickly as possible. If he declines to play this pawn sacrifice, he can soon fmd himself in deep water:

a) 10 b3, White prepares the use­ful resource Aa3. There are several possible replies:

a l ) 1O . . . :h6! The stro�Jest. 1 1 .17 ( 1 1 .£5+? �d8 12 _f4 Jlg6 [ 1 2 . . .• g7! is very strong, forking g2 and pinning the c3-knight] 13 l.b2? [ 13 .i.c4 still favours Black] 1 3 . . . .i.h6 14 � lOxdS 0-1 , Guidi­Minerva, 199 1 ; 1 1 .g3?! �6 12 .i.xc6+ Wxc6 13 1Dc2 .i.g4 14 .c3+ �d7 1 S d3 lDds 16 .c4 exd3! 17 .xg4+, Thoma-&rg, corr. 1982, and now 17 . . .• c6! is best, 1 8 .0 [ 1 8 .xc6+? Jbc6 19 cxd3 .i.g7 20 d4 :'e8 is over­whelming] 1 8 . . . .i.b4+ 19 .i.d2 [ 19 �fl? dxe2+ 20 .xe2 .xe2+ 2 1 Wxe2 Jle8+J 19 . . . dx�2 _ _ Black has a strong attack.) 1 1 ..�.xt7 ( 1 l .. .�6? is inferior, 12 .xe7+ 02 .i.xc6+?! Wxc6 13 .c4+ .cS favours Black, Svendscn-Hayward, corr. 1990J 12 . . . .i.xe7 13 .i.xc4 with four pawns for the piece, and a solid position, White must be favourite; l l . . .lOxdS 12 .xdS+ :d6? [ 1 2 . . . �e8 13 .i.b2 c6 is st1pCriorJ 13 .i.a3! JbdS 14 .i.xe7 Jlf5? [ 14 . . . .i.g7 I S lOxdS .i.xa l J I S .i.xf8 Jlxf8 16 lOxc4, the observation of the previous note applies, Kozlov-Gonsalves, corr. 1986) 12 .i.xf7 �6 ( 12 . . . .i.g7!? 13 .i.b2 1tlg4 14 .i.dS Niemand-Alloin, corr. 1983, 14 . . . .i.d4 I S Itldl .i.xb2 16 lOxb2 �ffi is equal) 13 .i.b2 with a new divide:

a l l } 1 3 . . . .i.cS! is best as, by at­tacking f2, Black deters White from quccnside casding, 14 h3 ( 14 o-O? even without queens the king is not safe here, 14 . . . �e7 IS .i.c4 .i.d6 16

144 Svedenborg 's variation 4 . . . d5

h3 .t.xh3! 17 Ilfel [ 1 7 ph3?? 1lg6+ I S ci>h I IlhS] 1 7 . . . llg6 win­ning, De Paz-Miguel, corr. I99S; meanwhile, if either white piece goes to dS, then the c2-pawn is lost, i.e. 14 lDds? lOxdS I S .t.xdS lOb4 16 .t.xe4 1le6 1 7 n lOxc2+ I S ci>dl lOxa I , Hemsley-Crimp, Australia 1994, when White could have re­tained some hope by playing 19 .t.fS) 14 ... .t.d4 IS (H)-()?! .t.xf2 16 g4 .t.h4 17 :dfl ci>e7 I S .t.c::4 .t.e6 +, Oren-Heap, corr. 199 1 .

a 12) 1 3 . . . .i.d6 is very popular, but it only leads to equality, 14 0-0-0 (14 h3?! lOes IS .t.dS lOxdS 16 lOxdS b6 17 0-0-0 .t.b7 �, Bamber­Jackson, corr. 19S7; 14 lOdS? lOxdS IS .t.xdS lOb4 16 .t.xe4 1le6, Sc::hneider-Vosselman, corr. 19S7, 1 7 d3 lOxc2+ I S ci>d I lOxa I 19 .t.fS lOxb3 20 .t.xe6+ ci>xe6 with every chance of triuIJY)hing) 14 . . . lOes I S .t.dS (best, I S lOIS? lOfg4! strands the light-squared bishop, 16 .t.xeS li)xeS 17 .i.gS IlhS I S lOf6+ ci>e7 0- 1 , Svec::-Tiemann, corr. 1979, and IS .t.c4?! devalues the pawn struc­ture, IS . . . li)xc::4 16 bxc4 b6 17 h3 .t.b7 with a dear edge, Pape­Landgraf, corr. 1990) I S . . . lOxdS 16 lOxdS li)g4!? ( l6 . . . b6 is quite play­able, 17 .t.xeS .t.xeS I S d3 e3 !? Ih-Ih, Leisebein-Winckelmann) 17 :del lOxf2 (17 . . . ci>c6 I S ll)c::3 does not change much, Tiemann-Gebuhr, corr. 1975) I S Ilhfl 1lxh2? (the main line, but it is inferior, and brings Black problems. I S . . . ci>c6 is best, 19 li)e3?! [ 19 ll)c::3 li)g4 20 h3 is level] 19 . . . li)g4 [ 19 . . . .t.xh2!? 20 1lxf2 .t.g3 and with the exchange for only one pawn, Black is better] 20 lOxg4 .t.xg4 2 1 1lxe4 .t.e6 22 g3 In-In, Heap-Diepstraten, corr. 1990, but Black can play on if he wants) 19 1lxf2 .t.g3 20 IlI7+ ci>d6! (20 . . . ci>e6!? 2 1 1lf6+! ci>xdS 22 c4+

ci>cs 23 1lxe4 c6? [23 . . . .t.d6 24 .t.a3+ ci>c6 might be better, al­though the king is still awkwardly placed after 2S lle7 .t.d7 26 b4] 24 .t.d4+ ci>b4 2S .t.c3+ �a3 [2S . . . ci>cS? allows mate by 26 .t.aS! and b4, or d4] 26 cS bS?? [26 . . . 1lh I + 27 ci>c2 bS is forced, but even here 2S cxb6 axb6 29 IlfS is strong, White will try to bring a rook to the a-file, 29 . . . bS 30 lleeS .t.f5+ 3 1 d3 1 -0, Alloin-Grivainis, corr. 19S5] 27 Ilfl ! IlhS 2S ci>bl 1-0 Graber-Leisebein, corr. 19S5) 2 1 lOxc7 (2 1 1lf6+ .t.e6!) 2 1 . . . .t.xel 22 .t.a3+ ci>eS 23 lOxaS Ilxg2 24 lle7+ ci>f6 2S Ilxe4 .t.xd2+ 26 ci>b2 .t.fS 27 Ilc4, Svendsen-Gllrd, corr. 19S9, and others when Black should be able to force a draw easily enough, e.g. 27 . . . aS 2S ll)c::7 Ilgl 29 .t.d6 .t.c::l + 30 ci>c3 1lg2.

a2) 1O . . . c6 I I .t.a3 (or I I .t.xe4 when, I l . . .llgS 1 2 .fS+ ci>dS 13 .n .i.g4 [stronger than 13 . . . llg4?! 14 �a3 ! .eS? {but 14 . . .. xa3 I S .xf6+ .t.e7 is better} I S .t.xfS Ilxe4+? 16 lOxe4 1-0, Heap­Tiemann, corr. 1991 ] 14 .d3+ [ 14 .e3 lObd7 { 14 . . . lOxe4 is also fine} IS d3 .87! 16 .t.b2 .i.b4 17 ci>fI lOds I S .i.xdS cxdS 19 l1)a4 .g6 20 a3 .t.d6 and Black has an attack, Heap-Krantz, corr. 1993] 14 . . . li)bd7 I S n .t.hS [ I S . . . li)xe4 is equally �ood, 16 .xe4 { 16 lOxe4 .t.fS 17 We2 .g7 hitting a i , and winning the crucial g2-pawn, Vitols­Borrmann, corr. 1992} 16 . . . .t.fS 17 .xe7+ .t.xe7 I S g4 .t.xc2 19 d4 .t.h4+ 20 ci>d2 .t.h7 Black has an edge, Delavekouras, Sotiris-Pellen] 16 0-0 .g7 17 g3 lOxe4 I S fxe4 .t.cS+ 19 ci>g2 Wc7 and White's king position is precarious, Van den Berg-Tatlow, corr. 1996; I I .t.c::4 bS 12 1.a3 .g7 transposes to I 1 .t.a3) I I . . ..g7 12 .xg7+ ( 1 2 .fS+?

WelS 13 .i.xfS IlxfS 14 .i.e6 �S IS �3 .i.xe6 16 .xe6 lOxc3 winning, Melchor-Clarke, COrT. 19S5) 12 . . . .i.xg7 13 .i.c4 ( 1 3 .i.t7?! IOg4! [threatening both 4.lOes, and to un­dennine the c3 knight by ... e3] 14 .i.g6 e3 I S O-O-O?! [ I S fxe3 lOxe3 16 g3 is a better tIy] I S . . . lOxf2 [ IS . . . �c7!? 16 fxe3 1Of2 is another way to take the exchange] 16 dxe3+ lOxdl 1 7 Jlxdl+ �c7 I S .i.d6+ � 19 �+ �S 20 b4+? [White still has chances, by 20 c4 for in­stance] 20 . . . �xa4 2 1 c4 �aJ! · [2 I . . .Jlh6?? 22" .i.c2+ �3 23 J:ld3+ �a2 24 .i.b I + � I 2S :'3 mate, Alloin-Tiemann, ·

COrT. 19S5] 22 .i.c2, ElbuFJ-Tiemann, COrT •• 19S9. when 22 . . . � 23 J:ld3 .i.b2+ 24 �d l .... f3 .ie- a very strong defence) 13 . . . bS (also 13 . . . Wc7 14 0-0-0 .i.fS IS h3 IObd7 16 g4 .i.g6 ; Alloin­�aard, COrT. 1991 ; but 1 3 . . . 1084 is 'ess effective here, 14 h3 .[ 14 .i.cS? b6 IS .i.e3'? bS 16 .i.e2, Alloin­Svendsen. COrT. 1990, loses to the elementuy 16 . . . lOxe3 1 7 fxe3 b4 I S O-O! �e7 { I S . . . bxc3? 19 Jlt7+ � 20 .*.c4+ �eS 2 1 Ilxg7} 19 lOxc4 .*.xa I when the extra black rook will surely nutke short work of the 'four whi� pawns] 14 . . . e3 I S f4, LOffier-Borrmann, COrT. 1991 , IS . . . �c7 [ l s . . . lOa 16 Jln iOc4? 17 lOxc4 .*.xa I I S c3 favours White] 16 .i.b2 .*.fS 17 lOds+ cxdS I S .i.xg7, Loffier-Borrmann, COrT. 199 1 , I S . . . Jlh7 19 .i.xdS Ilxg7 20 hxg4 exd2+ 2 1 �d2 Ilxg4 about equal) 14 .i.e2 �S (again, ., • . . . "'c7 I S 0-0-0 .i.fS, etc., is pos­sible. Although three pawns have the same nominal value as a piece. especially in the endgame, with so many pieces still left on the board and with all the white pawns so far back, Black has the edge in almost all these positions. e.g. 16 h3, Pape-

Svetlenborg 's variation 4 . . . tl5 145

Svendsen, COrT. 1990, 16 . . . aS 17 .i.b2 b4 I S lOa4 IObd7 19 g4 .i.e6 and the white knight is side lined) I S 0-0-0 .i.xc3 !? ( l S . . . lOxc3 16 dxc3+ �S is more logical) 16 dxc3 �c7 17 "'b2, Alloin-MalmstrOm, COrT. 1990, ( 1 7 c4 1Oc3 I S .i.d6+ "'b6 ; Malmstrom-Borrmann, COrT. 19901 91) 17 . . . b6 (to deter White from c4) I S hbS!? .i.xbS 19 c4 .i.xc4 20 bxc4 IOf6, the game is level.

aJ) 10 . . . 1Oxds?, as so often, is a mistake: 1 1 IOxdS _g7? ( I I . . .• CS 12 1Of6+ "'d6 is a better bet) 12 lOf6+ �6 13 .xc4+ �b6 14 .i.b2 and Black can " resign. k.ozlov­Skrastins, COrT. 1992.

a4) 10 . . .• 17!? is a playable alter­native, 1 1 xc4 lOxc4 12 .xc4 .i.d6 (menacing . . . JleS, and stopping White from cutling) 13 �n JleS 14 _a4+ 1Oc6 IS .i.b2. Bnvo­Borrmann. COrT. 1979, when Black should prepare further development by IS . . . �dS.

b) 10 .i.xc4?! As mentioned above this is one pawn too many! 10 . . . JlgS ( IO . . . "'dS?! has already been examined in the note to move 9) 1 1 .13+ �S 12 .0 1Oc6 ( l 2 . . . .i.g4 wasn't too bad either, in Korhonen-Nilsson, COlT. 1985: 13 .d3+ iObd7 14 b3 and now 14 . . . .i.fS I S O Jlxg2 is good) 1 3 0-0 ( 1 3 h3 [to stop .. . .i.g4] 13 . . . lOd4

146 SvedenbofJ 's variation 4 . . . d5

[ 1 3 . . . �e4 14 �e4 �S wins the g2-pawn immediatelr] 14 .d3 cS I S �dl �e4 16 t&e4 .i.f5 17 :e I ? [because of a tactic the rook fails to defend e4, 1 7 0 is forced] 17 . . . :Xg2 I S c3 .i.xe4 19 .xe4 .xe4 20 :Xe4 :gl+ 2 1 :el :Xel+ 22 �e I lOc:2+ 0-1 , Garcia-canal Oliveras, Barbera 1997; 1 3 d3? lOd4 14 .dl :Xg2 is considered in the note to move 9) and now:

bl) 13 . . . lDeS?! is the most popu­lar, but without justification, 14 .e3 lLlfg4 IS .d4+ ( l S .e2?! lLlxh2! with advantage, Stegmann­Borrmann, corr. 1992, as 16 �h2?? 1fh4+ 17 �gl lLlg4 wins) IS . . . .i.d7? ( l S . . . lLld7 holds the bal­ance) 16 h3 ( 16 .dS is also strong, 16 d3? less so, 16 . . . �h2! 1 7 �h2?? [ 1 7 .dS :hS I S .xb7 was the only chance] 1 7 . . . 1fh4+ I S �gl , Naeter-Leisebein, corr. 1982, and for some odd reason, Black re­frained from the evident I S . . . lLlo+ 19 .i.xO .xd4 !) 16 . . . 1fh4? 1 7 f4? ( 1 7 .i.xb 7 c6 I S 0 defends, and wins material) 1 7 . . .• g3 ! I S bxg4 lLlxg4 19 :0 .el+ 20 :n .g3 ih-1h, Turczynowicz-Borrmann, corr. 19S7, 2 1 :0 .el+ 22 :n .g3 leads to a repetition.

b2) 13 . . . �e4 is a good alterna­tive, 14 �e4... ( 14 .xe4 lLlb4) 14 .. . lOd4 I S .ffi ( IS .d3 .g7 16 g3 .i.d7 is also promising) I S . . . �c2 16 :bl .i.h3 and, having holed the white queenside pawn structure, Black has a considerable plus, Krongraf-Stummer, corr. 19S0.

b3) . . . as is 1 3 . . . lOd4 14 .d3 .d6 ( l4 .. . cS also looks good) I S b3 lLlg4 16 f4 .i.d7 1 7 .i.b2 � I S �hl lLlxh2! 19 �h2 .i.cs and the white king was in a painful predicament, Nyffeler-Elburg, corr. 19S5.

c) 10 O-O?! Black has too many open files in front of White's king for this to be playable. 10 . . .• d6! (the simplest, forkins. dS, and h2, 10 . . . lOc:6 1 1 .i.xe4 illeS 12 .g3 :gS1! , Drucke-Krantz, corr. 1990, is worse: 13 lLlds! :Xg3 14 .i.f5+ �eS I S �e7) i l l g3 ( 1 1 f4?! exO?! [ I l . . .lLlxdS 12 .f5+ �c6 13 .xcS 1fh6 is also strong, . . . .i.cS+ is the main threat, and somewhat sim­pler] 1 2 .i.xO .d4+?! [ 12 . . .• xh2+ 13 � .i.cS+] 13 �hl?? [ 1 3 :t2 �dS 14 d3 is not too clear] 13 . . . :Xh2+ 14 �h2 1fb4+ I S �gl .i.cS+ 0- 1 , Baiion-Me1c:hor, Barcelona I99S; 1 1 h3 is best, 1 1 . . .�dS 12 .f5+ �S 13 .xdS .xdS 14 �dS .i.f5 and White is only slightly worse) 1 I . . . �dS 12 .f5+ WdS ( 12 . . . �c6!? 13 .xcS 1fh6 14 h4 .i.d6 is also possible) 1 3 .xdS .xdS 14 �dS .i.f5, Baranowski-Grobe, corr. 1975, and Black's light-squared bishop is worth more than the white pawns ( I4 . . . .i.e6!?).

d) 10 h3?! does nothing to solve White's problems: 1O . . . c6 (or 10 . . . lOc:6 1 1 �xe4 :gS 12 .f5+ �dS 13 .0 transposing to a line above) 1 1 .i.c:4?! bS 12 .i.e2? (White loses control of gS) 12 . . . :gS 13 .f5+ �dS 14 .f4 :Xg2 =1= Druke-Elburg, corr. 1990.

t O ••• eIcl3+ I t .i.e3

1 I ••. c6! There is a wide choice for Black

at this point, and no real consensus as to which move is best. I, how­ever, rather like the text. The idea is to force the bishop away, and then play 12 . . . .t.h6xe3. The advantages of this plan are threefold: First, White will be forced to recapture with a pawn on e3, giving him an isolated pawn and closing the e-fiIe. Second, Black exchanges a danger­ous attacking piece, and third Black gains control of a lot of dark squares, i.e. his king is safe on c7, and he can install .a knight on eS. Here are the alternatives:

a) 1 1 . . . d2+?! , attempting to cause the irtax·imum inconvenience possi­ble, �ite wilUtave to castle 'by hand' after 12 �d2:

a l ) 1 2 . . .• d6 13 .f7+ ( 1 3 .t.cS?! "f4+ 14 .t.e3 In-In is a bit limp, sthirmer-Svendsen, corr. 1992; 1 3 .t.d4 .f4+ I� .t.e3 likewise, al­thQUgh 14 Wd3 !? is feasible;- but 1 3 Wc I ! ? is interesting, to vacate the d-file, and then if 1 3 . . . lOxdS 14 .fi+ .e7 IS .xdS+ We8 16 :e l ) 1 3 . . . .t.e7 14 :ad1 1If8 IS Wc l Zlxfi 16 .t.xfi .xdl+?! (quite UMeces­sary: 16 . . . W06! 17 Zlxd6+ .t.xd6) 17 IIxd l+ Wc6 1 8 a4! a6? 19 lDds .t.d8? 20 lOb4 mate, Nyman­Svendsen, corr. 1991 .

a2) 1 2 . . .• g7 13 .d3 .t.d6 14 Zlae l lOc6 IS h4 ( I S .t.xc6+?! bxc6, and certainly not I S . . . �c6?? 16 Wbs mate) I S . . . i0b4?! ( I S . . . tOeS) 16 .t.e6+? ( 16 .f5+! Wd8 1 7 .gS is pnpleasant for Black) 16 . . . Wd8 ( I6 . . . We7!? 1 7 .f5 [17 .t.xc8? lOxd3 1 8 .t.h6+ lOxe 1 19 .t.xg7 lOO+ 20 pO Zlbxc8] 1 7 . . . .t.xe6 18 .t.gS .f7 is u.nclear) 1 7 .f5 J.xe6 -18 .xe6 :e8 19 .f5 :eS 20 .0 .g6 2 1 IIc I with balanced play, Melchor-Budovskis, corr. 1993.

Sved6nborg 's variation 4 . . . d5 147

13) 12 . . . lOxdS? 13 lOxdS .d6 14 .g4+! (White also did well with 14 .d3 c6? in Melchor-Hayward, corr. 1990, as I S .f5+ wins on the spot, but 14 . . . lOc6 is better) 14 . . . Wc6 I S .c4+ Wd7 16 :ad I bS 17 .xbS+ with a big plus, Magee-Hayward, corr. 1991 .

a4) 12 . . . c6?! 13 .t.gS! ( 1 3 .t.b3 Wc7 14 .t.f4+ Wb6 IS :ae l [ IS a4!? as 16 .t.e3+ Wc7 17 :ael is also good, Nyman-BorrmaM, corr. 1991 ] I S . . .• d8+ 16 Wc I as? 17 J.gS lObd7 18 .t.e6 .t.e7 19 .t.xd7 J.xd7 20 Zlxe7 1-0, fape-Vitols, corr. 1994196)i 13 . . . .t.g7 14 :ael .f8 IS .t.f7 .t.h6 16 .f5+ 1-0, Nilsson-Stamer, corr. 1993.

b) I l . ..dxc2!? is somewhat risky �ut at least it means that White has only two pawns for the piece! 12 .xc2 and:

b l ) 12 . . . �7 13 .a4+ (other queen moves: 13 .d2!? .t.d6 14 0-0-0 [ 14 .t.0 lOc6 IS 0-0-0 tOes 16 J.e2 We7 looks quite reasonable for Black, Elburg-Destrebecq, corr. 1993] 14 . . . lOc6 IS .t.f4 tOe7 16 .i.xd6 cxd6 17 J.e4? [17 .t.c6+ bxc6 18 .xd6+ We8 19 .xf6 �6+ is about equal] 1 7 . . . lOxe4 18 lOxe4 .xe4! 19 .xd6+ We8 20 .d8+ wf7 2 1 .xh8 J.f5, the sting in the tail ! 22 Whs+ lOg6 0-1 , Logunov-Pape, corr. I 994/9S; 13

/48 Svedenborg 's variation 4 . . . d5

"'3!? c6 14 .*.e6+ [ 14 o-O-O! lOxdS { 14 . . . .*.d6 I S .*.e6+ �c7 16 lbd6 chd6 17 lid l+} IS lOxdS cxdS 16 .xdS+ �e8 17 :bel] 14 . . . �c7 I S .*.f4+ .*.d6 16 .*.xd6+ chd6 17 0-0-0+ �c7 1 8 lObS+ cxbS 19 .g3+ Ih-Ih, Leisebein­Pape, corr. 1995, as White has a perpetual check) 13 . . . �d8 14 0-0-0 (14 .*.gS .*.e7 I S .*.13 lOfd7 16 .*.f4 lOcs 17 0-0-0+ .*.d7 and Black can defend his king by blocking the d7 square. 1 8 .c4 JlfS 19 lOds .*.d6, unclear, Guido-Hector, Geneva 1989) 14 . . . lObd7 (14 . . . .*.d6 I S .*.13 .d7?, this is not an ideal file for a queen, 16 lObS a6 1 7 lbd6! cxd6 18 .f4 �e7 19 .*.d4 1 -0, Stummer­Svendsen, corr. 1994/95; 14 . . . lOxdS I S lOxdS .*.d7 16 "'3 �c8 1 7 .c4 .*.c6 1 8 .*.f4, Stummer-Reinke, corr. 1994/95, when 1 8 . . . .*.h6 ex­changes the bishop, with good chances) I S .*.gS .f5?! ( l S . . . .i.h6) 16 .*.xf6+! .xf6 1 7 .*.xb7 :b4 1 8 g4 L3! 19 lbd7+ .*.xd7 20 .xa3 Jlb8, Dahne-Stame, corr. 1986, and now 2 1 .xa7 favours White.

b2) 12 . . . .*.h6 exchanging an at­tacking unit, 13 0-0-0 .*.xe3+ 14 fxe3 c6 (opening another central file by 14 . . .• xe3+? is suicide, I S �bl c6 16 .i.xc6+ [flashy, and good, but 16 :bel 1 -0, Matheis-Kem, corr. 1998, has its points!] 16 . . . chc6, Terblanche-Nilsson, corr. 1 990, but now 17 lOe4+ �b6 1 8 lOxf6 is the most precise) I S .*.xc6+! �c7 16 .*.e4 1&6 1 7 Jlhfl .*.e6, Oren­Niemand, corr. 1994/97; Black has avoided the worst, and can now bring his final piece into the action.

b3) 12 . . . �d8? 13 0-0-0 lObd7 14 Jlhel .g7 IS lObs .*.d6 16 .*.xb7! .*.xb7 17 Jlxd6, the point, 1 7 . . . Jlc8 (for if now 17 ... cxd6?? 1 8 .c7+ wins) 1 8 Jld2 and with his king marooned in the centre, and no

co-ordination, Black is lost, De Smet-Keller, COI1. 1986.

c) 1 l . . ..*.h6!? can transpose into the main line, but there is one small problem with this particular order of moves

12 O-O! (Black is doing well after . 1 2 0-0-0 .*.xe3+ 13 fxe3 c6; and 12 .xd3 .*.xe3 13 fxe3 both transpos­ing, whilst 1 3 .*.e4+!? �e8 14 .*.g6+ �fS IS fxe3 1Oc6 16 0-0 lOes 1 7 .d4 cS 1 8 .f4 �g7 and play is balanced, Stummer-Stamer, corr. 1994/95) 12 . . . .*.xe3 ( 12 . . . c6? 13 .*.xh6 lOxdS? 14 lOxdS cxdS I S Jlae I is crushing, Jurgenson­Borrmann, corr. 1991 , as I S . . . �7? 16 .e6+ forces mate in six) 13 fxe3 JlfS ( I3 . . .• xe3+!? is risky, 14 �hl lOxdS [ 14 . . . lOhS?, Logunov-Clirke, corr. 1994/95, with a nasty threat to g3, but there is a large flaw: 1 5 .*.13 .eS 16 .*.xhS JlxhS 17 Jlt7+ �d8 1 8 .g8+, winning the queen, 1 8 . . .• e8 19 JlfS] I S lOxdS!? .eS 16 lOf4 dxc2?? [ l6 . . . d2 1 7 :adl �d8 1 8 lbd2+ J.d7 fights on] 1 7 JIae I c l =. 1 8 .d3+ .d6 19 .f5+ �c6 20 lbc I + and the black king is cut off from the rest of its forces, and soon dies, Kozlov-Elburg, corr. 1994/96) 14 .xd3 �e8 (an unfortu­nate necessity as 14 . . . c6? loses: I S .*.xc6+! [or I S lbf6! lbf6 1 6 .*.xc6+ chc6 17 "'5+ �c7 1 8

00+] I S . . . �c7 [ I S . . . Wxc6 16 Ilxf6+ .xf6 { 16 . . . 1lxf6 1 7 "'5+ �c7 I S OO+} 1 7 "'S+ �d6 I S ltlc4+ 1-0, Piazza-Leisebein, corr. 19S7] 16 1lxf6! :ds 1 7 J.d7! Ilxd7 I S lObS+ �S 19 .g6 .g7 20 .xg7 1lx�7 2 1 IlfS+ �d7 22 :d1+ �c6 23 IlxcS+ WxbS 24 IlddS lle7 25 h4 1 -0, Pfeifer-Dahlhoff, corr. 1990) I S lObS!? (This move asks a lot of questions of Black. White can't afford to hang around, I S e4?! c6 16 J.b3 lObd7 17 .g3 IOes I S � lOfg4 ;, Jurgenson-Gaard, corr. 1992; and I S Ilf4?! is also too slow, I S . . . c6 16 Ab3 J.e6! [ 16 . . . lObd7 1 7 :an IOes I S .d4 with strong pressure, · Nyman­Drao.cniek, corr.' 1991] 1 7 .g6+ �dS I S :dl+ lObdl'J9 ltlc4?! [ 19 J.xe6'.xe6 20 IlM .f7 is only equal] 19 . . . J.dS! 20 J.xdt cxdS 2 1 lOgS .xe3+ 22 �h I IlgS 23 .f5 llcS and Black is out of the woods, Krantz-Stamer, corr. 1994/95; I S :ad I is a vtry dangerous alterna­tive, I S . . . lDbd7 16 .g6+ �dS 17 �6 c6 [ 1 7 . . . �eS!?] I S J.e6! �eS 19 J.xd7+ J.xd7? [ 19 . . . lOxd7 is the only chance, although the black king is not Jooking too health):] 20 1l<i4! Iln � I lLlc4 lDxe4 22 1IfbS+ IlfS 23 1lxfS+ .xfS 24 1lxe4+, win­ning, Berner-Wegelin, corr. 19S5) IS ... lDa6?,\if the knight comes here to defend c7, then it cannot protect the king on d7, I S . . . lDxdS? 16 WxdS .i.d7? [a blunder, but if 16 . . . lDa6 then 1 7 IlxfS+ WxfS I S Iln + �8? 19 1lf3 also signals the end] 1 7 IlxfS+ .xfS I S lOxc7+ 1 -0 Scbirmer-Landgraf, corr. 199 1 ; but I S . . . c6! must be tried, i.e. 16 J.f7+! [ 1 6 lDd6+ �dS 1 7 J.f3 �c7 and the king is safe] 16 . . . Wxf7 1 7 lOc7 [ 1 7 �7+? �eS· I S lOc7+ �dS 1 9 .xe7+ Wxe7 20 lDxaS b6 and the knight is trapped; Black will

Svedenborg 's variation 4 . . . d5 149

eventually have three pieces for a rook and some pwns] 1 7 . . . �g7 [ 1 7 . . .• xc7?? I S 1Ifh7+ skewers the king and queen] I S lDxaS b6, again the knight is trapped, but this time the black king is under attack, 19 1lf3 J.g4 [ 19 . . . J.b7? 20 1lg3+ � 2 1 Iln J.xaS 22 1lxf6+ wins the queen] 20 Ilf4 lObd7 2 1 .c3 IOes 22 :an J.e2 with unclear j)lay. This line needs some tests) 16 .86+ �dS 1 7 :adl lDxdS I S IlxdS+ .i.d7 19 Ilfdl , 'pin and win' ! , Schirmer­Nolden, corr. 1991 .

11 J.b3 It can be important to retain con­

trol over the gS square sometimes, but 1 2 J.f3 is also played: 1 2 . . . J.h6 ( 12 . . . dxc2!? 1 3 .xc2 �dS, is play­able but dangerous, 14 0-0-0+ lObd7 [ 14 . . . J.d7 I S Ilhel �7 16 ltlc4 lDxe4 1 7 J.xe4 .xh2?! , too greedy, Melchor-Downey, con. 1993, I S J.f5 and the black king is under fire] I S Ilhel �7 an impor­tant resource, 16 h3 J.h6, unclear, Logunov-Grobe. corr. 1994) 13 0-0-0 (or 1 3 .xd3+ �c7 14 0-0-0 J.xe3+ IS fxe3 lDbd7! [but not I S . . . J.e6? 16 .d4! lDbd7 17 .f4+ �cS I S lZd2 lDb6 19 JZ.hd I with ad­vantage, Nyman-Gaard, corr. 1990, Black has lost the coordination of his rooks, and control of the all­important h2-b8 diagonal] 16 e4

150 Svedenborg 's variation 4 . . . d5

/t)eS ; lackson-Kozlov, corr. 1989/90) 13 . . . .i.xe3+ 14 fxe3 lIg8! (if 14 . . . Wc1?! I S �B3+!) IS .fS+ Wc7 16 .f4+ Wd6 1 7 .c4, Leisebein-Bornnann, corr. 1985, and now 17 . . .• eS 18 .xd3 .i.e6 and ... lDbd7 ; would have been the most straightforward.

12 ••• .i.b6! Continuing the plan, but again,

there is no shortage of alternatives: a) 12 . . . d2+?! 1 3 Wfl !? ( 1 3 Wxd2

see above) 1 3 . . . b6 14 .i.xd2 .i.a6+ I S WgI Wc7 16 lIel with a strong initiative, Kozlov-MalmstrOm, corr. 199112.

b) 12 . . . dxc2!? is not to everyone's taste (I don't like it!) 13 .xc2 Wd8 ( l 3 . . .• eS? 14 ()-()-()+ Wc7, Magee­Downey, corr. 1989, when White could have decided the game by I S .d2!, menacing both .d8 mate and �f4, I S . . . .i.d6 16 .i.d4 .f5 17 .i.c2 etc; 1 3 . . . Wc7?, Herling-van Will in-gen, corr. 1976, 14 lDbS+ Wd8 I S ()-()-()+ lDbd7 1 6 .i.xa7) 14 ()-()-()+ lDbd7 I S JIbe I .g7, Dissels­Raijmaekers, corr. 1993. White's position looks very aggressive.

c) 1 2 . . . b6? l3 0-0-0 .i.a6?! (to keep the d-fiIe closed, but lightning strikes elsewhere; better l3 . . . �c7 14 lIhel but White is much better) 14 .i.gS .i.g7 IS /t)e4 lIm 16 Jlbel and material is lost, Alloin-Strelis, corr. 1983.

13 t-O-O! The only move to test Black's set­

up; after 1 3 .xd3+ Wc7 14 0-0-0 .i.xe3+ IS fxe3 lDbd7 (intending . . . /t)eS, but I S . . . .i.g4 16 Jldel lDbd7 is a sound alternative) 16 .d4 b6 1 7 .f4+ Wb7 1 8 Jldfl lIm the d7-knight is coming to eS, or cS, and Black has an edge, Niewold­Van de Velden, corr. 1997.

13 ... .i.u3+ 14 fxe3 b6! Black prepares his king's safety.

14 . . . Wc7?! is inferior, I S .g3+! Wd8 ( I S . . . Wb6 16 Ld3 lDa6?! 1 7 lDa4+ WaS 1 8 .e I + 1 -0, Magee­Tiemann, corr. 1989; of course, l S . . .• d6 is impossible, 16 .g7+ picks-off the rook) 16 Ld3+ lDbd7 1 7 :bdl when Black had to suffer uncomfortable pressure, but man­aged to defend himself, Magee­Kozlov, corr. 1989.

15 l1xd3+ Neither I S .xd3+ Wc7 16 e4

lDbd7-eS ;, nor I S .g3?! .i.a6 ; offer anything to White, but I S Jlbfl may be best, although Black is fine after I S . . . lIm 16 Ld3+ Wc7 1 7 .g3+ Wb7 1 8 .d6 .xd6 19 Ld6 �7 20 lIf4 Wc7.

15 ... Wc7 16 e4 16 lIfl lDbd7 and Black is doing

well, he will play . . . h6 and con­nect his rooks next go. 16 h3 lDbd7 17 e4 .i.a6 1 8 lIO :am 19 .93+ Wb7 20 lie I /t)eS ;, Melchor­Elburg, corr. 1989 (by trans­position).

16 ••• lDbd7 17 liD .i..6 18 lII5 :'fl 19 .g3+ Wb7

20 :e1 lDcs 11 e5 lDxb3+ 22 axb3 lDd7 13 lIxfl lidS 24 Wbl lIO 25 lIxn .i.xn 26 .0 .i..6 27 .f5 �xe5

Black should be wIDDlng, lalcobsson-McDonald, corr. 1995.

A1 7 .sb8

If White gains no advantage play­ing for the attack, then perhaps he should just take the rook? However, although it is difficult to demon­strate a forced win for Black, the white position is constantly on the edge of a precipice. White has to take particular care with his queen, and the line has fallen into disuse since I wrote The Latvian Gambit.

7 ••• �f7 Fraser's move, threatening to en­

snare the white queen.

8 .d4 The pin on the d-pawn permits

White to extract his queen without losing material, although the queen will be mercilessly kicked around the board by the black minor pieces. Oddly, it reminds me very much of a school chemistry experiment to demonstrate Brownian motion. The alternatives are inferior, as the queen finds herself trapped on h7:

a) 8 .i.b3 .i.g7 9 �7 and: a l ) 9 . . . .i.e6!, by defending eIS,

Black prepares . . . 1t:xi7-fS to win the queen. With:

a l l ) 10 0-0 lOd7! 1 1 0 e3 ! (the f-fiIe must be kept closed) 12 .i.xdS (there is nothing else, White needs the h3-square for the queen, and 12 g4? lOh6 stops her getting there) 1 2 . . . .i.xdS 13 �3, Kaedy-

Svedenborg 's variation 4 . . . d5 151

Saunders, corr. 1995, when there is no particular reason Black cannot hold on to his e-pawn thus: 1 3 . . . e2 14 :Cl .e7 I S �hl lIe8 16 1&3 .i.xc3 1 7 dxc3 1Ogf6.

a 12) 10 d3?! lOci7! (the most ac­curate, 10 . . . exd3 1 1 cxd3?! 1Od7! 12 g3 [ 1 2 g4 is the other yossibility to save the queen, 1 2 . . . tOeS 13 -..u .i.xg4 but the white kingside resem­bles Swiss cheese!] 12 . . . �S 13 O-O? but White is in trouble, any­way, 1 3 . . . lOfti 0- 1 , Bomnann­Leisebein, corr. 1985, as 14 � lOO+, this seems to be the same, but allows White the possibility of 1 1 1Od2!) I 1 g3 ( 1 1 g4!? is again possi­ble, 1 I . . . exd3 12 gS!?) 1 I . . . tOeS?! (menacing ... 1Ofti, but now is the time for 1 I . . . exd3! and Black is at least equal) 1 2 lOd2! exd3 1 3 0-0 ( 13 f4! looks right, followed by lOO-gS) 13 . . .• ffi!? 14 f4! lDc4 IS f5 gxf5 Hage-Muysenberg, corr . . I99S, unclear.

a 13) 10 1&3 1Od7! 1 I .i.xdS ( 1 1 lDxdS!? lOfS 12 lDxc7! lDxh7 13 lDxe6 .c8 is fun, but Black won't be too worried about this) 1 1 . . . .i.xdS 12 lDxelS /l}fS 13 �3 .xdS is fairly level, Neukirch­Grobe, corr. 1972.

a14) 10 f4!? 1017 (if 10 . . . exO? 1 1 0-0, but with the c l-h6 diagonal closed 10 . . . /l}h6! must be right, 1 I 1&3?! [ 1 1 f5!? lDxf5 12 d3 1017] 1 1 . . .1017 12 lDxe4 .7 13 ... lOgs+ � when the white queen is a 8�ner) 1 1 f5 (forced) 1 I . . . gxf5 12 WhS+ WfS 13 0-0 I&s 14 1&3 Clarke-Niemand, corr. 199419S, and White has managed to sort out his most pressing problems.

a2) 9 .. . c6 is slower, 10 1&3 ( 10 d3? is bad, 10 . . . exd3 I 1 cxd3 .i.f5 12 0-0 [ 1 2 .i.c2? .e7+ 13 wn lOd7 +] 1 2 . . . �7 13 g3 [else . . . 1OfB wins] 13 . . . �S 14 /l}d2 /l}h6, trapping the

152 Svedenborg '$ variation 4 . . . d5

queen, AlIoin-Diepstraten, corr. 1983/4; 10 f31 exf3 1 1 O-O .gS! 12 lbf3+ lOa; 13 lOc3 .*.13 14 lbf5, obviously forced, 14 . . .• xf5 I S � IOg4 16 h3?? [ 16 lildl .*.d4+ 17 lDe3 holds on] 16 . . . J.ci4+ 0-1 , Zuchn�Erler, corr. 1990; 10 0-0 lild7 [ 10 .. . .*.13 I I f3 exf3 12 lbf3 also favours White, Clarke-Reinke, corr. 1994/9S] 1 1 �3 IOh6 1 2 f3 lOa; 13 �4 exf3 14 lbf3 .*.g4 IS Ilf4 1013 and the black pieces com­bine beautifully, always gaining time on the white queen or rook, 16 .a .*.h6, Clarke-Grobe, corr. 1994/9S, and now White should pre­fer the pragmatic 1 7 d3 ! .*.xf4 1 8 .*.xf4 with advantage) 10 . . . .*.13 ( l0 . . .• g�!1 I I g3 [and not I I IOxe41 .xg2 1 2 �g3 lOa; 13 � IOg4] I l . . .�a; 12 � .xh4 13 gxh4 and Black has a little compen­sation) 1 1 d3 ( 1 1 g31! IOh6 forces 12 g4 lOxg4 =1=; I I 0-0, White plays for f3, 1 1 . . . lOa6!1 [and not l l .. .lild71 12 lOxdS!] 1 2 f3 � 13 �hl lOa; 14 � Ilh8, prepar­ing to give the white kinS a hard time) l l . . .exd3 12 cxd3 .e7+ 13 .*.e31! ( 1 3 �dl ! lild7 14 Ile l is stronger, Koetsier-Rosso, corr. 1999) 13 . . . lild7 14 0-0-0 lOfS I S lOxdS cxdS 1 6 .i.xdS+ lDe6 1 7 .*.b3 Ilc8+ and Black has good play, Guillet-C�itaine. Fouesnant 1999.

a3) 9 . . .• gS1 is to be avoided: 10 h4! ( 10 g3 i s worse, 1O . . . lOa; [ 10 .. . lOh61! 1 I lOc3 with advantage AlIoin-Fiorit, corr. 1983] 1 I -.&4 .xh4 1 2 gxh4 breaking up the white structure) 10 . . .• xg2 I I .i.xdS+ �a; 1 2 Ilfl .*.h3 ( 1 2 . . . lDe7 13 lOc3 .*.h3 14 lOxe4+ �eS IS .xg7+ Elburg-Hayward, corr. 1992) 13 .*.c4 lOc6 14 d4! ( 14 b3 is also effective, 14 . . . lle81 I S .*.b2+ lDes 16 .*.xeS+ 1-0, Hergert­Tiemann, corr. 1989) 14 . . . e3 I S

.*.xe3 lOb4 16 lOa3 ( 1 6 .*.gS+ � 1 7 .xg7 is even quicker) 16 . . .• e4 1 7 0-0-0 hfl 18 .*.gS+ � 19 .*.xfl lOxa2+ 20 �bl 1-0, Clme­Svendsen, corr. 1994/9S.

a4) 9 ... cS1! plans to suffocate the b3-bishop, 10 d3 bS 1 1 lOc3 c4 12 lOxe4! cxb3 13 lOgS+ Wffi 14 � .e8+ I S .*.e3 bxc2 16 � lOc6 17 J:lhe I and the exposed black king is too great a handicap, Ruggeri Laderchi - Elburg, corr. training game 1998.

b) 8 .*.e2

8 .. . .i.g7 9 ·�7 lOc6 ( . . . lOd4' is a distinct possibility now that the c2-pawn is undefended, 9 . . .• gS!1 is interesting, 10 g3 lOa;! I I � .xh4 12 gxh4 i&6 13 d3 [ 1 3 lOa3] 1 3 . . . lOb4 14 lOa3 exd3 I S .*.xd3 .i.g4 16 h31! .*.f3 17 IlgI lOxd3+ 1 8 cxd3 lle8+ 19 .*.e3 d4 =1= Melchor-Kozlov, corr. 1985) and it is difficult to know what to suggest for White here:

bl ) 10 f3!1 lOd4 I 1 fxe4! (the best chance, I 1 lOc31 lOa; 12 � lOxc2+ 1 3 �d l lOxal 14 fxe4 d4 I S lOdS, Nyman-OWen, corr. 1988, I S . . . d3 =1=; 1 1 .*.dI1! .gS! 12 0-0 exf3 1 3 .*.xf3 lOffi 14 .i.d I .*.13 wins) l l .. .lOxe21 (incorrect, 1 I . . . �f6 12 Ilfl .*.f3! 1 3 � lOxc2+ 14 �dl .*.xe4 is unclear) 12 Ilfl+ lOa; 13 eS .*.13 14 lbf3! gxf3

I S exf6 .xf6 16 �S+ Wf'8 17 Wxe2 .eS+ I S Wfl 11 (but I S Wd I lleS 19 c3 defends easily) I S . . . lleS winning, Szilagyi-Pescod, corr. I 97617S.

b2) 10 d3? lOd4 1 1 .i.dl ( 1 1 Wdl lbxe2 1 2 Wxe2 .i.g4+ 1 3 Wfl .i.hS trapping the queen once again, and winning, Dogiel-Bomnann, corr. 1992) I I . ..exd3 12 cxd3 .e7+ 13 .i.e3 lbf6 14 �4 lbf3 I S .f4 .i.h6 16 .0 d4 +, Bomnann-Kozlov, corr. 1979/S I .

b3) to O-O?! lbd4 I I .i.dl .i.f3! (stronger than 1 1 . .. ..i.e6? 1 2 f3 exO 13 .i.xO lbf6 14 .i.dl [ 14 �? lbxc2, Strelis-Svendsen, corr. 1990] 14 . . .• d6 IS �4; I I . . .lbf3+?! is fun, but unnecessary, 12 .*.xf3 exf3 13 d4 .i.g4 14 lbd2 fxg2 I S Wxg2, Kraszewski-Mulleady, corr. 1995, IS . . .• e7! · 1 6 0 .i.hS when White has to play 17 lbe4! to save his queen, unclear) 1 2 d3 ( 1 2 O?! exf3 13 .i.xf3 lbf6 14 �4 lbxc2 I S lbc3 lbxa I 1 6 g4 is complicated, but must surely be winning. for Black after 16 . . . lbxg4 17 1IfxdS IlxdS I S .i.xj4 We6; 1 2 c3? lbe6, menacing . . . lilfS, 13 g3 lbf6 0-1 , Haletzki-Stammer, corr. 1979, as 14 �4 gS IS .i.hS+ WgS) 12 . . . exd3 13 c3, Hage-Collins, corr. 1999, 13 . . . lbc2 14 .i.xc2 dxc2 I S lb.3 when Black has the pleasant choice between I S . . .• e7 16 .*.e3 lbf6 17 �4 IlhS, and I S . . . d4!? 16 cxd4?! .xd4, trapping the white queen again.

b4) Perhaps to lb.3?! is possible, but 10 .. .• gS (threatening . . . lbf6) 1 1 g4!? lbd4 . 12 h4 lbf3+! 13 .i.xf3 1Ife7 14 .*.e2 lbf6 and again the queen's stay on the board proves to be short-lived.

8 ••• .*.e6 Again White must decide whether

to put his bishop on e2 or b3.

Svedenborg 's variation 4 . . . d5 /53

All 9 .i.b3 All 9 .i.el Other moves are hopeless, e.g. 9

lbc3? lbc6 10 .e3 .i.h6 1 1 "cS dxc4 12 lbxe4 lOd4 +, Ramazzotti­

. Destrebecq, France 1984. All 9 .*.b3

9 ... lbc6 Black can try to pick up the b3

bishop here, but it might be more · trouble than it's worth: 9 . . . cS to .e3 c4 1 1 .i.xc4! (the pragmatic decision, 1 1 .*.a4?! .c7!? [ 1 1 . . .a6 is also possible 12 c3 .i.h6 13 f4 lbtti 14 .*.dl lbW7 I S .12 .c7 16 0-0 lbcs and the hole on d3, with attendant development problems, give Black good play, Fiorito­Niemand, corr. 19S3] 12 lbc3 [ 1 2 f4 .i.cS 13 .e2 a6 14 c3 �f6 IS .i.d I lbc6 + Florito-Elburg, corr. 1984/SS] 12 . . . a6 13 f3 lbf6 14 fxe4? d4! IS .gS [ I S .xd411 .i.cS] I S . . . dxc3 16 eS .*.e7 17 Ilfl .i.f3 0- 1 , Korhonen-Elburg, corr. 19S5, although this is premature, as after I S lhfS gxf3 19 exf6 hf6 20 �S+ We6 2 1 bxc3 bS 22 .eS+, say, White has some pawns for the piece) I l . . .dxc4 12 lOc:3 lbc6 ( l 2 . . . .*.h6 13 .xe4 lOc:6 14 d3 � I S � lbd4!? 16 0-0 �f3 [ 16 . . . .*.g7!?] 1 7 �3 .i..B7 IS .0 .d7 19 .i.e3 cxd3 20 :adl lbxe3

154 Svedenborg '$ variation 4 . . . d5

2 1 .xe3 :hS 22 :Xd3 .c7 23 .g3 .xg3 and the two black bishops managed to hold their own in the ensuing endgame, Kozlov­Svendsen, corr. 1990) 13 �e4 .i.h6 14 .f3+ ( 14 .c5 .dS 15 lDd6+ Wg7 16 .xdS J.xdS 1 7 b3 cxb3 I S c4 J.xg2 19 JIg I should lead to equality, Bonmann­Niemand, corr. 19S3) 14 .. . J.fS!? (I4 .. . �f6 is simpler) 15 �g3 lOd4 16 .xb7+ .i.d7 17 .dS+ J.e6 Jackson-Kozlov, corr. 1 990, I S WxdS :XdS 19 Wd I �f6 and Black is ver:r active.

IO We3 There is another, bizarre, possibil­

ity in 10 �S? J.g7 1 1 Wh7 .g5 1 2 0-0 �f6 (as this is good, and simple, there is no real need to con­sider 12 .. . lOd4!? 13 �3 �f3+ 14 Wh I �f6 15 d4 [better than 1 5 �xe4?! �e4 16 d3 � 1 7 pf3 JlhS] 1 5 . . .• fS 16 �3 .xh3 17 ph3, Eydan-Eydan, when 1 7 . . . 00 keeps all the positional pluses) 1 3 d4 fuh7 14 .i.xg5 �g5 -/+ Slater­Karc, corr. 19S5.

IO •• ..i.b6

I I f4! Best, otherwise: a) 1 1 .g3?! �f6 1 2 0-0 ( 1 2 -eM

�S!? 13 0-0 lOd4 14 d3?? [ 14 Jlel M 1 5 Wh.3 is far superior] 14 . . . .i.xc l 1 5 .xhS lbhS 16 :Xci �2+ ( H , Driscoll-Downey, corr.

19S0) 12 . . . lOd4 13 Jle�, Heap­Elburg, corr. 1991 , and now, instead of 13 . . . �fS 14 .c3 .d6 15 h3 :hS 16 d4 when White had escaped the worst, 1 3 . . . �h5! 14 Wc3 J.g7 1 5 �

.e3 �fS 16 .e2 �f4 + is unpleasant.

b) 1 1 We2? lOd4 12 .dl?! (the bizarre-looking move 12 .n ! is better, as g2 is thereby defended, 12 .. . c6 [with the white queen so �assive, 12 . . . c5 seems right, e.g. 1 3 &3 c4 14 .i.a4 a6 1 5 lDdl b5 16 c3 �6 17 .i.c2 �5 with a lot of space] 1 3 c3 �b3 14 axb3, un­clear, Amdahl-Hartsook, National Open 1991) 1 2 . . .. g5 1 3 wn .i.g4! the most accuraw, ( 1 3 . . . �f6 is also strong, 14 �3 .i.g4 1 5 .e I [possi­ble now that e2 is covered, 15 d3?! �4 16 We l �f3 0-1 , Gardner­Fiorito, corr. 1977, although White can continue here] 1 5 . . . Wh4 16 f3 .xe I + 1 7 Wxe I exf3 I S pf3 JleS+ 19 Wf2 .i.xf3 20 Jln c6 and Black has good play, Salewski-Borrmann, corr. 1990) 14 .i.xd5+ ( 14 h4 is no help, 14 . . .. fS I � Wel .i.e2+ 16 Wgl .g4 17 Jlh3 1Wxh3; add not 14 .el? .i.e2+ 1 5 Wgl?? �f3 mate) 14 . . . Wg7 1 5 f3?! �f3 + Femandez-Gayobart, corr. 1990.

c) 1 1 .c5? �ge7 12 �3 b6 13 .a3 �fS 14 d3 .i.xc l 1 5 :Xc i .g5 16 0-0 �h4 with a swift mate, Harju-Gebuhr, corr. 1975176.

1 l • ..l'Dle7 Aiming for fS to continue tortur­

ing the white queen; the more com­mon 1 1 . . .�f6 is not so good:

a 1 ) 12 h3 !? trying to stop . . . �g4, but it is not necessary, 12 . . . Wd6 ( I 2 . . . .i.g7!? 13 c3 �7 14 0-0 �fS 1 5 .e l �S 16 d4?! exd3 1 7 00 lies is very promising, Magee­Clarke, corr. 1990) 13 0-0 .i.g7 14 c3 ( 14 d4? allows the simple tactic 14 . . . �g4 1 5 Wel .i.xd4+ 16 Whl

lOe7, Tiemann-Grabner, corr. 1975, and Black has attacking chan�) 14 . . . :I1S I S d3 exd3 16 .xd3 .tf5 17 "e3 "d7 I S 1If3 .te4 19 1Ig) lOe7. Black has a strong, and ultimately decisive, initative, Lawrence-Gillam, corr. 1 9S0.

al) 1 2 0-0 �&4 13 .el ( 1 3 "cS? .tf8 14 �S?? lOd4 I S .xb7 lOe2+ 16 <i>h I , Butzelaar-Diepstraten, Hil­versum 19S9, now the neatest win is 16 ... lLlg)+ 1 7 bxg) .tcS when a check on hS forces mate) 1 3 . . .• d6 14 d3 .cS+ ( I4 . . . exd3 I S cxd3 lLlb4 is an alternati� I S <i>hl e3 16 lLlc3 lLlf2+ 1 7 �gl i.xf4 I S llxf2!? exf2+ 19 "xf2 .xf2+ 20 <i>xf2 and White has a sound extra pawn, Wolny-Jablonski, Byd�szcz 1975.

a3) 12 d4 lLlg4 1 3 .gl .tg7 14 c3 �+ I S g3 � 16 f5! (played to free (he f4-square; 16 lLld2 lOe7-f5, with a light-squared bind on the kingside, Borrmann­Micheloud, corr. 1979/S0) 16 . . . gxf5 1 7 old I lIgS?! I S .txg4 .xg4 19 �f4 lIcS 20 lLld2 and at the cost of a .,awn White has solved his devel­opment problems, Doyle-Ev6ns, corr. 1993.

a4) 12 lLlc3 lOe7 ( 1 2 . . . d4? 13 .i.xe6+ <i>xe6 14 Wh3+) 1 3 0-0 lLlf5 14 "el "d6 I S lOe2 �g4 16 h3 "cS+ 1 7 <i>hl :I1S and Black has some play, but the white kingside is solid, Fiorito-Tiemann, corr. 1975.

12 0-0 lLl15 Black won very quickly in

Gemignani-Hamar, corr. 1974, with 1 2 . . . .tg7 1 3 c3 �f5 14 .e2 Wh4 I S �S?? (but I S g3 ! � 16 W81 is perfectly respectable.) I S . . . lLlcd4! 0- 1 , as 16 cxd4 .txd4+ 17 <i>hl lLlg3 mate.

13 "1'2 1 3 "e I is possibly superior,

1 3 . . ... d6 (alternatively, 1 3 . . . .tf8!? -cS. or first 13 . . . �S!?, but not

Svedenborg 's variation 4 . . . d5 155

1 3 . . . �? 14 c3 lL\bS [ 14 . . . lLlxb3 I S axb3 is also good for White] I S d4 .tg7 16 lOa3 �6 1 7 lLlc2 �S and White was doing well, Jackson­Jensen, corr. 1991 , although some­thing nasty happened along the h­file, later on.) 14 d3 exd3 I S cxd3 IOb4 16 .dl WcS+ 17 d4 Wxd4+ I S Wxd4 �d4 19 lOa3, Jackson­Borrmann, corr. 1990, when 19 . . . .tg7 stops the c l-bishop from developing, and promises Black fine play.

13 •• ..*.1'7 Black can also play 13 . . . WhS!? 14

lLlc3 lLlcd4 I S d3 e3 16 .txe3 !? �b3 17 axb3 d4 IS .td2 dxc3 19 .txc3 .tg7 but White is better here,

Bastian-Destrebecq, Ales 1979. Destrebecq suggests 13 . . . .tfS, in­

stead, e.&: 14 lLlc3 b6 I S We I .tcS+ 16 <i>hl i&d4 17 .ta4?! c6 I S lLldl bS 19 c3?!, Durville-Diepstraten, corr. 1993, when Black missed the shot 19 . . . lOe2! , winning, as 20 Wxe2 �g3+ 2 1 bxg) Whs+ mates.

14 c3 �b4!

15 d4 00 16 .e2 � 17 14 1 7 . . . :I1S is a big threat, and the 17

g3 of Reinke-Heap, COIT. 1994/96, allows 17 . . . �cl I S Lcl �g) 19 .12 lLlf5 with dangerous play around the white king.

17 ... �llcl l. llllCl lLlb6 19 15?

156 Svedenborg 's variation 4 . . . d5

19 gS was the only move, al­though 19 . . . lLlf5 20 lln llh8 is strong, with the menace of ... lLlg3.

19 ••• pf5 10 liD 20 gltf5?? .gS+. 10 ... lLl:lg4 11 .A.:ld5?! .A.:ld5 11

1l:lf5+ And now Black faltered on the

brink of victory. 1l ... �e6? 13 11f4 Unclear, Borrmann-Leisebein,

COrT. 1986, when simply 22 . . . .A.ffi! 23 llxdS llg8 wins.

All 9 .A.el

The disadvantage of this move is that c2 is not defended, adding force to moves like .. . li:ld4, but, on the other hand, from this square the kingside is given assistance.

9 ... 1Dc6 10 .e3 White has little choice, as 10

.a4? leaves the king short of defenders:

a) 1O . . . .A.cS 1 I ll:lc3 ( 1 1 d3?! � 1 2 0-0? [ 1 2 g3 .A.xf2+ 1 3 �dl only loses a pawn, not the king] 1 2 . . . lLlffi 1 3 ll:ld2 li:ld4 14 .A.d I llh8 1 5 h3 .A.xh3 16 g3 .xg3+ 0- 1 , Pire­Kolzon, USA 1985; 1 1 o-O? is al­ways very risky with the h-file open, and no queen to defend, 1 1 .. .ll:ld4 1 2 ll:lc3 -..4 1 3 d3 lLlffi 0-1 , Moser-Stader, COrT. 1979, as

there is sim�y no defence against . . . llh8) 1 1 . . .Wh4· 1 2 li:ldl lM4 1 3 g3? .ffi 14 llgl (what else? 14 �3 loses to 14 . . . ll:lxe2 1 5 ll:lxe2 .A.xf2+ 16 W .0 1 7 11n .A.g4 1 8 �S � and .. . a6) and instead of 14 .. .lle8?, Aker-Magee, USA 1992, 14 ... ll:lxe2 15 �e2 .0+ 16 �el .A.g4 puts an immediate end to White's miseries.

b) 1O . . .• gS 1 1 g3 .A.cs 1 2 ll:lc3 ( 1 2 �S .A.b6 1 3 �3 .f5 14 lln, Markus-Melchor, COrT. 1982, and now 14 . . . ll:ld4 almost forces White to play 1 5 g4 .eS 16 .a4 lLlffi when the white kingside is falling apart as 1 5 .a4? .A.d7 16 .A.bS lOxbS 1'1-ll:lxbS a6 Wins material) 1 2 . . .• f5 1 3 li:ld l d4 (to stop the white knight from hopping to e3) 14 �S?, Dartnell-Chadwick, COrT. 198 1 , ( 14 t3!) 14 . . . d3! (threatening . . . .A.xf2+, winning the queen) 1 5 .xb7 ll:ld4 16 cxd3 exd3 1 7 ll:le3 ll:lc2+ winning.

10 ... .A.h6 1 1 f4 The pawn is best placed in front

of the queen, 1 1 .g3 ll:ld4 ( I l .. .lLlffi?! is slofier, and allows 1 2 0-0 [or 1 2 c3 !!Crhaps] 1 2 . . . ll:ld4 13 .A.d I lLlf5 14 Wlt3 .A.g7 leaves the white gueen exPosed, 1 5 ll:lc3? d4 16 �bS, Meester-Diepstraten, Hilversum 198 1 , 16 . . . ll:le3) 1 2 li:la3 .A.f8!? <an interesting idea, 1 2 .. .tl�ffi is more typical) 1 3 .A.dl .A.d6 14 .e3 .ffi 1 5 O-O?! (Into the lion's den! 1 5 c3 lLlf5 16 .e2 leaves the same open) I S . . . ll:le7 16 c3? ( 16 h3 &f5 17 .c3 .eS is also unpleas­ant for White) 16 . . . ll:lef5 17 .e l .A.xh2+ 1 8 �h2 llh8+ 19 �g I lLlO+ 0- 1 , Beaulieu-Saalfeld, COrT •

1996, as 20 .A.xO exO 2 1 g3 gS and . . .... 6 forces mate.

1 l ••• d4 Black can rarely resist this thrust,

but 1 1 . . . lLlge7 is also interesting:

a) 12 g4!? played to oontrol f5, 12 . . . d4 1 3 .g3 d3 14 Adl � I S 1&3 .d7 16 c3 lObdS 1 7 0-0, Koskinen-Grabner, corr. 1987, when I would prefer 1 7 ... llfB to gang up on the f4-pawn.

b) If 1 2 0-0 then IJ . . . lOf5 1 3 .a Ag7 14 c3 d4 loob sharper than 12 . . .• d7 1 3 AbS lOfS 14 wa Ag7 IS c3 1lh8 of Downey-Nyman, oorr. 1993.

c) 12 c3 lOfS 1 3 .a, Thomson­Bemard. oorr. 190 1 , 1 3 . . . d4 (else White will plal.d4 himself).

d) 1 2 lOa3 1Of5 1 3 .a Ag7 14 c3 d4 IS 0-0 ( lS AbS!?) IS . . . d3 16 .i.dl . Nilsson-Svendsen. corr. 1998. when 16 . . . a6 keeps White bottled up. he will find it difficult to bring his quecnside into the game.

11 .11 On this square the queen is

slightly less exposed than after 1 2 .g3 (not 1 2 .xe4? .i.ds 1 3 &c4 [ 1 3 .d3? lOb4 14 · � lOxc2+] 13 . . . Axc4 14 d3 .i.ds IS .e2 �+ 16 .0 1le8+ 1 7 �fl .g4. Lonsdale-Bornnum. oorr. 1986. and the check · on d l is decisive) when. although Black's position is very promising. he is material do.wn. and the white position is solid. H�w then to proceed?:

a) 1 2 . . . lOa; 1 3 0-0 (instead, 1 3 d3 .i.f8!? 1 4 h4!? e3 I S hS lOxhS 16

Svedenborg 's variation 4 . . . d5 157

AxhS gxhS 1 7 IlxhS .a; 18 fS .i.xf5 19 .xc7+ flJc7 20 .f4 looked unclear in Holecek-Pu, oorr. 1972174. but quickly went Black's way; 1 3 a4?! looks odd, 13 . . . d3 14 cxd3 IOd4 [ 14 . . . lOb4 IS 1&3 exd3 16 AD .d4! 1 7 fS .i.xfS 18 .xc7+ �fB 19 lObS? { 19 .c3 } 19 .. . 1&2+ 20 � ... =1= Downey-Ha}'WUd. oorr. 1992] IS .i.il1 It)hS?! [ I S . . . M 16 .a exd3] 16 AxhS gxhS 1 7 1&3 h4 ;t Downey­Malmstrom. corr. 1993) 1 3 . . . lOb4 ( 1 3 . . . flJc7!? 14 1&3 lOfS IS .el d3 is also reasonable. Mierzynski­Hunstock, corr. 1993) 14 1&3 d3 IS cxd3 exd3 16 .i.dl. Simango­Carlsson. cOrr. 1990. and, instead of grabbing the a2-pawn, I think that Black should play 16 .. . c6 to keep the knight stuck on a3.

b) Apart from the previous line. which is the most popular, but a lit­tle slow. Black can essay the ultra­sharp 12 . . . d3 !? 1 3 cxd3 IOd4 14 .i.dl ( 14 1&3?! lOxe2 IS �e2 exd3+ 16 � ltlffi 17 llel .d4+ 1 8 � Axf4 and White is reeling. Fleckncr-Svendsen, Merano 1988) 14 . . . lOfS (or 14 . . . ltlffi!? IS dxe4 lOxe4 16 .e3 .a;! 1 7 .xe4 [not forced, 1 7 lOc3 is possible, for i!l­stance] 1 7 . . . 1Ie8 1 8 0-0 Ac4 19 .xb7 Axfl . Downey-80mnan, corr. 1992, and White is in trouble as 20 �fl .xf4+ 2 1 AD .xh2 leaves his king looking worried) I S .12 exd3 16 1Oc3 Ag7 17 .D lOa; 18 .xb7 lLldS! 19 AD Ilb8 20 .xa7 .t.d4 2 1 .as Ab6 22 .a4 1Od4. White has lost time with his queen. and the black pieces are swarming all over his position. Downey-Kozlov. oorr. 1989.

c) 12 . . . lOb4 13 1&3 d3 is similar to the above. 14 cxd3 exd3 I S AD .d6 ( I S . . . flJc7!? 16 .i.xb7!? IlbS 1 7 .41 loses instructively, [ 1 7 Ae4!

158 Svedenborg 's variation 4 . . . d5

is necessuy] 1 7 . . . Jbb7! 1 8 exh6 ed4 19 WfI 1&2! 20 lOxc2 dxc2, threatening . . . �c4 and White has no defender for his light squares, 2 1 �el �c4 22 d3 :b6 23 Wh7+ �e8 24 �3 :e6+ 2S �d2 .12+ 26 �c3 .i.dS 27 eb8+ �d7, Lonsdale­Kozlov, corr. 1986) 16 0-0 c6 1 7 f5 ed4+ 1 8 �hl gxf5 19 �h5+ � 20 eg6 efti was also fine for Black, Bertoni-Malmstrom, corr. 1999.

Il ... lOf6 I am surprised that moves like

1 2 . . . d3 !? have not been tried in this position; play may continue: 1 3 cxd3 exd3 14 �dl �g7 when White can no longer castle.

13 0-0

13 .•• d3!? This is by far the most popular,

but is far from obligatory, as, in a way, it does dissipate the space advantage.

14 cxd3 exd3 15 �dl White keeps an eye on c2, but I S

�O seems fine, I S . . . lOd4 16 1&3 .i.f8 17 �xb7?! (walking into the tactics; 1 7 �hl is safer, when 1 7 . . . �cS?! would be met by 1 8 lOa4) 1 7 . . . :b8 1 8 �a6 ( 1 8 .te4 is no improvement, 1 8 . . . �cS 19 �hl lOg4 20 eel [20 eg3 lOe2]

20 .. . 1OxIa!) 1 8 . . . �cS 19 �hl lOb3, regaining his rook, Valverde Lopez-Borrmann, corr. 1989.

15 ... �4 16 � �f8 Bringing the bishop to the impor­

tant gl-a7 diagonal, rather than the 16 . . . bS?! of Gaard-Leisebein, corr. 1990.

17 �bl !

17 •• ..i.c5?! In the light of White's obvious re­

ply, this is premature. Black has to look at ways to prepare this, like 1 7 .. . bS!? for instance, e.g. 1 8 :el �cS 19 :e� �b6 20 eg3 lOf5 2 1 ee I ed4, menacing another . ..It�g3+ mating attack after, say, 22 Jbe6?? �g3+ 23 bxg3 :b8+.

1' �4! Simple, and effective, 1 8 1Ih4?!

eb8!? 19 exh8 Jbh8 with a pow­erful attack, Scott-Paz, corr. 1974, despite the exchange of queens, 18 eel?! a6 (alternatively, 1 8 . . . eb8!? is tempting, 19 � [ 19 lOa4?? exh2+ 20 � lIh8+ forces mate] 19 . . . lOxe4 20 exe4 lOf5 2 1 exd3 1Ih4 leads to a strong attack) 19 �! lOxe4 20 exe4, Nilsson­Evans, corr. 1990, when 20 .. . �f5! threatens mate, with .. . �g3+, etc. 2 1 ee I ed4 and Black has a strong position.

18 .. ..*.e7 19 b3 ll:)g4 20 .i.xg4 .i.xg4 21 .i.b2

White has successfully completed his development, and can now exploit his material advantage, Alderden-Goedhart, corr. 1984.

B 6 ••• ll:)r6

7 .e5+ .i.e7 8 �b5+! An important finesse. a) The immediate 8 �e7 (or first

8 .xe7+) is worse because it allows Black to use the c6 square for his queen's knight: 8 . . .• xe7 9 .xe7+?! (this is the last chance for 9 .i.bS+, trall$posing to the main line) 9 . . . �xe7 10 .i.e2 ( 1 0 .i.b3!? de­fends the c-pawn, but leaves the bishop on the wrong sector of the board to defend the kingside light squares. 10 . . . :g8 1 1 liX3 c6 1 2 g3 �h3 1 3 d3 exd3 14 cxcla �7 IS �e3, Mayet-Hanstein, Berlin 1 837, when IS ... lOes 16 0:0-0 �f3 will win the exchange by · ... .i.g2) 10 . . . :g8 1 1 8.3 (of course not I I O-O? �h3 12 1Oc3?! liX6!? 13 �S J:txg2+ 14 �h I :ag8 IS d3 J:t2g6 0-1 . Lonsdale-Littleboy, corr. 1987) I I . . .liX6 (or I I . . .�h3 1 2 liX:3 � 1 3 �S? a6 14 ll:)xc7?! :ac8 + Moreno-Jurgenson, corr. 1(71 ) after which:

a I ) 1 2 h3 is the best idea. White tries to control the weak h3 and g4

Svedenborg 's variation 4 . . . d5 J 59

light squares, 12 . . . � 13 10&3 cS 14 c3 �e2 I S �e2, Lescot­Turian, corr. 1995, when Black could best exploit his advantage on the light squares by IS . . . b6 and . . . h6+, with a likely draw.

al) 1 2 d3 � 1 3 1O&3? �e2 14 �e2 .i.S4+ I S � �d7 16 .i.d2 .i.f3 17 :gl �g4, winning quickly, Vidigal<lasta, corr. 1974.

a3) 1 2 O-O?! � 1 3 Adl .i.g4 14 10&3 .i.xdl (weakening White's light squares by exchanging their defender) I S J:txdl �g4 16 d3 :am 17 .i.e3 �f3+ 1 8 �g2 d4 19 .i.d2 lOit4+! 20 �gl J:txa +, Nymann­Koslov, corr. 1991 .

84) 1 2 c3?! stopg . . . �, but fur­ther exposes the light squares, 1 2 . . . lOes 1 3 10&3 .i.h3 (not bad, but 1 3 .. . .i.g4 gains f3 and d3 for the black knights) 14 liX2 cS I S b3 .i.g2 16 :gl .i.f3; Black has a posi­tional edge because of his custom­ary light-squared bind, Zalomik­Akopian, Bundesliga 1989.

b) Meanwhile, taking the rook is perhaps not as bad as is generally thought: 8 1Oxh8 dxc4, then:

bl) 9 b3! is intcres: 9 . .. � 10 -..,S .i.e6 I I .i.b2 1 (this is not the place for prevarication, 1 1 . . . a6 is strong, 1 2 .gS [but not 1 2 .xb7?? W -1+] 1 2 .. . �d7, best, as Black will capture the knight next

160 Svedenborg 's variation 4 . . . d5

go; 12 . . . lOi4!? 1 3 .eS lOxc2+ 14 �dl .d3 I S .xe6 lOxal is messy) 1 2 .gS 1OI4?! 1 3 .eS!? ( 1 3 lOa3) 13 ... 'Oxc2+ 14 �dl lOxa l IS .xe6 .dS and instead of 16 �3? cxb3 1 7 1Dc3 .c4 1 8 Axa I bxa2, Cardenas-Rodriguez Izq�iero, Cor­doba 1992, simply 16 .xdS lOxdS 17 bxc4 -*.ffi, else 1 8 -*.xa I 1 8 �c I with advantage.

b2) 9 �S+!?, greedily grabbing another pawn, 9 . . . 1Dc6 10 .xc4, with the further choice:

b2 1 ) 10 . . .• dS Black decides to play the ending as he will remass the h8-knight and can then play on the light squares, 1 1 .xdS ( 1 1 1&3?! Ae6 12 .xdS hdS, recap­turing with the bishop allows Black to defend his e4-pawn, 1 3 c4?! -*.e6 14 d4 li:)xd4 I S Ae3 Ab4+ 16 �f1 li:)f5 17 AgS �e7 and Black is win­ning, Crimp-Budovskis, corr. 1999) 1 l . ..li:)xdS 1 2 0-0 lOi4?! ( l 2 . . . Ae6 1 3 lie 1 �d7 14 IIxe4 :Xh8 is more accurate, as Black has a consequent lead in development) 1 3 lOa3 -*.ffi 14 lIel � I S c4 li:)b4 16 IIxe4 -*.xh8 1 7 d4 and for once White is clearly on top, Crimp-Niemand, corr. 1999.

b22) 1O . . . lOi4!? is also possible, 1 1 1&3 ( 1 1 .f7+ �d7 doesn't appear to lead anywhere, 1 2 .c4 [ 1 2 �?! .xh8 1 3 0-0 �d8, Nilsson-Migala, corr. 1998, and here White has to extract his queen by 14 .c4 cS with a plus for Black] 12 . . . �8 1 3 1&3 [White can take a draw with 1 3 .f7+! of course, and he probably should] 1 3 . . . Ae6 14 .c3 as per the main line, Downey­Svendsen, corr. 1990; 1 1 d3? is a blunder, depriving the queen of sev­eral retreat squares, I l . . .Ae6 12 .a4+ [ 1 2 .c3 Ab4] 1 2 . . . bS 1 3 .a6 li:)xc2+ 14 �dl .xd3+ I S � , Brusey-Simons, England

\.992, and Black missed the simplest mate: I S . . . �3+ 16 �el lOxg2+ 17 �dl Ag4+ 18 f3 Axf3 mate but found a more convoluted one!) 1 1 . . .Ae6 12 .c3 (12 .a4+?! �f8!? [ 1 2 . . .• d7 1 3 .xd7+ Wxd7 is also rather good, and easier to calculate] 1 3 c3 tOfS 14 � c6 I S �3 li:)xe3 16 dxe3 �g7 17 0-0 Wxh8 1 8 lie I .g8 19 b3? Ah3 20 g3 .g4 0-1 , Druke-Svendsen, corr. 1989) 1 2 . . . bS! ( 1 2 . . .• dS is also reason­able) 1 3 b4 ( 1 3 lOxbS? lOxbS 14 .c6+ Ad7 I S .c4 1016 16 �3 -*.f8 and .. . Ag7 will gain three pieces for. the rook) 13 . . . aS 14 Ab2 axb4 1 S .xd4 bxa3 16 .xd8+ IIxd8 17 Ac3 b4 1 8 AeS �d7 19 0-0 IIxh8 and White is lost, Downey-Svendsen, corr. 1990, the white a2-pawn is weak, and the a3-pawn is poised for S!C8tness.

b3) 9 0-0 1Dc6 10 .gS 1014 1 1 1&3 -*.e6 12 c3? �2+ 1 3 �hl .d3, completely paralysing the white position, Cabral-Gedult, Paris 197 1 . •

b4) 9 1DcY?! 1Dc6 10 �S -*.e6 1 1 .xb7, Cordir-Barbaut, French ch, Chambery 1994, when 1 l . . .�7 is good, there is a passing threat of ... a6, and . . . 1la7 to cope with, so 12 d4 IIb8 ( 1 2 . . . a6 1 3 &4) 1 3 .a6 .xh8 and the white queen is still not out of the woods. .

8 ••• 00 In the exciting game, Marshall­

Forsberg, New York 192 1 , Black played 8 . . . .i.d1!? 9 lOxh8?! (some­what risky, 9 lOxe1 .xe1 10 .i.xd1+ �d1 [ IO . . . lObxd1?! I 1 .xc1 0-0 1 2 0-0 { 1 2 b3 is almost certainly very good for White} 12 . . . llac8 1 3 .g3+ �h8 14 lOa3 lOhS IS .e3 1Of4 16 d3?! { 16 �h l is much mo� circumspect} 16 . . .• g1 1 1 .g3?? { 1 1 g3 exd3 1 8 �hl should still favour White, although - it is dlessy} 1 1 . . . lOe2+ 0- 1 . Purins-Littleboy, corr. 1983] 1 I • xe1+ �e1 would have left White a pawn up, but facing some difficul­ties exploiting it, not the sort of po­sition Marshall would have found at all interesting, of course! 1 2 d3 lObd1 1 3 dxe4 dxe4 14 -*.gS J:thg8 I S -*.xffi+ �f6 16 0-0 and White has a solid extra pawn. Abramauv­Diablo, Minitel 1991 ) 9 . . . .i.xbS 10 lOc3 ( 10 .e6!? extricates the knight from its tomb, 1O . . . lOc6 I I lOa3 [ 1 1 .f7+? is a waste of a check: I t . :.�d1 12 lOc3 ."h8 1 3 lOxd5 J:tf8. 14 lOxf6+ .xf6 and, with his king stranded in the centre, and f2 weak. White is lost, Edwards­Littleboy, corr. 1985] I I . . ..i.a6 12 lOf7 .d1 1 3 lOgS lOd4 14 .xd1+ [ 14 .17+) 14 . . . �xd1 1 5 Ic3 lOe2 and White's weak light squares give ample compensation for the ex­change, Krustkains-�ittleboy, corr. 1986) 10 . . . -*.a6 I 1 d3 lOc6 12 .g3 .i.d6? (improving White's position, Marshall indicates either 1 2 . . . lOd4, or 1 2 . . .• d1 with the idea of 1 3 . . . 0-0-0, . as better in his marvel­lous • . . . Best Games of Chess', for example [ 1 2 . . .• d1) 1 3 -*.g5 .e6 14 0-0-0 0-0-0 15 dxe4 1Ob4, 16 exd5 .f5 11 J:td2 lOhS 1 8 .i.xe1 lDxg3 19 hxg3 lOxa2+ 20 lOxa2 :Xh8 2 1 lOc3 but White is a bit better here,

Svedenborg 's variation 4 . . . d5 161

Doyle-Nilsson, corr. 1994; perhaps Black's best move is 1 2 . . . �b4, e.g. 1 3 0-0 lOxc2 14 J:tb I hd3) 1 3 �4 d4? 14 lOxe4 .i.e1 IS lOxffi+ -*.xf6 16 �S+ �d1 1 1 -..u+ �e8 1 8 .xh1? (after this Black starts gaining the ascendancy) 1 8 . . .• e1+ 19 .xe1+ lOxe1 20 0-0 hhS=.

9 1Ou7 Once again, 9 lOxhS? is possible,

but must be even worse than before, as Black manages to keep his centre intact: 9 . . . cxb5 10 lOc3 �6 1 1 .g5 � 1 2 0-0 b4 13 �I .i.e6 14 lOe3 �d1 I S c3 lOe2+ 16 �hl .xhS + Krustkalns-Kozlov, corr . 1986181.

9 •••• u7 10 .lle7+ �e7 11 -*.e2

Sometimes White 'undevelops' his one developed piece with 1 1 .i.f1 !? and this makes some sense, as White can thereby avoid weaken­ing his kingside: I I .. .cS ( l 1 . . .J:tgS?! does not threaten the g-pawn here: 1 2 d4 [ 1 2 g3?! is wrong, 1 2 . . . .i.g4 1 3 .i.g2 ltla6!? 14 b3 � I S h3 cS 16 .i.xb4 cxb4 17 a3 IEacS IS axb4 :Xc2 19 lOc3 equal, Nolden­Stummer, corr. 1991] 1 2 .. . exd3 1 3 cxd3 lOa6 [ 1 3 . . . .i.f5 14 h3 �7? { 14 .. . hS } I S g4 .i.g6 16 f4 and White's pawns are on the move, MalmstrOm-Granik, corr. 1994195] 14 .i.f4 .i.f5 I S h3 hS 16 g3 IlaeS 1 1 �d2 �f1 I S lOa3 and Black's

162 Svedenborg 's variation 4 . . . d5

compensation is insufficient, Gunderam-Hempel, corr. 1970) 1 2 d3 ! ( 1 2 g3?! 1Dc6 1 3 .i.g2 lDd4 14 lDa3 .i.g4 I 5 h3 .i.f3 16 0-0 hS ! with a powerful initiative, Harju­Destrebecq, corr. 1978) 1 2 . . . exd3 ( 12 . . . .i.f5?! 1 3 .i.gS! exd3 14 lOc3! :d8 I S O-o-O! d4?! 16 .i.xd3 .i.xd3 1 7 :Xd3 lDbd7 1 8 lDdS+ 1 -0, Crimp-Lescot, corr. 1995) 1 3 .i.xd3 :g8 14 0-0 .i.h3 I S Ile l + Wf7 16 g3 1Dc6, Niemand-Budovskis, corr. 1998, and now White could have kept some advantage by 1 7 .i.e3.

1 1 •. ':g8 12 g3

11 ..• cS If we take stock of the current

situation, we see that White has a pawn more and the bishop pair, but that in return Black has a strong centre and good piece play. Further, White's kingside light squares are sensitive. Keres considered that White has a clear advantage, but I think ;J= is somewhere nearer the truth. Black has tried several moves in this position, the text, an idea of Gunderam's, intends .. . lOc6-d4. Other moves are:

a) 1 2 .. . .i.fS trying to dissuade White from playing d3, 1 3 b3 (bet­ter 1 3 d3 ! exd3 [ 1 3 . . . lDbd7 14 �3 exd3 I S .i.xd3 .ixd3 16 cxd3 lDes 1 7 We2 Wf7 18 f4 lDed7 19 .i.e3 with advantage, Magee-Stummer,

corr. 1990] 14 .i.xd3 .i.xd3 I S cxd3 cS 16 lOc3 [ 1 6 0-0 lOc6 1 7 lOc3 lDb4 1 8 :dl , Hage-Koetsier, corr. 1999, 1 8 . . . d4!? 19 lDe4 lDxe4 20 dxe4 c4, with potent c and d pawns] 16 . . . lOc6 1 7 J:f4 Wd7 1 8 0-0-0 d4!? [ 1 8 . . . llae8 19 11hel lOd4 20 h3 lDf3 2 1 lle3 is easier for White, Melchor-Bomnann, corr. 1995] 19 lDe4 lDxe4 20 dxe4 bS 21 Ilhe I as Gouat-Sireta, corr. 1995, and White is better, but the· black queenside pawns certainly provide practical chances) 1 3 . . . cS 14 .i.a3 Wd6 I S d4 ( I S b4!? b6 16 lOc3 d4! [ 16 . . . lbbd7 17 bxcS+ bxcS, Oren-Jackson, corr. 1994197, 1 8 d4 exd3 19 cxd3 fa­vours White] 1 7 lDa4 lbbd7 18 bxcS+ bxcS 19 0-0 Ilab8 and Black's imposing centre is worth a pawn, Oren-SCn«haud, corr. 1994/95) I S . . . exd3 16 cxd3 lOc6 17 lOc3 lOd4 18 Wd2 bS! ( I 8 . . . llae8 19 Ilac I [ I9 1lae I?! .i.g4 20 f3 :Xe2+! 2 1 :Xe2 .i.xf3 = Melchor-Grivainis, corr. 1992] 19 . . . llg7?! 20 .i.dl b6 2 1 b4 Ilc7 22 bxcS+ bxcS 23 h3 and White is on his way to consolidat­ing, Ruggeri Laderchi-Elburg, corr. training game 1998) 19 b4 c4 20 Ilae I a6 2 1 dxc4 dxc4 22 J:d I llae8 23 :Xe8 :Xe8 24 f3 lDds 25 lDxdS WxdS and Black's activity and extra space compensates for the pawn, Melchor-SCnechaud, corr. 1994/95.

b) 1 2 .. . .i.h3 13 d3 ( 1 3 d4?! lbbd7 14 J:e3 hS I S lDd2 �g4 16 J:fl lDxe3 1 7 fxe3 .i.xfl 1 8 :Xfl �«; 19 We2 h4! 20 ph4 �-�, Ballard­Cano, corr. 1971 172) 1 3 . . . lbbd7 14 dxe4 (14 �?! exd3 IS .i.xd3 llae8 16 .i.e2 Wd8 1 7 f3 lDes 1 8 Wdl lle7 19 b3 Ilge8 with a strong initiative, Kozlov-Destrebecq, corr. 198 1183; 14 J:f4 exd3 [if 14 . . . ItlcS then White should play I S dxe4 rather than I S lOc3?!, Faig-Perez Pietronave, Doea ch 1997, when

1 5 . . . � 16 .*.d2 lOd4 equalises] 1 5 cxd3 d4 16 lDa3 lOd5 17 .i.d2 llaeS I S 0-0-0 White has a plus, N�n­Destrebecq, COlT. 199 1 ; 14 1&:3?! li)e5! IS .i.f4 [ 1 5 dxe4 dxe4 16 .i.e3 .i.g2 17 Ilg I �f3+ is fine for Black] I S . . . .*.81 16 Ilg I �f3+ 17 .i.xf3 .*.xf3 and Black is master of the light squares, Campbell-Steam, COlT. 1994) 14 . . . .i.81 1 5 Ilgl .*.xe4 16 lDa3? ( 16 <i'dl !? is interesting, White frees the e-fiIe for his rook: f6 .. . li)e5 1 7 lie I <i'fl I S � �fg4!? 19 .i.xg4 �g4 20 f3 .i.f3 21 �fI and White has a clear ad­vantage) 16 . . . li)e5, the knight will reach f3, and Black's opening prob­lems will be solved, 1 7 f4 �f3+ I S .i.xf3 .i.xf3 19 Ilfl .i.e4 20 .*.e3 �g4 Black has all the play, Melchor-Niemand, COlT. 1 999.

c) 12 ... .*.g4?! , the idea is to ex­change the protector of White's light squares, but I don't like this move for two reasons, one: Black is It pawn down and should therefore not encourage simplification, and two: Black's bishop controls more squares, and has more scope than its counterpart. 1 3 d3 ! (White should al�ays try to exchange the e4 pawn, 1 3 d4 .*.xe2 14 Wxe2 lDa6 1 5 b3 <i'd7 16 .*.a3 liJc7 � Schirmer­Stummer , COlT. 1 99 1 ; but not 1 3 .i.xg4? �xg4 14 h3?! �5 1 5 b3 cS 16 1&:3 <i'e6 17 �b5 IlcS I S <i'e2 a(). 19 liJc3 �bc6 20 .i.b2 �+, White virtually forced Black to put his knights on their optimum squares, Kott-Elburg, COlT. 1992) 13 ... .*.xe2 (Black can still try to get a knight to f3 via e5: 1 3 . . . �bd7 14 dxe4 dxe4 15 .i.e3 �e5 but then 16 lOd2 and f3 is covered a second time, with a plus) 14 <i'xe2 lDa6 1 5 h3 (White has other promising op­tions: 1 5 dxe4 �e4 16 lOd2 with

Svedenborg 's variation 4 . . . d5 J 63

advantage, or 16 .i.e3 with advan­tage) 1 5 . . . �b4 16 lDa3 exd3+ 17 cxd3 llaeS I S .*.e3 as 19 .i.b6 IlaS 20 Ilhe l lOd7 2 1 <i'd2+ with an ex­tra pawn, Krongraf-Stummer, COlT •

1990. lJ eIJ If White wants to play actively,

then 13 b4!? cxb4 14 a3 bxa3 I S .*.xa3+ <i'fT 16 liJc3 is interesting, 16 . . . liJc6 ( 16 . . . a6!? 1 7 f3 .*.h3 IS fxe4 dxe4 19 .i.c4+ .*.e6 20 .i.xe6+ Wxe6 2 1 0-0 and White has ob­tained activity, but at the cost of his pawn, Nyman-Strelis, COlT. 1990) 17 �b5 �h3 ( 1 7 . . . .*.g4!?) I s lOd6+ <i'g6 19 �xb7 � 20 .i.d I .i.81 (20 . . . llgbS! , first, avoids any prob­lems) 2 1 Ilgl .*.f3? (2 1 . ..�f3+ 22 .i.xf3 .i.xf3 is good, so that if 23 .i.b2 IlgbS 24 1la6? d4 25 .i.xd4 Ilxb7) 22 .*.b2 (22 .*.c5?! �! [22 ... .i.xd I 23 .*.xd4 transposes to 22 .i.b2] 2� .*.d6 �g4 24 1la6 1lgeS White is playing without his king's rook, Bertoni-Koetsier, COlT. 1999) 22 . . . .i.xdl 23 .i.xd4 (23 Wxdl ? �f3 24 1lhl llgbS 25 .*.xf6 llxb7 is very good for Black, Elburg-Tait, COlT. training game 1996) 23 . . . .i.xc2 24 1la6 IlgfS 25 liJc5, Bertoni-Elburg, COlT. 1999, the pin on f6 causes Black serious problems.

Alternatively, 1 3 b3 liJc6 ( 1 3 . . . .i.f3 transposes into line (a) of move 1 2, above) 14 .i.b2 d4 1 5 d3 .i.f3 16 lOa3, Fiorito-De Boer, COlT. 1984, and now 16 .. . �b4 17 <i'd2 llae8 is best.

1J ••• exeIJ However 1 3 . . . liJc6?! is worse, 14

dxe4 �xe4 15 f3 � 16 lOa3 �xe2 1 7 Wxe2 lOd6 I S .i.f4 with advantage, Schlenker-Stummer, COlT. 1991 .

14 ClleIJ?!

164 Svedenborg 's variation 4 . . . d5

This creates a weakness on d3, White should play 14 .i.xd3 ! e.g. 14 ... .i.h3 ( I4 . . . c4 I S .i.e2 lOc6 16 c3 .i.f5 1 7 .i.e3 lhe8 1 8 (H) <i'd7 19 :etl Ag4 20 .i.xg4+ Jhg4 2 1 � and Black's compensation is non-existent, Larsen-Nilsson, COlT. 1991) I S .i.f4 lOc6 16 lOc3, with a white advantage, Edwards­Riegsecker, COlT. 1985; 16 . . . c4 17 Ae2 � , followed by capturing on e2, might represent Black's best chance, bringing opposite-colour bishops.

14 ••• 1Dc6 15 � Ah3 Black has other promising lines,

too: a) I S . . . a6!? intends to set a queen­

side pawn roller in motion, 16 Ae3 d4 1 7 Ad2 Af5 1 8 Ilc 1 /Od7 19 lOc4 bS 20 �S lOceS 2 1 b4!, Kuelaots-Erler, COlT. 1990, 2 1 . . .1Oxd3+ 22 .i.xd3 .i.xd3 23 bxcS Ae4 with equality.

b) I S . . . b6!? Black will attack d3 from another angle, 16 Ad2 b6 17 lOc2 <i'f7 1 8 o-(H) llae8 19 Afl d4

and Black's pressure on d3 gives him good chances, Nitols· Budovskis, COlT. 1994195.

16 .i.dl?! <i'd7

17 li.k2? King evacuation was called for:

1 7 0-0-0 � 1 8 .i.fl .i.xfl 19 Ilbxfl iOO, equal.

17 •• .Ilae8 18 <i'dl iOg4 19 .i.el IlgfB 20 (4 1le7

White's extra kingside pawns arc firmly blockaded and Black has tied White down to passive defence, Elburg-Krantz, COlT. 1990.

9 Other replies to 3 .i.c4 I e4 eS l lllo IS l .*.e4

Although the previous chapter showed that Black can obtain a very good position by 3 . . . fxc4 and 4 . . . d5, players of the black pieces still sometimes search for other, less well-known, methods of conducting the defence. These are as follows:

A l ••• hS B l ••• lllr6 C l .•. lOc6 Otherwise: a) 3 ... d6 is perfectly rdpe<:table,

and transposes to. a Philidor counter­attack, e.g. 4 d4 (4 d3?! is insipid, 4 . . . lllffi 5 /llg5?! d5 6 exd5 illxd5 7 .0 c6 8 o-o?! .*.e7 9 h4?! .d6! 10 LI h6 1 1 1llh3 0-0 � Bclloulou­Kosten, Paris 1993) 4 ... exd4 5 IllgS! (5 Illxd4 fxc4 6 �5+ g6 7 .d5 .e7 8 .*.g5 Illffi = Melchor-: Svcndscn, corr. 1985) 5 . . . lllh6 6 0-0 (6 illxh7!? is risky, 6 . . . /l)g4! 7 illxf8 �f8 8 .xd4 1OC6 9 .d5 .cS with a dangerous lead in development for Black, Sorokin-Maliutin, USSR ch 199 1 ) 6 . . . 1&6 (6 . . . f4!? 7 .*.xf4 .ffi

avoids the opening of the c-file, Glcz-Perez, corr. 1994) 7 exf5 hf5 8 Ll+ Wet7 9 .h6+! .*.xc6 10 illxc6 � 1 1 hh6 gxh6 1 2 .0 and the threat of .f5 causes Black some grief, Zcmitis-Svcndscn, corr. 1997. But the moves .. .

b) 3 . . .• ffi? 4 d4 fxc4?! 5 illxcS 1llc7 6 .*.t7+! w.ct8 7 1&3 d6 8 illxe4 .15 9 1llg3 .ffi 10 0-0 with enormous advantage Monsalvo­Barbosa, corr. 197 1 .

c) and 3 . . .• e7? 4 d4 (but not 4 1&3?! fxc4 5 1lld5 [5 lOxc4 c6] 5 . . .• d6 6 Illg5 c6 7 1&7+? [7 Illt7 �t7 8 1&7+ Wc7 9 illxa8 � is also losing, Black controls the cen­tre, and will win the trapped aB-knight; 7 1&3 is the best move] 7 . . .• xc7 � Grotts-Brieger, USA 197 1 ) 4 . . . exd4 5 e5 ! Wb4+ 6 1llbd2 with advantage, Studicr-Gundcram, corr. 1959, cannot be recommended.

A l ••• hS!?

This attempt at distracting the c4 bishop was introduced by Strautins in 1970. The idea is to play similar

J 66 Other replies to 3 �c4

positions to those of chapter seven but with the presence of a pawn on bS (and a possible bolt-hole for the black king on b7).

4 .i.b3!? Not clearly the most persuasive of

White's possibilities: a) 4 .i.xbS? is just what Black is

hoping for: 4 . . . fxc4 S lOxeS (S IOgl ? 1Of6 6 c4 .i.cs 7 1Oc3 �O :J: Tatls-Jozols, corr. 1989; S .i.c4!? dS?! [S . . .• e7! ] 6 lOxeS dxc4 7 �S+ g6 8 lOxg6, Monsalvo­Leeners, COrr. 1974, when 8 . . . hxg6 9 .xh8 1Of6 should be fine for Black) S . . .• gS! 6 d4 .xg2 7 :fI (not 7 �S+? g6 8 IOxg6 _xhl+ 9 .i.fI hxg6 10 .xh8 Wf7 :J: Schwibbe-Atars, corr. 1973) 7 . . . 1Of6 (7 . . . c6!? is possibly even stronger, 8 .i.a4 �a6 9 c4, Korsmaa-Jackl, corr. 1975, 9 . . .• xh2 with advantage) 8 .i.f4 c6 9 .i.e2 d6 (9 . . . .i.a6 10 .i.xa6 IOxa6 I I .e2 1Oc7 is also pleasant for Black, Atars-Zemitis, corr. 1 970) 10 lOc4 dS l l lOe3 ( l l lOd6+ .i.xd6 1 2 �xd6 �h3 =; I I lOes .i.d6 1 2 c4 � with a good position, Lynch-Nickel, Washington 1972) 1 1 . . ..g6 1 2 c4 �b4+ 1 3 1Oc3 .i.h3 14 :hl .i.xc3+ I S bxc3 � and the black position is preferable, Strelis­Eglitis, corr. 1977/80.

b) 4 �xg8! , swapping this bishop for the unmoved knight on g8 is normally doubtful, but here the tempo gained is important, 4 . . . hg8 and:

bl ) S .e2 is the most popular, S . . .• e7 6 .xbS (6 lOC3 c6!) 6 . . .lbc6 7 .dS (7 1Oc3 fxc4! 8 IOxM gS! 9 g4?! :g6 10 c4 :b8 was an aesthetic, and effective, method of bringing the black rooks into the fray, Gunderam-Strautins, corr. 197 1 ) 7 . . . fxc4! 8 lOxeS!? (per­haps too risky, the game Krustkalns­Eglitis, corr. 1977178, continued 8

.xc4 gS!? 9 1Oc3 g4 10 lOds .g7 unclear; 8 .xg8 is also possi�l

_et. but

just as hai�raising for White, 8 . . . exO 9 .xh7 fxg2 10 :gl [ 1 0 .g6+? "n I I .xg2 .i.b7 1 2 :gl 0-0-0, White is already helpless a.gainst the threat of . . . lOd4, 13 .g6 .xg6 14 hg6 lOd4 I s lOa3 �xa3 etc., Buckley-Johnsrud, corr. 2000] 10 . . . lOd4 I I hg2 .i.b7 1 2 :g3 0-0-0 and Black has a strong attack on the light squares, Melcho� Strautins, corr. 1997) 8 . . .• xeS 9 .xg8 1Ob4 10 Wdl ( 10 lOa3?! lOd3+! 1 1 We2 �a6 with a vicious attack, Grava-Strautins, corr. 1970) 1O . . . dS! menacing . . . .i.g4, with a strong initiative, rather than 1 0 . . .• f5 1 I :el unclear.

b2) S exf5! is the best,

... then S . . . c4 (S . . . 1Oc6 6 .e2 .e7 7

1Oc3 a6 8 lOds .d6 9 c4 bxc4? [9 . . . .i.b7] 10 .xc4 with discovered threats to the g8-rook that shou\d be decisive, Motivay-Atars. corr. 1970) 6 lOd4! (6 .e2 .e7 7 lOd4 a6 8 1Oc3 .i.b7 8orrmann-Micheloud, corr. 1974176, doesn't provide much in the way'. of inconvenience for Black) 6 . . .• gS Motivay-Monsalvo, corr. 1970, when the elementary 7 � a6 8 d3 is crushing.

b3) S IOxeS?! .e7 6 d4 d6 7 1.00 .xc4+, Gabrans-Budovskis, corr. 1974178.

b4) S d4 fxe4 6 .i.gS! .i.e7 7 iOxeS! g6 8 .i.xe7 .xe7 9 .e2 c6 10 llXI2, Kozlov-Franco, corr. 1974, gives White an edge.

4 ••• flle4 Crucial to the evaluation of this

line, but also 4 . . . d6 S d4 fxe4 6 iOxeS!? (6 lClgS dS 7 dxeS c6 8 iOxe4 Strelis-Alberts, corr. 1978 [8 e6 lClh6 � Schild-Bullockus, corr. 1988], when Black should play 8 . . .• c7 [8 ... dxe4?? 9 .i.f7+ �e7 10 .i.gS+ lDf6 1 1 .xd8+ �d8 12 exf6] 9 lC!d6+ .i.xd6 10 exd6 .xd6 which is not too bad) 6 ... dxeS 7 Whs+ �d7 8 .xeS (8 .f5+!? �ppears tempting, but 8 . . . �c6 9 Wxe4+ �b6 10 lC!c3 [ 1 0 .xa8?? .i.b7] 1O .. . .i.b7 when White might just have enough compensation for the piece, but no more) 8 . . . .i.d6 9 .e6+? (9 .i.e6+ �e8 10 .xg7 .i.xe6 1 1 .xh8 "'4 is still reason­able for Black) 9 . . . �c6 10 .dS+ �b6 1 1 0-0 lClf6 and, as the rook is still untouchable, Black is winning, Hebrard-Jester, corr. 1987.

And 4 .. . lC!c6?! S d4 iOxd4 6 iOxd4 exd4 7 .xd4 .e7 (7 .. . .i.b1?! 8 lC!c3 fxe4?! Melcho�Strelis, corr. 1987, 9 iOxbS I[c8 10 .i.f4 is ove� whelming) 8 lC!c3 c6 9 0-0 b4 10 lClbS! cxbS 1 1 .i.ig8 .i.b7 ( 1 1 . ..1[xg8? 12 .dS .xe4 13 .xg8 .i.b7 14 0 is winning) 12 .i.dS

Other replies to 3 .i.c4 J 67

.i.xdS 13 exdS �f7 14 .d3, White wins a pawn, Bertoni-Rozzoni 1999.

5 lClu5 This may not be so wonderful,

White has several other contenders for best move here, all based on the weakness of the two white-squared diagonals, a2-g8 and hI-a8:

a) S lC!c3!? and: a l ) S . . . lC!c6 Black's best move, 6

iOxeS .gS (6 .. . iOxeS? 7 ... S+ lClg6 8 .dS) 7 d4 (this time, 7 lClf7? is false, 7 . . .• xg2 8 I[fI lC!d4 with a strong attack) 7 . . .• xg2 8 Whs+ g6 9 .i.f7+ �d8 10 .gS+ .xgS I I .i.xgS+ .i.e7 12 h4 ltIxeS ( 1 2 . . . iOxd4?? 1 3 0-0-0 cS 14 .i.dS winning material, Alloin-Strelis, corr. 1987) 1 3 dxeS .i.b7 14 1[h3, Pellegrino-Mathieu, corr. 1975, and others, 14 . . . h6! I S .i.d2 b4 and Black has no problems.

al) Of course, not S .. . exO? 6 .xO, forking f1 and a8.

aJ) S . . . lClh6? is not particularly useful: 6 iOxeS .gS 7 d4 (7 .i.dS is also very strong, e.g. 7 . . .• xeS [7 . . . c6 8 d4 .xg2 9 .i.xe4] 8 .i.xa8 c6 9 0-0 dS 10 d4, White will open play with 0, and Black is a long way from capturing the a8-bishop) 7 . . .• xi2 8 I[fI .xh2 9 .i.dS c6 10 .i.xe4 9h3 1 1 iOxbS .i.b4+ 12 c3 .i.aS 13 I[h l .e6 14 "'S+ �f8 IS

J 68 Other replies to 3 JLc4

.i.xh6 +/- Gaard-Strautins, corr. 1991 .

a4) S . . . Ab7?! is also doubtful, 6 lOxeS .gS and:

a4 1 ) With the bishop on b7, 7 LOf7! .xg2 8 Itfl is worth serious consideration, and NCO even rates this +/-. Let's have a look: 8 . . . LOf6 (8 . . . b4 9 lOxh8 bxc3 10 .i.xg8) 9 iOxh8 LOg4 (9 . . . b4 10 lOdS) 10 LOxbS lOxh2 1 1 .e2 ( 1 1 Ac4? LOf3+ 12 We2 .g4 deserves to lose!) I I . . .lOa6 12 d4 Ae7 13 Ae3 ( 1 3 LOf7!? LOf3+ 14 Wdl LOgI I S Itxgl .xgl+ 16 .el .g4+ 1 7 Wd2 is probably good, too) 13 . . . lOxfl 14 .xfl .xfl+ I S Wxfl c6 16 lOc3 dS 1 7 We2 Wf8 and the knight is lost, although White is still better. However, there may be many improvements for White lurking in this analysis, although it is clear that Black still has counterplay.

a42) 7 d4? .xg2 8 WhS+ g6 9 JLf7+ Wd8 10 .gS+ .xgS 1 1 JLxgS+ Ae7 1 2 h4 (followins. 1 2 Axe7+ LOxe7 1 3 lOxbS d6 14 lDg4 1tf8, Jackson-Strelis, corr. 199 1 , Black has the better chances) 1 2 . . . d6 1 3 O-O-O? ( 1 3 LOg4 b4 still favours Black, who has the better structure) 1 3 . . . AxgS+ 14 hxgS dxeS �, Krongraf-Jensen, corr. 1 99 1 .

as) S . . .• e7!? 6 LOxe4 (6 lOdS?! .d6 7 LOgS LOf6 [if 7 . . . JLb7 then 8 LOxe4 .g6 9 LOg3 rather than the 8 LOxc7+?! .xc7 9 LOf7 dS 10 LOxh8 86 1 1 d3 exd3 1 2 .xd3 e4 1 3 WxbS+ .i.c6 14 .e2 .g7 of Rosso­Viola corr. 1999, which is none too clearL8 lOxf6+ .xf6 9 Whs+ g6 10 LOxel [ 1 0 .g4 lOc6) 1O . . .• xf2+ 1 1 Wxf2 gxhS with a level position, Zerbib-Baudoin, corr. 1 997) 6 . . . c6 7 0-0 (7 d4! and if 7 . . . dS? 8 lOxeS dxe4 9 .i.f7+ Wd8 10 Axg8 Itxg8? 1 I AgS wins) 7 . . . dS 8 LOg3 .i.g4 9 h3 Axf3 10 .xf3 White's lead in

development gives him an edge, Dufek-Fritz 3 1996.

a6) S . . . d6? 6 LOxeS! _SS (6 . . . dXeS 7 Whs+ g6 [7 . . . Wd7 8 WxeS is also hopeless] 8 .xeS+ .e7 9 .xh8 LOf6 10 �S LOxdS 1 1 AxdS c6 12 .i.g8 and Black has no compensa­tion for the exchange, Borrmann­Purins, corr. 1983) 7 d4 .xg2 8 Whs+ g6 9 Af7+ Wd8 10 .i.xg6 .xhl+? ( l0 . . . hxg6 1 1 AgS+ Ae7 1 2 JLxe7+ Wxe7 puts up much more resistance, although Black is still losing) 1 1 We2 lClf6 12 LOf7+ ( 12 AgS would have shortened the game somewhat) 12 . . . We8 1 3 lDxd6+ Wd8 14 LOf7+ We8 I S lOcs+ Wd8 16 JLgS .xa I 1 7 Axf6+ leading to mate in a few moves, Alloin-Vlasic, corr. 1988.

b) S o-O?! .f6! (s ... LOf6 6 d4!? [6 LOxeS dS 7 d4 transposes to S d4) 6 . . . exd4 [6 . . . exf3!? is risky, 7 dxeS JLb7 8 exf6 .xf6 9 Itel+ Ae7 10 AdS { IO lOc3!?} 1O . . . AxdS 1 1 .xdS c6 12 Whs+ .f7 with a small plus to Black) 7 LOxd4 dS 8 JLgS c6 unclear, Schmidt-Tiemann, corr. 1978) 6 lOc i (6 .i.dS? c6 7 Axe4 dS) 6 . . .• fS!? 7 lOc3 c6. Jackson­Strautins, corr. 1990, 8 d3! , opening the position with chances of an advantage. ..

c) S d4!? exd4 (S . . . dS is also per­fectly reasonable, 6 LOxeS [6 dxeS!? c6 {6 . . . exf3?! 7 AxdS fxg2 8 Itgl leaves Black facing the twin threats of Axa8 and Af7+} 7 lOd4 a6 8 a4 cS 9 axbS!? cxd4 10 .xd4, Lladors­Esnoala, corr. 198 1 , when the sim­�lest would have been 1O . . . lOc7 t I &3 Ab7 with every chance of de­fending) 6 . . . LOf6 7 0-0 .i.e7 8 lOc3 c6 9 .i.gS, Lescot-GHrd, corr. 1997, 9 ... aS 10 a3 � and White is only slightly better) 6 lOxd4 (6 lOcS!? dS 7 Whs+ g6 8 lOxg6 LOf6 [8 . . . hxg6!?) 9 .eS+ JLe7 10 lOxh8

1Dc6 I I .gS �e6 1 2 .g7 .d7 and although temporarily a rook down, Black has a strong position once he castles long, and captures the bS-knight, Clarke-Strautins, COlT. 1995) 6 . . . �ft) 7 lOxbS dS S �8S (S �f4 �6) S . . . �e7 9 �lc3 {9 .i.xft) �xft) 10 .xdS .xdS I I hdS c6 1 2 �xe4 is messy, but might favour White) 9 . . . c6 and the strong black centre promises equal chances, Grivainis-Strautins, COlT. 1997.

d) S d3 ! is a good move, and may even be White's best here, S . . . �ft) (S . . . exd3 ! 6 .xd3 .e7 7 .xbS � appears stronger) 6 dxe4 and Black will have difficulty develop­ing, Harding-Reed, corr. 1 984.

S •••• gS

6 d4 Just as in Chapter 7 White is

obliged to let his g-pawn go, 6 �f1? .xg2 7 :fl dS (necessary, to threaten . . . .i.g4, and interfere with the t7-knight's defence) and now:

a) S .e2 �g4 9 We3 (Piquero­Cenal COlT. 1 990, 9 f3!? is the alter­native, 9 . . .• xe2+ 10 �e2 �xf3+ 1 1 �el �t7 1 2 �xdS+ �eS 1 3 �xaS �e7, menacing . . . �h4+, Black has excellent compensation for the exchange) 9 . . . .i.h3 10 .e2 �t7 ( l 0 . . . �g4 draws. or forces 1 1

Other replies to 3 �c4 169

f3 as per the previous note) I I �xdS+ �eS 1 2 �xaS .xfl + 1 3 .xfl �xfl 14 �fl c6 which looks fairly level.

b) S �S? �ft) 9 .eS+ (9 � �dS 10 �f1+ �e7! 1 1 .gS .xgS 1 2 lOxgS c6 and .. . h6 will win the knight) 9 . . . �e7! (9 . . . �t7 10 �xdS+ lOxdS I I .xdS+, Kullesenko-Matwienko, Voronezh 1973, when 1 1 . . .�g6! wins because of the threat of .. . �h3, i.e. 1 2 d4 [ 1 2 .xaS? �h3 forces mate ] 1 2 . . . �b4+ 1 3 c3 �h3 14 lD<i2 c6 I S .eS �7 1 6 .c7 :acS) 10 d3 �h3 +, Heiermann-Carlsson, COlT. 19S I .

c) S lOxhS? �B4 9 .xg4, forced, (9 f3? exf3 10 .i.xdS �e7 I I lixf3 �h4+ 1 2 :g3 .gl mate, Taylor-Stewart, COlT. 1992) 9 . . .• xg4 10 �xdS c6 I I �f1+ �7 1 2 .i.xgS �6 and the white king is too poorly defended, and his king­side pieces misplaced.

d) S �xdS .i.g4! (S . . . �h3 9 .e2 �ft) is also playable) 9 f3 (9 .xg4?! .xg4 10 �xhS c6 trans­poses to S lOxhS?) 9 .. . �e7 10 .e2 ).h4+ I I �dl �xf3 12 lixf3 .xe2+ 13 �e2 exf3+ 14 �f3 �ft) I S �xaS �t7 and Black's lead in development makes up for the white bish� pair.

6 ••.• :lg2 7 ... S+ White won very quickly in Roiz­

Fernandez, Oviedo 199 1 , following 7 :fl? �f6 S �3 (S �t7 :gS) S . . . .i.b4 9 .e2 �xc3+ 10 bxc3 dS ( 10 . . . �a6!?) 1 1 �a3?! �a6? (but l l . . .c6! 1 2 f3 .xe2+ 1 3 �e2 as wins the b3-bishop) 1 2 �b4?! ( 1 2 �xdS! b4 1 3 �f1+ �dS 14 �c4) 1 2 . . . �cS?? ( 1 2 . . . �bd7) 1 3 .xbS+ c6 14 .cS 1 -0.

7 ••• g6 8 �f7+

J 70 Olher replies 10 J j.c4

8 ... Wd8 Strautins' move S . . . We?? is

worse: 9 j.gS+ (9 ..,..+ lLlffi 10 .*.gS transposes) 9 . . . lLlffi 10 ..,.. .xh I + I I We2 ( 1 1 Wd2?! allows some resources, I I . . .e3+ 1 2 Wxe3?! [ 1 2 We2! .*.g7 1 3 lLlc3 .g2 14 lLle4 should win for White, Melchor­Asensio. Barcelona 1996] 1 2 . . . j.h6! the only chance, 1 3 lLlc3! [ 1 3 .*.xh6? gives Black a valuable tempo for his counterattack. 1 3 . . .• e l+ 14 Wd3 lLlc6 I S lLlxc6+ dxc6 16 lLld2 .i.fS+ 17 Wc3 b4+ I S Wb3 .xal 19 .*.g7. O'Keefe­Downey. corr. 19S I . 19 . . .• d l 20 .xffi+ Wd7 2 1 lLle4 .0+ 22 Wa4 .xe4 23 j.xhS .e2 winning] 1 3 . . .• xa l [ 1 3 . . . j.xgS+?? 14 .xgS .xa l ? I S lLlds+ mates] 14 j.xh6 [ 14 lLlds+ WdS I S .xh6 .e l + 16 Wf4 .i.a6 is messy] 14 . . . j.b7 [ l 4 . . .• e l+ I S Wd3 .i.b7 16 j.gS WdS 17 .*.xffi+ Wc8 IS j.xhS j.g2 19 lile2 j.fl is less clear] I S j.gS .c l+ [unfortunately. it is too late for I S . . .• el+? as 16 lile2 wins] 16 Wd3 .xgS 1 7 .xgS ltill I S j.xg6 ltgS. Meldrum-Peet, Scotland­Holland, corr. 19S2. and now White could have concluded the game in a number of ways, 19 lLlc6+ j.xc6 20 .eS+ WdS 2 1 .xffi+ WcS 22 j.xh7 being possibly the simplest) I I . . ..*.g7 1 2 lLlc3? ( 1 2 lLld2! is com­pletely decisive. 1 2 . . .• xa l [it is

difficult to believe that Black wil\ last Ions after 1 2 . . .• g2 1 3 lilxc:4 Will 14 l&ffi; and now 12 . . .• 0+ IS easily dealt with: 1 3 lLldxo exO+ 14 Wd2 d6 I S j.dS etc.] 13 lilxe4 forces mate in a few moves. 13 . . . lLlc6 14 j.xffi+ Will I S j.xg7+ Wxg7 16 .ffi+ Wh6. Leiros Vila-Bustos. corr. 19S7. 17 lLlg4+ WhS I S .gS mate) 1 2 . . .• 0+ (the only try. 1 2 . . .• xal ? 13 lLlds+ [ 1 3 j.xf6+ forces mate in six moves] 1 3 . . . WilI 14 j.xffi 1-0. Garcia­Bravo. corr. 19S6) 1 3 lilxo exO+ 14 Wfl ! (probably better than 14 Wd2?! Wxf7 IS .f4 when IS . . . j.b7 16 .xc7 j.c6 is somewhat stronger thanpS . . . d6? 16 .xO c6 1 7 � 1-0. Thorn Leeson-Wyers. corr. 19S5; 14 WxO? Wxf7 IS lLle4 lLldS. Black is defending successfully. Navarro-Baltar Femandez, corr. 19S0) 14 . . . Wxf7 I S .f4 j.b7 (bet-ter than I S . . . lteS? 16 lLldS?? [played the wrong way round! First 16 j.xffi j.xffi then 1 7 lLlds wins] 16 . . . lte6?? [Black returns the favour. 16 . . . ltc4! 17 .xO j.b7 defends. as if IS lilxffi there is IS . . . lIe l+ 1 9 ltxe I .i.xO] 17 lte l 1 -0. Catalan-Herranz, S.pain 19S5) 16 j.xffi j.xffi 17 .xc7 j.a6? (getting his pieces in a tangle. 1 7 . . . j.c6 I S dS .i.xc3. with the king on d2 this would be check. 19 dxc6 j.ffi 20 cxd7 ltds is reason­able for Black. i.e. White must avoid 2 1 Wb?? lilxd7 22 ltdl We7 23 ltxd7+?? ltxd7 24 .xaS?? ltdl mate) I S lLlds ltill ( IS . . . lIeS 19 .f4 lte6 20 /0;7 is also hopeless) 19 lilxffi Wxffi. Torija-Bustos. corr. 19S6. when the most precise move order is 20 .eS+ Wf7 21 .dS+ Wg7 22 lte I winning.

9 .g5+! White enters the ending as the

rook sacrifice does not work here: 9 j.xg6? .• xh l+ 10 We2 h6!

vacating c8 for the king, I I .i.gS+ �c8 (or 1 l .. ..i.e7 =1= Padula-Purins, 197 1n2) 1 2 .i.f5 (alternatively 1 2 lDc3 .xal 1 3 .A.xe4 b4+ 14 �f3 lOf{;!? I S .i.xf{; .i.g7 -/+ Benlolo­SCnec:haud, France 1988; or 1 2 lOd2 b4+ 1 3 �e3 .xal 14 .i.xc4 .el+ I S �f4 lOf{; [ I S . . .• xd2+ 16 �g3 lOfti is more effective] 16 .i.xf{; .i.h6+ winning, Bellin-SCnec:haud, France 1988) 1 2 . . . .i.h6 1 3 .i.xh6 lOxh6 14 .xh6 b4+ I S �d2 .0 16 .e3 1lf8 and Black has a strong at­tack and extra material, Bar-Hilbert, corr. 1989.

Also, 9 .i.gS+? .i.e7 10 .i.xg6 .xh I + 1 1 �d2 ( 1 1 �e2?! .i.a6! 1 2 lOd2 [ l 2 lOt7+ �c8 1 3 lOxh8 hxg6 14 .xg6 b4+ I S �d2 .0 16 .xg8+ �b7 and White is defence­less against the black mating attack] 12 . . . b4+ 1 3 �e3 1Wxa l 14 lOxc4 .el+ I S �f4 .A.b7 -/+, Milev­Metodiev, Primorska 1975) 1 l .. ..i.a6 12 lDc3 e3+!? 13 .A.xe3 .xal leaves White a rook down with no attack, Sanchez Rodenas­Pascual, corr. 198 1 .

9 •••• xgS 1 0 .i.xgS+ .i.e7

/1 1 .i.dS c6 11 .i.xe7+ r 1 2 �t7+?! is bad as ever:

1 2 . . . �e8 13 ' lOxh8 cxdS 14 h4,

Somerset-ozols, USA 1970, 14 . . . �f8 and .. . �g7 wins the knight immediately. I •

Other replies to 3 .i.c4 1 71

1 2 .i.xc4 .i.xgS 13 lOf7+ �c7 14 lOxgS lOf{; is harmless for Black, who enjoys a sUper!or pawn forma­tion, e.g. I S .i.d3 lOa6 16 c3 .i.b7 1 7 0-0 lObs 1 8 Ilel lOf4 19 .i.o :afS, Perez Millan-Canal Oliveras, corr. I992.

11 . . • �e7 13 .i.xe4 Anghar-Vincent, France 1977,

when 13 ... lOf{; 14 .i.g2 .i.b7 I S 0-0 d6 16 00 lOb<i7 or 16 . . . lOa6 is fairly level.

B 3 ••• lOr6

Named after the Argentinian player Morgado who had a number of successes with this variation at the beginning of the I 970s, al­though the main line had already been cited by Leonardo as long ago as 1 575.

4 lOxeS Critical, both 4 0-0 fxc4 (4 .. . dS?!

5 exdS .i.d6 is fun, Feist-Soruco, corr.) 5 lOxeS dS, and 4 d4 fxc4 5 lOxeS dS (Chapter 1 1 ) are relatively innocuous, but 4 lDc3 is quite dan­gerous (see Chapter 1 2).

4 •••• e7 The rook sacrifice 4 . . . fxc4?! was

mentioned in Chapter 7 S d4 lOc6 With S . . . fxc4 we transpose into a

line considered in Chapter 6 (from the move order 3 lOxeS .e7 4 d4

/ 72 Other replies to 3 JJ..c4

lCJ1'6 S JJ..c4 fxe4) whilst S . . . d6 should also be compared to that ma­terial, e.g. 6 .*.f7+! (6 lCJti!? dS 7 lCJxh8 dxc4 8 eS lCJds 9 "'S+ g6 10 lCJxg6 bxg6 I 1 .xg6+ .ti is fine for Black, Marldand-Simm, Eng­land 1967) 6 . . . Wd8 7 JJ..b3 dxeS 8 dxeS+ .*.d7 9 ex1'6 .xe4+ 10 JJ..e3 ! (better than 10 Wf) gx1'6 1 1 lCJc3 _g4 1 2 JJ..e3 JJ..d6 Pogats-Szilagyi, Hungary 19S0) 1O . . . gx1'6 ( l0 . . . f4 1 1 lCJc3) 1 1 lCJc3 .*.b4? ( l 1 .. ..eS) 1 2 JJ..dS .eS 13 .*.xb7 f4 14 .d2 with advantage, Downey-Destrebecq, carr. I992.

6 0-0 Simple development is the most

effective recipe here. The alterna­tives are:

a) 6 lCJxc6?! .xe4+ 7 .e2 .xe2+ 8 Wxe2 bxOO 9 lie I dS =

Kotek-Morgado, corr. 197 1 b) 6 lCJf7!? dS 7 lCJxh8 dxc4 8 eS

lClds 9 "'S+ g6 10 lCJxg6 bxg6 1 1 .xg6+ .f7 :f, Pupols-Morgado, corr. 1970/72, and Black has a strong grip on the light-squares.

c) 6 JJ..f7+ Wd8 7 0-0 lCJxeS 8 dxeS lCJxe4 9 e6 d6 10 lIe l .1'6 is­n't so bad for Black, the ti bishop is not particularly useful, Holthuizen­Diepstraten, Hilversum 1977.

d) 6 lCJc3!? lCJxeS 7 dxeS .xeS 8 0-0 (8 f4?! .cS 9 .e2 fxe4 10 lCJxe4 lCJxe4 1 1 .xe4+ .e7 leads

nowhere, WolI-Morgado, carr. 1974) 8 . . . fxe4 (8: .. JJ..d6!? 9 g3 fxe4) 9 lCJds lCJxdS 10 JJ..xdS c6 1 1 J.xe4 J.e7 ( 1 1 . . .dS 1 2 JJ..d3 [ 1 2 JJ..g6+ Wd8 1 3 :e I .1'6 achieves nothing) 1 2 . . . JJ..e7 13 lie I .1'6 14 'ehS+ Wd7 and Black's position is uncom­fortable, Grivainis-Diepstraten, carr. 1977, but the king move cannot be avoided, as 14 . . . g6?? loses to I S JJ..gS) 12 lIe l .1'6 1 3 .d3 ! (John Nunn prefers this in NCO; 13 c4 0-0 14 J.e3 ;t Gunderam-Grivainis, carr. 1 970) 1 3 . . . g6 ( 13 . . . dS 14 JJ..xh7 Wd8 might be better, but is hardly enticing) 14 J.dl lIfS ( l 4 . . . 0-0? IS J.c3 .xf2+ 16 Whl threatens JJ..xg6, and wins quickly) I S..JIe2, preparing to double rooks along the e-fiIe, Castelli-Grivainis, corr. 1970.

6 ••• r�e4 Rather than 6 . . . lCJxeS?! 7 dxeS

lCJxe4 8 lCJc3 lCJxc3 9 bxc3 .cS, Schirmer-SCnechaud, carr. 1992, 10 .e2 ( 10 .d3 !? with advantage) 1O . . . b6!? 1 I JJ..e3 ! .xeS?, Kozlov­SCnechaud, corr. 1994, 12 :ael and the multiple threats ( 13 �S+, and 13 JJ..gS to name but two) force im­mediate resignation.

7 JJ..t7+ In the game Eckenfels-Morgado,

corr. 1967, Black had good play fol­lowing 7 lCJxOO dxOO 8 f3 J.rs 9 fxe4 .J.xe4 10 lCJc3 O-O-O!, but the best move is certainly 7 lCJc3! with a likely transposition to d), above. 7 lCJf7!? can also be played, of course, 7 . . . dS 8 JJ..xdS lCJXdS 9 lCJxh8 � 10 f3? JJ..d6 1 1 f4 JJ..g4 12 .el .e7 13 c4 lCJxd4! 14 cxdS lCJc2 1 S .c3?! ( I S .g3) I S . . . JJ..cS+ 16 Whl lCJxal and Black has the edge, Alberts-Grobe, corr. 197 1 .

7 ••• Wdl I lCJ�c6+ 8 f4?! lCJxeS 9 CJ.eS .xti 10 ex1'6

gx1'6 I I "lCJc3 dS 1 2 JJ..e3 lIg8 1 3

.d2, continuing by doubling rooks on the f-file with some pressure, Grobc-Padula, COlT. 197 1n2, but Black is better.

.... bxc6 8 .. . dxc6!? 0 "

9 �b3 d5

10 �15 b6 1 1 .i.f4 -*-14 11 .d1 .e6 131llal

White is a little better, Alvarcz­Morgado, COlT. 1969.

C 3 ••• �

Strictly speaking. this is an Italian Game ( I e4 eS 2 �t3 � 3 .i.c4 5), and therefore I will only deal with it briefly.

4 d4 I suppose 4 d3 is quite reasonable,

4 . . . d6 (or 4 . . . �ffi S 0-0 �cS =) S � iDh6 6 �7!? �g4!? is un­clear Schmidt-Bosc, COlT. 1 964;

Olher replies 10 J .i.c4 1 7 J

6 . . . lhh7 7 -...S+ �7 8 .g6 IOd4 may also be possible; but 4 hg8? giving up the bishop pair, is terrible (actually, I must admit to having played this once, as a lad!) 4 . . . lhg8 S ex5 .ffi (S . . . dS ! is simpler, as 6 ee2 h5 7 1&eS?? 1Od4) 6 0-0 (6 �3 .x5 7 0-0) 6 .. . dS 7 LI h5 �, Donny-Strelis, COlT. 1971n2.

4 •• .exd4 a) The position after 4 . . . fxe4?! S

l&eS dS 6 .i.bS ed6 (6 . . . �ge7 is possible) 7 0-0 .i.fS 8 hc6+ bxc6 9 t3 �ffi 10 fxe4 �xe4 1 1 �3 is very unpleasant for Black, Elburg­Raijmackers, COlT. 1985.

b) 0 And Gunderam's move 4 . . .• e7?! is too artificial: S �gS �ffi (S . . . -..,..+ 6 �bd2 cxd4 7 ex5 with advantage Padula-Vitols, COlT. 1974) 6 �3 -..,.. (6 . . . fxe4? 7 10JS .d6 8 dxeS l&eS 9 1&eS .xeS 10 .i.f4 .xb2 1 1 Jlbl wins material, Rost-Jurack, COlT. 1987) 7 hffi (7 dxeS? 1&e4 [7 .. .• xc4 8 exffi with advantage Zsc:hom-Schwibbe, con:. 1 979] 8 .i.b3 [8 .dS I&gS 9 I&gS .xb2 may just be good for Black] 8 . . . l&c3 9 bxc3, Fiorito-Khulmann, COlT. 1982, 9 . . .• e4+ and Black has good chances) 7 . . . gxffi 8 J.b3 cxcM (8 ... fxe4 9 � .xcM 10 �) 9 a3 .cS 10 �S White has the initiative, Gucnzcl-Feichtqcr, COfr. 1 986.

1 74 Other replies to 3 j.c4

c) 4 . . .• f6? S j.gS .g6 6 exfS .xfS 7 dxeS (7 O-O?! e4 8 Ile I . Atars-Gunderam. corr. 1970. 8 . . . lLlge7! 9 j.xe7 j.xe7 appears to favour Black. who will play .. . dS on his next go if the knight retreats) 7 . . . h6 8 j.d3 .e6 9 j.e3 lLlge7 10 0-0 gS!? Dravnieks-Gunderam. corr. 197 1 . 1 1 lila3!? a6 (else lObS) 12 ltXi4!? lLlxd4 13 j.xd4 lilc6 14 .i.c4 .g6 IS e6. exposing the black king position.

S IOxd4 Otherwise. S O-O!? fxc4 6 1Oxd4 lLlf6 7 j.gS

lLleS!? (if 7 . . . dS 8 j.xf6 .xf6 [8 .. . gxf6!? 9 �S+ �e7 is certainly playable. though] 9 lLlbS might be good for White) 8 j.b3 c6 9 c4! h6 (9 .. . .i.cS? 10 lilc3 lild3? 1 1 lLlxc4

with advantage) 10 j.xf6 .xf6 1 1 lilc3 00 12 lLlxc4! (White proves himself the eql.Ull of the position) 1 2 . . .• xd4 1 3 Ilel �8! 14 j.c2?? (Oops! 14 1le3 .xb2 I S 1lxd3 left everything to play for) 14 . . . lOxe l ! I S .xd4 IOxc2 16 .dl lOxa l 1 7 .xal d6 1 8 .dl �c7. winning eas­ily. Fortc-Kosten. Clermont-Ferrand 1994 .

S ••• lLlxd4 6 .xd4 rxe4!? 7 j.x,,?!

7 0-0 is better. 7 . . . lLlf6 8 j.gS c6 9 j.xf6 .xf6 10 .xe4+ j.e7 I I lilc3 d6 ( l 1 . . .dS?! 1 2 j.xdS j.fS 1 3 .e2 wins a pawn. although Black is still alive after 13 . . . <HH>!) 1 2 1lfel j.fS �e3 �7 14 1le2 llhe8 I S Ilae I and Black is tied up. GUrd-Schott. corr. 1989.

7 •• .Ilxg8 8 .dS llb8 9 ... S+ g6 IO .e5+ Wf7!

1O . . .• e7?? 1 1 .xh8 +/- Kozlov­Diepstraten. corr. 1977179.

1 1 0-0 Not 1 1 .xh8? j.b4+. 1 1 j.gS

.e8 1 2 .xc7 j.g7 =. 1 1 •• ig7?! 1 1 .. ..f6 is equal. 12 .d5+. Lewis-NN. London

1840. 12 . . . �f8 13 f3 .e7. Black's king is awkwardly placed. but this is not overly serious.

10 3 exfS

I e4 e5 l lbo f5 3 exf5 Currently one of the most popular

variations, White takes the proffered pawn even though it will entail the displacement of his king's knight. The move was cited by both Polerio in IS7S, and by Greco in 1 62 1 , and was later analysed by Bilguer in his famous Handbuch ( 1843).

3 ••• e4 The critical reply, although Black

has also tried other moves here: a) 3 . . . d6 is reasonable, transposing

into Philidor's Defence. It is not really within the scope of this work, but here is a brief overview: 4 d4 (else Black will simply recapture on 5) 4 . . . c4 S lbgS .*.x5 (S . . . lbf6 6 lbe6 [6 lbc3 .i.x5 transposes to S . . . .i.x5] 6 . . . .i.xe6 7 fxe6 dS S g3 cS 9 .i.e3 � 10 lbc3, Gonsalves­Kozlov, corr. 1 9S6, when Black can

try 10 . . .• xe6 I I dxcS lbbd7) 6 f3 (6 g4!? also 'promises some advan­tage, 6 . . . .i.g6 7 lbe6 .d7 S lbxm �m, Littleboy-Borrmann, corr. 1976) 6 . . . lbf6 (6 . . .• e7 is safer, 7 fxc4 .i.xc4 S lbxc4 .xc4+ 9 .e2

.xe2+ 10 .i.xe2 lbc6 I I c3 lbf6 12 0-0 0-0-0 ;t because of the bishop pair, Hayward-Downey, corr. 1990) 7 lbc3 dS (7 . . . exf3 S .xf3 .cS 9 .i.c4 is unpleasant) S fxc4 dxc4 (S . . . lbxc4 9 lbcxc4 [9 lbgxc4 dxc4 to .i.e3 -*.d6 I I .lc4 ;t, Vitols­Maly, corr. 197 112] 9 . . . dxc4 [9 .. .• e7? 10 .i.bS+ which is almost winning for White, Hristodorescu­Pessi, Bucharest 1993] to .i.c4 lbc6 Svendsen-Nilsson, corr. 1991 , 1 1 .i.e3 again, clearly better for White) 9 .i.c4 lbc6!? (9 . . . .i.g6 10 0-0 lbc6?, Jaunozols-Eglitis, corr. 1970, I I lbe6 .d7 1 2 1Ixf6! gxf6 1 3 lOdS is devastating) 10 -*.17+ ( 10 lbf7? leads to a draw, but no more, 10 . . .• xd4 I I .xd4 lbxd4 1 2 0-0 JIgS 13 lbgS JlhS 14 lbf7 JIgS etc., GUrd-oren, corr. I 994/9S) to . . . �e7 I I dS h6 12 .i.e6 -*.g6 1 3 lbgxc4 lbxc4 14 lbxc4 .i.xc4 I S .e2 .d6 16 .xc4 �+ 17 .xb4+ lbxb4 I S �d I lbxdS 19 .i.xdS lidS 20 c4 c6, almost equalising, Magee-GUnl, corr. 1999 ,

b) 3 . . . lbc6!? 4 d4 (or 4 .i.bS .i.cs S .i.xc6? dxc6 6 lbxeS .i.xfS [6 . . . .i.xf2+!?] 7 �S+ [7 0-0] 7 . . .• S lbxg6? bxg6! 9 .xhS .e7+ 10 �dl .l.xf2 I I .xgS+ �d7 1 2 .c4 JleS 0- 1 , . Schlechter-Chigorin, I S7S) 4 . . . exd4 (4 . . . c4? S lbes lCsf6 6 .i.e2 menacing the powertPl .i.hS+) S lbxd4 lbxd4 6 .xd4 .f6 7 .e3+ (7 .i.e3 ;t.) 7 . . .• e7? (7 . ./ac7 S lbc3 .xf5) S lbc3 lbf6 9 .i.d3 with a pawn more, Rossell-Koetsier, corr. 1999.

1 76 3 exf5

c) 3 . . .• ffi?! 4 d4 exd4 (4 . . . e4?! S �S �7 [ugly. but if S . . .• xf5 6 .*.c4 with advantage] 6 �S+ g6 7 fxg6 hxg6 8 .gS and White enjoys a sound extra pawn. Tener­Dreibergs. corr. 1965) S lCJxd4 �7 6 lCJc3 .eS+ 7 .*.e3 lCJxfS 8 lCJxf5 .xfS 9 .*.d3 with a lead in develop­ment. Miljkovic-Baier. corr. 1998.

d) 3 ... .i.cS?! is quite popular. 4 lCIxeS .*.xf2+ S �f2 �+ 6 �o! (objectively stronger than 6 g3 .d4+ 7 �g2 .xeS 8 .e2 d6 9 .xeS+ dxeS 10 g4 although this is simpler. and also assures an advan-tage. Vitols-Vashegyi corr.) 6 . . . bS (the possibility of playing .. . b4 provides Black with extra resources. and so this has replaced 6 . . . b6 when 7 d4 lCJffi 8 lCJc3 .t.b7+ 9 dS 0-0 [9 .. . .*.xdS+ 10 lCJxdS .e4+ 1 1 �f2 .xfS+ 1 2 .0 is also hopeless] 10 g3 lCIxdS 1 1 lCIxdS lhf5+ 12 .*.f4 1 -0. Gaberc-Rosso. 1970). Never­theless. the line is still completely unsound. 7 d4 lCJffi 8 lCJc3 .*.b7+ 9 �e2 b4. Tiemann-Elburg. corr. 1993. 10 lCJbS Black has a few checks. but no real compensation for the piece.

e) 3 . . . lCJffi 4 lCJxeS transposes to 3 . . . lCJffi 4 exfS in Chapter 6.

White has four main responses to the attack on his knight:

A 4 lCJe5 B 4 .el C 4 lCJd4 D 4 lCJgl

A 4 lCJe5 lCJf6

Very occasionally. Diepstraten' s 4 . . .• gS!? sees the light of day when. S d4! (S g4!? .e7 6 lCJc4 dS 7 �3 d4 8 lCJc4 hS. unclear. Krantz-Diepstraten. corr. 1991 ) S . . .• xfS 6 .*.c4 lCJc6! with:

a) 7 g4! .ffi 8 lCJrl (now this is possible) 8 . . . dS! 9 .*.xdS lCIxd4 10 lOxh8?! ( 10 lCJd2! is more_prudent. e.g. 10 . . . lCJh6 1 1 lCIxh6 .xh6 1 2 � � 13 c3 and White has an extra pawn) 10 ... lCJO+ 1 1 �fl .*.xg4 1 2 .t.f1+ �e7 13 .dS .t.h3+ 14 �e2 lCJd4+ 1 S �e3 ( I S �d2 lCJO+ 16 �e3 �+ 17 �e2 lCJd4+ probably leads to a draw by perpet­ual) I S . . . lCIxc2+ 16 �d2 :d8. quite unclear. Geervliet-Diepstraten. Hilversum 1969.

b) 7 lCJrl? is a typical mistake. 7 . . . dS 8 lCJxh8 (8 .*.xdS is no better. 8 . . .• xdS 9 lCJxh8 .t.e6 10 .i.f4 0-0-0 1 1 lCJc3 .*.b4 12 a3 .*.xc3+ 13 bxc3 lCJffi. winning easily. Bakker­Diepstraten. Hilversum 1968) 8 . . . dxc4 9 dS lCJb4 10 lCJa3 lCIxdS 1 1 lCIxc4 .*.cS 1 2 .*.e3 lC1xe3 1 3 fxe3?! .gS 14 .d2 .*.e6 IS b3 .*.xc4 and Black can soon recuperate the h8-knight. and win. Oordijk­Diepstraten. Hilversum 1969.

c) 7 .*.rl+ �e7 8 g4 .ffi 9 .*.xg8 (rather than 9 h4?? lCIxeS 10 .*.gS �O+. winning a piece. Eydan­Svendsen. corr. 1990) 9 . . . llxg8 10 lCJc3 .e6 1 1 0-0 dS 1 2 0 lCJxeS 1 3 dxeS c6 1,\ fxe4 d4 1 S .xd4 .xg4+ 16 �h l �e8 messy. GUrd-MalmstrOm. corr. 1994/9S.

d) 7 "lCIxc6 dxc6 is nothing special.

5 .i.e2! The critical response, threatening

the disruptive check on hS. a) White can also try to hang onto

his extra pawn by 5 g4?! although this must be dubious, as Black has more than one strong reply, S . . . d6 (also, S . . .• e7!? is strong, 6 1Oc4 dS 7 lDe3 d4 8 lOg2 [8 �?! hS 9 gS lOlls 10 h4 .i.xfS and Black has a massive space advantage, Keiser­Jackson, corr. 1982] 8 . . . hS!? [the logical 8 . . . 1Oc6 must be good, too, 9 gS lOlls 10 Whs+ Wd8. Black is better] 9 gS [9 lOf4 may be stronger, 9 . . . lOxg4 10 lOg6 .cS Criel­Tiemann, corr. 1980, which is un­clear, as I I lOxh8?? is impossible, 1 1 . . .d3 and f2 is undefendable] 9 . . . lOg4 10 h4 .i.xf5 I I .i.c4? lDes 12 .i.e2 d3 and White is in a bad way, Vashegyi-Grobe, corr. 1970; or S . . . .i.cS? which is much worse, objectively, but good fun, 6 gS 0-0 7 gxffi .xffi 8 lOc4?, Danan-Gedult, Paris 1969, and Black already has a forced win! [8 d4 exd3 9 lOxd3 .xf5 10 .e2 defends comfortably] 8 . . . .i.xf2+! 9 We2 [9 �f2 .xfS+ 10 We3 1Oc6 forces mate in two] 9 .. .• xf5 10 d3 .g4+ I I Wd2 e3+ winning tile queen) 6 1Oc4 dS (6 . . . hS? is weak, 7 gS lOlls 8 1Oc3! �c3 9 dxc3 .i.xf5 10 .dS! [ 10 lDe3 Neukirch-Alvarez, corr. 1971, 1O . . . .i.e6 equal] 10 . . .• c8, Gubats-

3 exf5 / 77

Strautins, corr. 1970, when I 1 lDe3! is critical, 1 l . . ..i.d7 [ 1 l . . ..i.h7 12 .i.h3! .xh3 13 .xb7] 12 .xe4+ Black is in a bad way) 7 gS (7 lDeS? .i.d6 [this is good, but 7 . . .• e7! is almost winning, the knight has no retreat] 8 d4 exd3 9 lOxd3, Morgado-Atan, corr. BetinA Memorial (superfinal) 1976, and now, the typical 9 . . .• e7+ 10 .e2 lOxg4 wins Black a pawn; 7 lDe3 d4 8 lOg2 is uncomfortable for White, but playable) 7 . . . dxc4 8 gxffi .xffi 9 .i.xc4?! (9 Whs+ Wd8 10 d3 leads to rough equality) 9 . . . .i.xf5 10 .e2? 1Oc6 I I lOc3, Vereb-Eberth, Agria Eger 1999, l l . . .lOd4 12 .dl lOf3+ and wins!

b) A more sensible idea is 5 lOg4 again similar to a King's Gambit (2 . . . lOffi variation), S . . . dS (but if Black wants some fun M'can try S . . . .i.cS!? 6 1Oc3 dS 7 lOxffi+ [7 d3 is safer] 7 . . .• xffi 8 Whs+ [8 lOxdS .i.xf2+ 9 �f2 .xfS+ 10 Wel .xdS favoun Black] 8 . . . g6 9 fxg6 [9 .i.bS+ c6 10 tl)xdS!? .eS!] 9 . . . bxg6 10 .xd� .xf2+ I I Wdl .i.g4+ 12 .i.e2 .i.xe2+ 1 3 lOxe2 "'xg2 14 .eS+ Wd7 I S lIel [ I S "'g7+ Wc6 16 .ffi+ forces a per­petual] I S . . . lIhS, unclear, Melchor­Tiemann, corr. 1991) 6 lOxffi+ .xffi 7 Whs+ "'fl 8 .xfl+ Wxfl 9 1Oc3 c6 10 d3 ( 10 g4?! hS 1 1 d3 exd3 1 2 .i.xd3 bxg4 clearly favoun Black, Backhuijs-Den Hertog, corr. 1986) 1O . . . exd3! ( 1O . . . .i.b4 I I dxe4 .i.xc3+ 12 bxc3 lIe8 13 .i.e3 lIxe4 � Melchor-Svendsen, corr. 1990; 1O .. . .i.xf5 1 1 dxe4 dxe4 12 .i.c4+ Wg6 13 .i.e3 Nilsson-Elburg, corr. 1988, is a little better for White) I I .i.xd3 1Od7! 12 .i.f4 lOcs 13 g4 lOxd3+ 14 cxd3 .i.cS ( 14 . . . g6!? I S fxg6+ bxg6 16 gS lIb3 17 d4 .i.b4 gave Black reasonable play in GromitChess 1 . 1 - GNU Chess 4.0

/ 78 J exf5

computer game 1997) 1 S o-O?! hS 16 1014 Chess 4.7-Levy, Toronto 1978, and now Black missed 16 . . . .i.e7! 17 h3 bxg4 18 bxg4 .i.xf5! 19 gxf3 1lh4, skewering the bishop against the knight, and thereby gain­ing an appreciable advantage.

c) If 'White does nothing, Black equalizes immediately: S d4 d6 6 lDc4 (6 �g4 hf5 7 �3 tnns­poses) 6 . . . .i.xf3 7 �3 .i.g6 8 c4 (8 dS .i.e7 9 .i.e2 0-0 10 tOe3 �bd7 =

Colas-Destrebecq, France 1978; 8 h4!? dS 9 c4 lOc6! 10 lik:3 dxc4 1 1 dS �S 1 2 �c4 �c4 1 3 .i.xc4 a6 14 .i.e3 .i.d6 I S hS .i.f5 16 1lh4 0-0 ; Crouch-Kindennan, London 1978) 8 . . . .i.e7 9 .i.e2 0-0 10 lik:3 lik:6 1 1 0-0 .d7 with balanced chances, Gibert-Melchor, Spain 1999.

d) S d3?! is simply met by S . . . d6 6 � dS (6 . . . .i.xf3 =) 7 I&d2 .i.xf5 8 dxc4, Bychkova-Mikhailova, St Petersburg 1997, 8 . . . �c4 (or 8 . . . dxc4) 9 �c4 .i.xc4.

At this point Black has a choice: AI 5 ••• d6 Al 5 ••• .i.e7!? A3 5 ••• d5? A4 5 •• ..i.c5?! The themes to bear in mind, in all

these lines, are that: White can win the h8-rook, but

that his knight then becomes trapped on h8. If Black can success­fully recapture this he may be bet­ter. On the other hand, White can often save this piece, by g4-gS-g6, and it is then a question of whether Black has enough compensation for the exchange.

Otherwise, White can play �d6 and then capture the bishop on c8, but it is normally a mistake, as the knight will then also be trapped.

AI 5 ••• d6

Traditionally the main move, and the most forcing, but bearing in mind that White wants to play � and �f1 anyway, it may be that Black has more useful moves.

6 .i.b5+ We7 7 �f7 7 � .i.xf5 8 d4 � NCO, but

8 . . . .i.e6 9 �3 dS doesn't look par­ticularly wonying for Black.

7 •••• e8 8 �xb8 a) It is a moot point whether this

is stronger than 8 1&:3, as follows:

a l ) 8 . . . �hS 9 lOds+ (the most exciting, although 9 �h8 �ffi 10 84 c6 [Betin! gives 10 . . . �! 1 1 gS 1&14 12 gxffi+ Wd8 when, after 13 f7 .eS 14 .g4? { 14 Wfl hf3 IS Wg2!?} 14 . . . .i.xf3 IS �+, Diepstraten-Eglitis, corr. 1977, Black tould, indeed, have obtained a powerful position by I S . . . Wd7 16

cM1 �e7 1 7 .g3 lhhS, but White has stronger tries at move fourteen] 1 1 gS � 12 .g4 �d8, Svendsen­Diepstraten, COrT. 1992, is also promising)

a l l ) 9 . . . Wxt7! 10 .xhS+ g6 1 1 fxg6+ �g7 12 �c7

( 1 2 b3?! is less accurate, 1 2 . . .• eS 1 3 .xeS+ dxeS 14 �c7 �d6 I S �aS hxg6 16 �b2 �f5 1 7 0-0-0 :cS!? and, provided he captures the knight soon. Black will have reason­able prospects, Miraglia-Ruggeri Laderchi, COrT. 1997) 12 . . .• eS (I prefer 1 2 . . .• xg6! 1 3 .xg6+ Wxg6 [ 1 3 . . . hxg6?! 14 b4 �e7 I S �b2+ .*.ffi 16 �xffi+ Wxffi 1 7 �aS lOa6, Evans-Reinke, COrT. I 994/9S, and the white knight escapes, I S bS �b4 19 0-0 �c2 20 :acl etc.] 14 �xaS lOa6 IS d4 �e7 16 �f4 �d7 1 7 IOb6 axb6 I S a3 ; in such a position, where the white rooks have no open files, the black pieces will dominate, Le Deuff-Destrebecq, COrT. 19S5) 1 3 .xeS+ dxeS 1 4 ltlxaS ( 1 4 gxh7 � I S �xaS �d6 [trapping the knight, I S . . . �!? is more ambi­tious, 16 0-0 { 16 :b l !? �c2+!? 1 7 �dl �b4 IS �7 �g4+ 19 �el 00+ 20 � �cS is interesting} 16 . . . �xc2 1 7 :bl �d6 I S b3 �f5 19 �b2 lhaS and Black is on top, Zeidaks-Budovskis, COrT. 1974] 16 c3 [16 c4!? b6 1 7 d3 exd3 I S �e3

3 exf5 1 79

�e6 19 �b6 axb6 is wild] 16 . . . .*.g4 1 7 h3 �f5 I S g4 �e6 19 b3 lhaS 20 �b2 Wxh7 21 0-0-0 :18 22 :h2 �cS Black has good play, Evans-Sireta. COrT. 1994/94) 14 . . . �d6 I S 0-0 ( I S c4!? [time is of the essence! ] I S . . . lOa6 16 b4, Salenga-Laurinavivius, COrT. I99S, 16 . . . �b4 1 7 0-0 00 is promising for Black, I S :b I Wxg6 19 �b6 �g4! ; I S gxh7 transposes to the previous note) I S . . . hxg6 16 c4 lOa6 (if 16 . . . b6 1 7 a4) 1 7 b4! White must find a way to free his knight, 1 7 . . . �xb4 ( l7 . . . �d7 IS cS �e7 is also feasible) I S �b2?! ( I S :bl !? �d6 19 :bS) I S . . . �d6 19 cS �cS 20 .h3 b6 2 1 :tb I .i.f5 22 �b6 axb6 23 lhb6 :as and Black has some drawing chances, Charushin­Schrcyer, COrT. 19S6.

a 12) 9 . . . �7?! is clearly worse, 10 .xhS :gS ( l 0 . . . c6!? is no better, 1 1 �f4 :gS 12 � lOa6 13 .xh7! �e7 [ I 3 . . .• x fi 14 �gS .ffi { 14 . . .• dS I S �} I S .xgS .xgS 16 .xf8 .xg2 17 .fi+ �dS I S :n must be good for White] 14 �fgS with a significant advantage, Zeidaks-Grobe, COrT. 1970) 1 1 ffi! (1 1 �f4 �e7 12 � �6 1 3 g4!? �S 14 �fxdS .xhS I S gxhS �xdS 16 :gl �ffi 1 7 h6 �c6 I S �g7 �xfS-Betini, is not so clear) I I .. .lOa6 ( I l .. .c6? is hopeless, 12 lOes+! dxeS [ 1 2 . . .• xeS 1 3 .fi+] 1 3 fi .e6 14 fxgS= •• xgS, Vitols-Grobe, COrT. 1973, I S .g4+ .e6 16 .xe6+ Wxe6 17 �7+; and 1 1 . . ..e6? loses immediately, 12 �f4 .xffi 1 3 �3+ �c6 14 .xcS .xfi I S �; but I I . . .�! is Black's best try, 12 �f4 [ 1 2 .g4+ .e6 1 3 .xe6+ Wxe6 14 �c7+ Wxfi 1 S �aS �b4 is not too clear] 12 . . . � [ 1 2 . . . gx.ffi?? 1 3 .f5+ �e7 14 �S+ Wxfi IS .xffi mate, Vitols-Gebuhr. COrT. 19751 1 3 11'h3+

180 3 exf5

�c6 14 1I'c3+ �d7 I S 1I'xd4 1I'xt7 16 .xe4 gxft) and White is only a ;wo up) 1 2 0-0 ( 1 2 fxg7? Axg7 13

xh7 .xf7 14 �ft)+ .xft) I S 1I'xgS c6 is good for Black, and 1 2 �S+? comes one move too soon, 1 2 . . .• xeS 1 3 11'f7+ �c6 14 1I'xgS Ae6 IS �7+ �b6 which is un­clear) 12 . . .• e6 ( l2 . . . c6? 1 3 �S+ dxeS 14 t7 .e6 I S fxgS= • • xgS 16 �3 analysis of Keres 1 -0, Popa­Steinhauer, corr. 19S5, although, in the game Geisler-Reichardt, corr. 1992, Black reached the same posi­tion and decided to play on for a while, 16 . . . Ad6 before resigning) 1 3 f.3 h6 14 fxg7 ( 14 fxe4!? g6 I S Wh3 1I'xh3 16 ph3 c6 1 7 liJe7 of- . fers less advantage) 14 . . . Axg7 I S fxe4 White is better, the black king will have difficulty finding a safe resting place, Pape-Heap, corr. 1994/96.

a2) S . . . g6!? 9 fxg6 (9 �hS gxhS 10 d3, Kell-Nortje, corr. 1997, 10 ... j.g7 I I dxe4 .i.xhS 12 AgS fa­vours White) 9 . . . bxg6 10 �xhS ( 10 j.xg6 :gS I I AhS J:txg2 1 2 �d6 1I'xhS l 3 �cS+ �d7 14 1I'xhS �hS I S �e4! �cS [ I S . . . �6 1 6 �g3 �g3 1 7 hxg3 .i.cs may be stronger) 16 �g3 �f4 1 7 d4 �3 I S J..e3 :gl+ 19 J:txgl �xgl 20 � when Black experienced cer­tain difficulties stopping the white pawns, Reinke-Melchor, corr. 1994/96) 10 .. _:SXhS 1 1 1I'e2 Ag7 ( l l . . .dS?! 1 2 We3!? [ 1 2 d3 Ag7 l3 dxe4 is strongest) 1 2 . . . Ag7 l 3 1I'cS+ �e6 14 �bS �6 IS .xc7 �c7 16 �c7+ �f5 1 7 �xeS �eS, unclear, Stummer-Schlenker, corr. 199 1 ) 1 2 �e4 1I'xhS may not be to everyone's taste, but is per­fectly playable, 1 3 �ft)+ ( 1 3 �gS+ Wd7 14 11'e6+ �c6 I S 11'c4+ �d7 16 11'f7+ �dS 1 7 0-0 is a little awkward for Black, but no more,

Lindeberg-Nilsson, corr. 1993) \

l 3 . . . �ft) 14 d4 .i.f5 I S 0-0 �6 16 .e3 �f7 17 1I'b3+ �g6 I S c3 with a plus, because of the black king, Pape-Roelofszen, email 1995.

a3) s . . . Axf5?! is worse, though; 9 �d6! (following 9 �hS .xhS 10 IOds+ �dS I I .xhS �hS 1 2 �t7+ �7 l 3 �gS h6 14 �3 j.g6 I S �3 Black has some compensa­tion for the exchange, Grivainis­Kotek, corr. 1965) 9 . . .• d7 (9 . . .• xhS 10 .xhS �xhS I I fuf5+ �e6 12 �g3 �g3 l 3 bxg3 Ae7 14 �e4 with a clear advan­tage, Stummer-Spiegel, corr. 199 1 ) 10 �b7 (or 10 �xf5+ .xf5 I I Ae2 with advantage Gubats-Cuba, Latvia) 10 . . . �hS I I 1I'xhS g6 12 .e2 is convincing for White, Therkildsen-Zerbib, T ourcoing 1995.

b) S d3 is a reasonable alternative, S . . . g6 (stronger than S . . . �S 9 �hS 10ft) [9 .. . �S 10 0-0 �ft)?! { l0 . . . exd3} 1 1 dxe4 �6?! 12 eS! lOxeS 1 3 :el winning a piece, l 3 . . . Ae7 14 J:txeS 1I'xhS I S c4 with a winning advantage, Haba-Agnos, Bundesliga 1 9S9) 10 0-0 [ 10 dxe4) 10 . . . Axf5 I I dxe4 Ae6 12 AgS 1I'bS!? the queen is coming round for the hS-knight, l 3 Axft)+! [ 1 3 11'd2 is less helpful] l 3 . . . gxft) 14 �3 1I'gS I S IOds+ Wd7 1 6 f4 1I'g7 1 7 eS?! [ 1 7 f5!) 1 7 . . .• xhS I S �ft)+ �cS the position is not so clear, Saavedra-Atars, corr. 1 969) 9 �S (9 fxg6? bxg6 10 AXg6?! [ 10 �S gxhS I I AgS Ag7 12 dxe4 AxhS may be slightly better for White) 1O . . . :gS I I lOes J:txg6 1 2 �g6+ 1I'xg6 l3 �fl Af5 and the white king is more exposed than the black, Basmaison-Sireta, Auvergne IV 1993) 9."gxhS 10 0-0 Ag7 1 1 dxe4 AxhS 12 AgS with advantage, MalmstrOm-Sireta, corr. I 99419S.

8 .••• xb5 As ever, with the king so exposed,

it is better to plar the endgame. The move S .. .l��xhS?! didn't work

out too well in Cano-Ciprian, corr. 1970, 9 g4 (9 d3 transposes to S d3 in the last note) 9 . . . �f6 10 d3 exd3 I I O-O!? dxc2 12 .xc2 with advan­tage, the black king is very awk­wardly placed.

9 .xb5 �xb5 10 g4 �r6

1 1 :gl 1 1 lOc3?! is doubtful, I I . . .�xg4

(or I I . . .c6 12 :gl with advantage as the annoying . . . lOc:6-d4 is no longer possible) 1 2 �S+ �d7 1 3 �t7 1Oc6 and . . . �4 is threatened.

However, 1 1 h3 is a good, solid alternative, I I . . .lOc:6 12 lOc:3 ( 12 O-O?! �4 I 3 lOat3 dS?! [ 1 3 . . . gS! 14 fxg6 .i.g7 wins the knight] 14 d3 c6 I S dxe4 dxe4 16 :dl with advan­tage, Clarke-Diepstraten, corr. 1995) 1 2 . . . � 1 3 �dl �O 14 d3 exd3 I S cxd3 .i.d7 16 gS �hS 17 �S+ �dS I S �f7+ and White has got over the worst of his problems, Van de Velden-Melchor, corr. 1 995.

II ... � The knight threatens to hop into

o via d4, or eS, and . . . �b4 can also be effective. Black can also win the knight by a well-timed .. . gS, and . . . Ag7. Meanwhile, White's queen­side is undeveloped, and he must try to- find some way to either extricate

3 exf5 J8J

his knight. or gain sufficient com­pensation for it.

Il llg3 There are others: a) 12 �d l (to avoid forks on c2,

or 0) 12 . . . �S 13 h3 Zirnis-Silavs, corr. 1970, when I think I would re­capture the knight immediately, 1 3 . . . gS!? 14 fxg6 .i.g7 I S,AXh7 .i.xhS 16 lOc:3 .i.e6 and. providing Black retains his e4-pawn, the white kingside pawns shouldn't go far.

b) 1 2 gS? weakens fS and lets the cS-bishop out. 1 2 . . . �hS ( 12 . . . �!?) 13 g6 .i.xfS ( 13 . . . �?? 14 ph7 �c2+ I S �dl �f4 16 �g6+ 1 -0, Gaard-Knorr, corr. 1990; but Destrebecq's 1 3 . . . h6! is interesting. as although the hS-knight can go to f7, he cannot return from Black's territory) 14 gxh7 .i.xh7 I S �g6+ .i.xg6 16 Jlxg6 �S and Black has good play for the exchange, Riegsecker-Lonsdale, corr. 19S6.

c) 1 2 h3 is always a useful move, and I actually think it is the best here, 12 . . . lOd4 (immediately 12 . . . gS 1 3 fxg6 .i.g7 may be better) 1 3 �dl gS!? ( 1 3 . . . dS 14 b3 �O IS :g3 �eS 16 d3? .i.d6 17 :12 exd3 1 8 cxd3 .i.eS 19 �e2? [ 19 :g3] 19 . . . �h4 0-1 , Borsdortr-Grava, corr. 1965) 14 fxg6 .i.g7 I S lOc:3 ( IS gS?! �hS 16 gxh7 loosens the white kingside pawns, 16 . . . .i.fS 1 7 g6 �O and White is worse, he will

182 3 exft

shed all of his advanced pawns,) Ruhbaum-Schreyer, corr. 1979, when IS . . . .i.xhS 16 gxh7 .i.e6 1 7 d3 slightly favours White.

d) 12 d3?!, Schulien-Glass, corr. 19S0, allows the simple 1 2 . . . lOes 1 3 �e2 exd3+ 14 cxd3 lllexg4.

e) 12 lik3?!, likewise, 1 2 . . . lOes 13 gS lOhS ( l 3 ... 1Of3+ 14 �f1 lOg4 is also reasonable) 14 lOds+ �dS I S lOe3 100+ 16 �f1 lOxgl 1 7 Wxg I almost equal, Staak-Rubach, Pinne­berg 1994.

Il ... lOd4 1 2 . . . lOeS!? 13 h3 (the pawn needs

defending, 1 3 lik3?, Tiemann­Diepstraten, corr. 1990, is far weaker, 1 3 . . . lOexg4 14 lOxe4 lOxe4! I S lbg4 .i.xf5 16 11f4 g6 1 7 d3 10£6 and . . . .i.g7 with a clear ad-vantage) 13 . . . 100+ 14 �dl <ReS I S 1Oc3 dS 16 lObS .i.d6 1 7 lOxd6+ cxd6 I S d3 .i.d7 19 .i.f4 �e7 with chances for both sides, GUrd­Diepstraten, corr. 1 994/9S.

13 �dl p! Leading to some long forcing se­

quences, 1 3 . . . dS?! 14 d3 100 I S dxe4 dxe4 is advantageous for White after 16 lik3 8..6 ( l6 . . . lOxh2? 17 .i.gS 100 I S lOxe4 Garcia­Kapitaniak, corr. 1975) 1 7 lOxe4 Black is losing, Szilagyi-Downey, corr. 1976.

14 c3

This is the only test, others: a) 14 d3?! is a mistake, 14 . . . gxf5

IS SS (with advantage-Keres!) I S . . .tllg4 16 .i.e3 ( 16 c3? lOxf2+ 17 �el 1Oxd3+ I S �dl? f4 19 1lxd3 .i.g4+ 20 �e I? lik2+ 0- 1 , Elburg­Heap, corr. 1991) 16 . . . .i.g7!? (or 16 . . . f4 1 7 .i.xd4 fxg3 I S hxg3 lOes 19 dxe4 .i.g4+ 20 �c I .i.g7 2 1 f4 00+ 22 cxd3 .i.xd4 which is not easy to win, Hansel-Schreyer, East Germany 19S5) 1 7 Ilh3 �! IS 1lxh7 this is not too clear, I S . . . lOo 19 lOd2 �gS 20 g6 (20 Ilxg7+? Wxg7 2 1 lOif3 lOxe3+ 22 fxe3 exO is quite hopeless, Grabner-Tiemann, corr. 1975) 20 . . . .i.xhS 2 1 h4, McDonald-GUrd, corr. 1991 , when the rook on the seventh, and passed pawns, provide counterplay.

b) 14 fxg6?? is a blunder, 14 .i.xg4+ I S Ilxg4 lOxg4 � 16 gxh7 .i.g7 1 7 lOg6+ �f7 I S hS=. l.xhS 19 lOxhS+ IIxhS 20 d3 lOxf2+ 2 1 �d2 e3+ 22 �c3 lOdl+! 0- 1 , Prietz-Pupols, corr. 1967/68, as

23 Wxd4 1lh4+ 24 �dS e2 2S .i.d2 el=. 26 .i.xel lOe3 mate.

c) 14 gS?! is far from critical, 14 . . . ltlhS I S £6+ �eS 16 IIgI? ( 16 1le3 .i.g4+ 1 7 �e I is superior, although Black is at least equal) 16 . . . 100 1 7 1Ibl , Tortosa-Melchor, Barcelona 1993, 1 7 . . . lOxgS winning easily.

14 ••• pf5! IS gS IS cxd4 f4 16 IIc3 ( 16 IIgI?!

.i.xg4+ 1 7 �e I .i.h6 favours Black, Jackson-Pape, corr. 199G) 16 . . . .i.xg4+ 1 7 <Re I ( 1 7 �c2 lOdS I S IIb3 .i.g7 19 11bS .i.e6 20 Ilxb7 :i.xhS leaves Black with strong positional com�tion for the ex­change) 1 7 . . . IlcS!? ( l7 . . . lOdS! I S IIb3 .i.g7 19 IIbS .te6 20 Ilxb7 hhs is perhaps simpler, when Black has more than sufficient com­pensation fm the exchange, e.g. 2 1

lOt3 :g8) 1 8 d3 e3 ! 19 :b3? ( 19 fxe3 fxe3 20 .i.xe3 .i.g7 2 1 :b3 is much safer, but Black still does have some compensation) 19 . . . .i.h6 20 lbb7 .i.f3 2 1 lba7, analysis of Peel ( 1989), when 2 1 . . .:g8 catches White in a mating net, i.e. (peel's 2 1 . . .lOls is also good�� fxe3 fxe3 23 .i.xe3 .i.xe3 24 :g 1+ 25 �fl lDds.

15 ••• �g4 16 cxd4 �d2+ 17 Wel

17 ••• f4?! This doesn't look very logical to

me. 1 7 . . . lDd3! 1 8 lOt3 .i.e6 must be

fine for Black, his d3-knight is worth a rook. For instance, in the line 19 dS .i.d7 20 g6 .i.g7 2 1 �ti bxg6 22 ll)gS Black is a rook down, but better! 22 . . . J:h8 23 h3 .i.eS 24 :gl J:hS and White is playing with­out his queenside.

18 :c3 �3 19 :u7+ WeB With obscure complications ana­

lysed at some length by Pupols. 10 � 13+ 11 We3 fl 11 :1'7

.i.e6 23 :xfl �xfl 14 Wxfl, Matz­Leisebein, East Germany 1 990, and now 24 . . . .i.g7 25 dS .A.f5 is better for White, but Black has drawing chances.

A1 5 •• ..i.e7!?

Probably the better square for the bithop, since, unlike variation A4,

3 exf5 183

White's advance d2-d4 will not gain a tempo.

6 .i.bS+ 6 d3 d6 (6 . . . exd3 7 .i.xd3) 7

.i.hS+ WfS 8 �f'7 .e8 is similar, 9 �d6!? .xhS 10 .xhS �hS 1 1 �c8 .i.cs 1 2 dxc4 � 1 3 �a7, gaining four pawns for the piece, Wittrnann-Gunderam, corr. 1972.

6 ••• WfI

7 � This does not seem to ofTer White

much, but neither do the alternatives:

a) 7 d3 exd3 8 �d3 dS 9 .i.g4 lOt6 1 0 �f4 �g4 1 1 .xg4 .i.ffi 12. c3 .e7+ 13 �2, Wallwork­Saunders, corr. 199 1 , 1 3 . . .• eS equal.

b) 7 d4 d6?! risky, (7 . . . exd3 trans­poses to the above) 8 �ti (8 � .i.xf5 9 �3 .i.e6 10 dS 1.£1 1 1 h£1 Wx£1 1 2 0-0 �7 1 3 lOt3 Jensen-Magee, corr. 1991 , doesn't ofTer White anything. Generally, by comparison with the King's Gambit, the slight inconvenience suffered by Black's king displacement is offset by the tempi White has lost in play­ing .i.e2-hS-e2) 8 . . .• e8 9 g4 (the most precise, effectively forcing Black to capture on hS with his knight, 9 �8 .xhS 10 .xhS IOxhs 1 1 g4 �ffi 12 gS lOls 1 3 g6 [ 1 3 ffi gxffi 14 :gl � I S c3 f5 is

184 3 exf5

fine for Black] 1 3 . . . h6 [ 1 3 . . . .i.xfS 14 gxh7 .i.xh7 1 5 :gl �ffi is also quite reasonable] 14 h4 .i.xfS 1 5 hS ll)d7, whilst the h8-knight is now safe for a moment, it has nowhere to go after fl, Kerstens-Snuverink, Hengelo 1 997; 9 �xd6 .xhS 10 .xhS �hS 1 1 �c8 �h4 1 2 .i.e3 1Oa6 1 3 �a7 Zlxa7 White has enough compensation for the piece, Clarke-Stummer, corr. 1994/95) 9 . . . �hS 10 �xh8, Melchor­Magee, corr. 1989, 1O . • . �ffi 1 1 gS forcing Black to concede his knight, as 1 1 . . .�S 1 2 ffi gxffi 1 3 :gl .i.fS 14 c3 is awkward.

c) 7 �fl?! is premature, 7 . . .• e8 8 �xh8 (this time, 8 g4 is less effec­tive, 8 . . . �hS 9 �xh8 �f4 10 d3 exd3! 1 1 .i.xf4?! [ 1 1 cxd3 is more circumspect] 1 1 . . ..i.b4+ 1 2 �n .e4 13 :gl ? dxc2 14 .c l b6 0-1 , Pugh-Mamsell, corr. 1978, . . . .i.a6 is threatened) 8 . . .• xhS 9 .xhS �xhS 10 g4 �ffi 1 1 gS �S 1 2 g6 h6 ( 1 2 . . . �g8?! 1 3 lOfl d6 14 gxh7+ �xh7 Paavilainen-Niemand, corr. 1991 , 1 5 :gl .i.xfS 1 6 �gS+ with chances of an advantage) 1 3 f3 exf3 14 0-0 �6 1 5 c3 d6 16 d4 ( 16 :xf3 ll)es 1 7 :n �3 is hardly an improvement) 16 . . . �g8 1 7 �fl .i.xfS 1 8 lOxh6+ gxh6 19 Zlxf3 �xg6 20 �xh6, the black pieces are worth more than the rook and pawn, Evans-Clarke, corr. 1994/96.

7 ••• d6 8 �r7 .ea 9 �xb8 There is something to be said for

9 g4! �xhS 10 �xh8 �ffi, trans­posing to 9 . . . �xhS, but cutting down on Black's options.

Not 9 �xd6? .xhS 10 .xhS �xhS 1 1 �c8 .i.d8, trapping the knight, Jaunozols-Strautins, Latvia.

9 •••• xh5 9 . . . �xhS!? looks eminently play­

able: 10 g4 �ffi 1 1 gS .i.xfS!? ( 1 1 .. .�g8? cannot be right, 1 2 �S

1Oa6, Vitols-Clarke, corr. 1990, 1 3 .g4 c6 14 �f4, threatening an -ap' propriate �, must be good) 12 gxffi .i.xffi 1 3 �S .i.d8 14 ll)e3 ( 14 c3?! �7 15 ll)e3 ll)eS 16 �fS �f3+ 17 We2 Wbs+ 18 d3 .xfS 19 dxe4 .xe4+ 20 .i.e3 ll)es 21 :gl Wg8 and, after recapturing the knight on h8, Black had sufficient compensation for the exchange, Diepstraten-Morgado, corr. 1977) 14 ... �d7 1 5 b3 lOc6 16 d4 ( 16 .i.b2 �g8 1 7 :gl ll)eS) 16 . . . �d4 17 .xd4 .i.ffi unclear, Canal Oliveras­Kozlov, corr. 1998.

IO .xb5 �xb5 1 1 14 �f6 1l 15

-" � ,- " �-� 1 r� r��. r� 1 � � � �

'�' '� '� �- '� � ;1, � � % � � � . . ", ,� � � ft �Ri

� � 1 � � � � � � 'f;� ,'f.'\' � � ?}';f, � � � � �;e;� � v� /$, /� ..l.1. r� ..u.. �� �ll1 fJS, .� �. � � �

1l ••• lDe8! 1 2 . . . �g4 is more active, but after

1 3 00 ( 13 �xe4 is also a problem, 1 3 . . . .i.xfS 14 d3 Wg8 [ 14 . . . �S 1 5 f4] 1 5 f4 menacing h3) 1 3 . . . .i.d8 ( 1 3 . . . 1Oa6? 14 ffi! gxffi I S h3 �S 16 gxffi is worse, Koser-Schirmer, corr. 1993) 14 g6 .i.xfS 1 5 �fl 1Oa6 16 gxh7 .i.xh7 1 7 �d8 Zlxd8, Black's pawn, and light-squared control, provides some compensa­tion for the exchange, Keskinen­Kilpela, Finland 1997.

13 16 �g8 14 �5 .i.b4 15 �r7 .i.xf5 16 d3

-

Koser-SCnechaud, corr. 1993, whe.f 16 . . . .i.xg6 17 �gS .i.xgS 18 .i.xgS exd3 would have been rea­sonable for Black.

A3 5 ••• d5? This move combines the disad­

vantage of S . . . d6 (the awkward placement of the black king on e7) without the advantage-the immedi­ate attack on the eS-knight. There­fore White has a greater choice of seventh moves, and more than one method of cooking Black's goose.

6 �b5+ We7

7 d3! The best. White can capture on e4

when appropriate. to open the cen­tral files. There are two other impor­tant possibilities: .

a) 7 IOn is less effective. al­though the knight will be difficult to attack on hS, 7 . . .• eS with the fur-' ther choice:

a I ) S iOxhS "xhS 9 .xhS lOxhS 10 g4 ( 10 1Oc3 c6 transposes to S 1Oc3, but 10 . . . 1Of6 is perhaps better) 10 . . . 1Of6 I I h3 (the tactical I I :gl?! 1Oc6 12 :g3 iOd4 1 3 Wdl just helps Black. 1 3 . . . g6 14 c3 lOO Black has already gained the ascen­dance. Van Swol-Van Willigen, corr. 19S2, as I S gS [ I S fxg6 �g7) I S . . . .i.xf3 16 gxf6+ Wxf6 1 7 d3 �d6 regains material) 1 l . . .1Oc6 1 2 d3 WeS 1 3 iOd2? ( 1 3 �gS) 13 . . . lOb4 14 Wdl exd3 I S cxd3 lOxd3 16 We2 lOf4+ 1 7 We3 �d6 IS lOO hS!? and Black has excel­latt counterplay, Nobbe-Wynia. corr. 19S2.

3 exf5 185

a2) S 1Oc3? can transpose. but allows Black a couple of extra possibilities:

a2 1 ) S . . . g6 9 fxg6 (9 iOxhS gxhS) 9 . . . hxg6 10 lOxhS ( 10 �xg6? :gS) 1O . . . gxhS I I .e2. Downey-Tiemann. corr. 19S7, 1 l . . .�8? 12 d3 �xhS 1 3 dxe4 dxe4 14 ll)xe4 iOxe4 I S .xe4+ wn with every chance in the endgame.

a22) S . . . �xf3!? leaving the d­�wn to its fate, 9 iOxhS .xhS 10 WxhS ( 10 iOxdS+ Wd6 should transpose) 1O • . • iOxhS 1 1 iOxdS+ Wd6 1 2 iDe3 �e6, if Black finds time to play . . . g6. and . . . �g7. he will be fine. 1 3 g4!? lOf4 14 b3 ( 14 d3 exd3 I S lOc4+?! does not work. I S . . . �xc4 16 �xf4+ Wd7 1 7 cxd3 .i.dS I S 0-0 1Oc6 and the white knight will be captured by the black rook. after the ffi-bishop moves) 14 . . . lOd7 I S d3 ( I S �b2 iDeS) I S . . . Wc6 16 1013 �b4+ 1 7 Wfl �c3 I S :b I lOds 19 dxe4 lOb4 20 .i.b2 �xb2 2 1 J:lxb2 :xhS. unclear. Salvador-Goedhart, corr. 1979.

a23) S . . . c6?! is too acquiescent, 9 IOxhS (9 d3 lOxhS 10 IOxhS exd3?! I I 0-0 WdS 12 �gS+ 1Of6 13 �xf6+ gxf6 14 "xd3 �d6 IS lOxdS cxdS 16 "xdS Wc7 17 IOn and White has a raging attack, plus extra material, Modena-Alloin, 1990) 9 . . .• xhS 10 "xhS iOxhS 1 1 g4 1Of6 12 h3 ( 1 2 gS?! just weakens the pawn structure here. 1 2 . . . lOg4 1 3 g6 �xf3 14 1Of7 �xg6 I S lOgS and White has extracted his knight, but at too great a cost. Cimmino­Bonavoglia, Turin 1975) 12 . . . lObd7 13 d3 iDes and White held the upper hand, Gaard-Jackson. corr. 1992. as it is difficult to see how Black will capture the knight on hS.

b) 7 b3?! cS (7 . . . lObd7 S �a3+ cS 9 d4?! [9 1Of7 .eS 10 1Oc3 is tempting) � . .• c7 10 f4. Polaczek-

186 3 exJ5

Oppitz, Torremolinos 1986, 10 . . . exO I I j,xO lOxeS 1 2 .e2 j,xf5 is �layable.) 8 'On .eS 9 lOxhS (9 1&3 g6, unclear, Valenti­Redon, COlT. 1991) 9 . . .• xhS 10 • xhS lOxhs I I g4 10ffi 1 2 gS lObS? ( l2 . . . lOe8 1 3 tOe3 tOe7) 1 3 tOe3 d4 14 lOxe4 b6 I S IOg3 and Black is losing, De Boer-Fiorito, COlT. 1984.

c) 7 d4?! j,xf5 8 j,gS (S 'On .e8 9 j,gS g6 10 lOxh8 gxhS I I tOe3 c6 12 .d2 j,g7 1 3 .f4, Gaillard-Van Mulder, Paris 1996, when 1 3 . . . j,e6 14 0 exO leaves Black ahead) S . . . g6 9 tOe3!? (inter­esting; retreating.Jhe bishop would also offer some small advantage, i.e. 9 j,e2 j,g7 10 0-0 IObd7 I I c4 c6 12 tOe3 h6 13 j,xffi+ IOxffi?? [ 1 3 . . . Wxffi is perfectly reasonable] 14 g4 1 -0, Burk-Leisebein, COlT. 19S2; 9 84!? �e6 10 .i.xg6 bxg6 I I IOxg6+ wn 1 2 lOxhS+ Wg7 1 3 :gl i.d6, Black wishes to ca�ture on h8 with his rook, 14 lOd2 1&6 I S c3 Wd7 16 0 Kiihn-Leisebein, East Germany 19S2, and now 16 . . . j,xh2 1 7 :g2 J:[xhS is unclear) 9 . . . j,g7 (9 . . . gxhS?? 10 .xhS wins immedi­ately) 10 O!? gxhS (Black might as

well take the piece, although 10 . . .• d6 is not too bad, I 1 fxe4 j,xe4 12 lOxe4 dxe4 1 3 j,xg6 bxg6 14 lOxg6+ Wd7 I S IOxh8 J.xh8 16 0-0 with a slim advantage Svendsen-Tiemann, COlT. 1990; 1O ... e3?! I I g4 j,e6 1 2 j,xg6 bxg6 1 3 lOxg6+ wn 14 lOxh8+ .xh8 I S h4 tOe6 16 j,xe3, the three mobile white pawns are too strong, Svendsen-Magee, COlT. 1991) I I fxe4 j,xe4 1 2 0-0 for the piece White has opened the Hile, 12 . . . :f8! (this might be the most �rudent, 1 2 . . .• d6 13 .xhS .e6 14 t&e4 dxe4 IS J:[xffi! .i.xffi 16 j,xffi+ .xffi 17 :n .e6 IS dS .d6 19 'On � 20 Whl lOd7 2 1

.gS+ �ffi 22 d6+! 1 -0, Kozlov­Borrmann, COlT. 1992, because of 22 . . . Wd7 23 lOes+ Wxd6 24 J:[xffi+; 12 . . .• eS 1 3 .el :f8 14 lOxe4 dxe4 I S .xe4 IObd7 16 :ael lOb6 17 .xh7 1-0, Krantz-Melchor, COlT . 1989) 1 3 .el WeS 14 � .d6 I S :ael iObd7 16 lOxe4 dxe4 1 7 J:[xe4 lOxeS 1 8 dxeS?? ( 1 8 J:[xeS+ main­tained the attack) I S . . . lOxe4! 19 exd6 j,d4+ 0-1, Svendsen-Downey, COlT. 1993.

7 •• ..i.xf5 a) 7 . . . :gS is a real computer

move, although it may be best, ob­jectively, S 0-0 (S dxe4 dxe4 9 .xd8+ WxdS 10 J..n J:[bS I 1 j,b3 is simple, and effective) · S . . . j,xf5 9 dxe4 he4? (9 . . . dxe4! leads to a playable position) 10 tOe3 .d6 I I .i.f4 .cS 1 2 j,n :h8 1 3 :e I and Black has no answer to the attack along the e-file. Rise 2S00-chess Genius 3, Liebert 1993.

b) 7 . . . exd3?!, Black should not open the e-file with his king in such a state, S cxd3 (8 O-O!? i.xfS 9 :el ! dxc2 1 0 IOn+ Wd7 1 1 Wo cxbl=. 1 2 J:[xb I and Black is lost, Harper­Pape, COlT. 1990) S . . . j,xf5 9 tOn .i.g4?! (9 . . .• eS 10 .e2+ j,e6 I I j,� g6 1 2 IOxh8 gxhS 1 3 f4 j,g7 14 15 Wd7 is slightly better) 10 0 .i.xhS 1 1 lOxd8 Wxd8 12 0-0 and Black's compensation for the queen

is woefully inadequate, Martinez­Thomsen, Copenhagen 1982.

l 1.g5 By far the most common move

here, despite the fact that 8 �f7! wins outright: 8 . . .• e8 9 .i.gS g6 10 iOxh8 gxhS ( lO . . . exd3 1 1 0-0 dxc2 12 1WxdS is also hopeless) 1 1 �3 c6 12 dxe4 dxe4 13 .d4 .i.g7 14 1Wb4+ �e6 I S .xb7 and Black's position is in ruins, MalmstrOm­Vitols, corr. 1994195.

1 ... i.e6! Other, inferior moves: a) 8 . . .• d(; 9 �f1 1rh4+

(9 . . .• e6!?) 10 .i.d2 .xb2 1 I .i.c3 � 1 2 iOxh8 d4 1 3 dxe4 +/­Stummer-Harper, corr. 199 1 .

b) 8 . . . g6? 9 dxe4 .i.xe4 10 f3 ! (even simpler than 10 .i.f3 .d6 1 1 lOg4 �bd7 [ 1 1 . . ..i.g7 1 2 �3 .i.xf3 1 3 .xf3 .e6+ { 1 3 . . . c6, then 14 0-0-0 �bd7 transposes to I I . . .�bd7) 14 �f1 .a6+ I S �gl lObd7 16 iOxdS+, crushing, Stummer-Kron�f, corr. 1990] 1 2 �3 .i.xf3 1 3 .xf3 c6 14 0-0-0 .i.g7 I S :hel+ �fT 16 ltie4! dxe4 1 7 :Xe4 .c7 1 8 :xd7+ 1 -0, Kiihn­Leisebein, East Germany 1982) 1O . . .• d6 1 1 lLlg4 lLlbd7 12 fxe4 gxhS 1 3 iOxf6 �xf6 14 0-0 .i.g7 I S eS 1 -0, Radovic-Pape, email 1998.

9 dxe4 dxe4 10 lLlc3 .d6 1 1 lLlg4 b6 Il lLld6 .xdl+ 1 3 :xdl gd6 14 .i.r4 IS 15 lLld5+ .i.xd5 16 :xd5

For once material is balanced, but White has a clear positional advan­tage, Stummer-Chmilewski, corr. 199 1 .

•

A4 5 ••. .i.c5?!

Tempting, but, it now seems that this is dubious because of the ex­posed position of the bishop.

3 exf5 187

6 .i.b5+ 6 d4 should amount to the same,

6 .. . exd3 7 iOxd3 .i.b6 8 O-O? (the last chance for 8 .i.hS+ transposing) 8 . . . 0-0 9 �3 dS 10 .i.f3 c6 1 1 1.gS .i.xf5 12 .i.h4 lLlbd7 Roach-French, corr. 1967/68, Black is already at least equal, and following 13 b3? he could have gained a clear advantage by 13 . . . .i.d4.

6 • • • �f8 7 d4! Simpler than 7 lLlfT .e8 8 iOxh8

(8 _.84?! iOxhS 9 iOxh8 �f4) 8 . . .• xhS (8 . . . lLlxhS!? is also possi­ble, 9 .e2 dS 10 d4 .i.b6 1 1 tOe3?! [ I 1 g4 lLlf6 1 2 gS is more to the point] I I . . .�g8! 12 iOxdS �xh8, unclear, Stummer-Schirmer, corr. 199 1 ) 9 .xhS iOxhS 10 d3? (the standard 10 g4 �tO I 1 gS lLlg4 12 �3 asks more questioos of Black) 1O . . . exd3 I 1 cxd3 lLltO with advan­tage to Black, Vitols-Krauklis, Lat­via 1970, as at some point Black will play .. . �g8xh8 with two pieces for a rook.

7 ••• exd3 l lOxd3 .i.b6 8 . . .• e7+ changes little, 9 1.e2

and: 9 . . . .i.b6 10 0-0 d6 I 1 g4!? hS? ( l l . . .�6) 12 �f4! +/- Koser­Evaos, corr. 1993; or 9 . . . dS 10 iOxcs .xcS I I .i.d3 ( 1 1 i.e3, Eidan-<>ren, corr. 1990, 1 1 . . . Irb4+ 12 �3 .i.x(5) 1 l . . .�f1 1 2 0-0 with advantage, AuerbachlTauber­AntoniadilEliascheff. Paris 19 17.

J88 3 exf5

•

9 0-0 lllc6 10 .i.g5 d6 1 1 :e1 Menacing thl black queen. 1 1 ••• .i.d7 1 l . . . �7 12 lOes Ae6?

( 12 . . . lOxhS 1 3 tOe6!) 1 3 lllg4 lOxg4 14 .i.xg4 1 -0. Stimpson-Hall. corr. 197 1 .

Il lllr4 lOe5 I3 lOd5 With a large plus. Koser-Hansson.

corr. I992.

4 •••• e7 The move 4 . . . dS is more in keep­

ing with gambit play. but White can gain an appreciable advantage: S d3 (S 1t:xi4?! is less logical. and was severely punished in Bertin­Destrebecq. corr. 1978. S . . . lllft) 6 f3 cS 7 Wbs+? [7 � .i.xe6 8 fxe6 tOe6! 9 fxe4? {9 c3 } 9 . . . lOd4 10 1I'dl lllxe4 and the threat of . . . Wh4+ is decisive) 7 . . . .i.d7! 8 1I'xb7 cxd4 9 .xa8 1I'c7 10 lOa3? [ 10 a4 is the only chance. so that 10 . . . .i.c6? can be met by I I .i.bS. but 10 . . . .i.d6 is still very good for Black) 10 . . . .i.xa3 I I bxa3 .l.c6 ;: and 1 1 . . .0-0 ma� be even stronger. objectively) S . . . lllft) (S . . . .i.xfS!? 6 dxe4 [6 lOd4 is also tempting. 6 . . . 1I'd7 and. instead of 7 f3?! .i.g6 8 fxe4 dxe4 9 lOb3 .e7 10 dxe4 Wh4+ I I g3 .xe4 1 2 .i.g2 1I'xe2+ 1 3 Wxe2 with equality. Miiller-Wynia. corr. 1982. 7 lOxfS

.xfS 8 lOd2 lllft) 9 f3 wins a pawn. although 9 . . . tOe6' gives Black a lead in development as compensation) 6 . . . dxe4 7 tOe3 [7 WbS+?! .i.d7 8 .xb7 {8 1i'hS+ 86 9 1I'eS+ .e7 10 .xh8 exf3+ I I .l.e3 lllft) is not that clear. the white queen is trapped} 8 . . . .i.c6? Greenwalt-Wall. Dayton 1983, {8 . . . exf3! 9 .xa8 .i.c6 is much stronger} 9 .i.bS .i.xbS 10 1I'xbS+ c6 I I .e2 with advantage) 7 . . . lllft) 8 .i.gS .i.e7 [8 . . . Ab4!? 9 Wbs+ tOe6 10 .xfS exf3 I I 1I'xf3 1I'e7+ is more interesting) 9 lOd2 with a small advantage Radikevich­Khavin. Kiev 1937) 6 dxe4 (this is more accurate than the more com­mon 6 tOe3?! .i.b4 7 dxe4 [7 .i.d2?! 0-0 8 dxe4 .i.xc3 9 .i.xc3 lOxe4 and Black enjoys a very pleasant posi­tion, e.g. 10 0-0-0 { IO .i.eS?! tOe6 I I 0-0-0 l%e8 leaves Black well on top. Wild-Buchicchio, Saint Vincent Italian Ch (corr.) 1999} 10 . . . lllxc3 I I bxc3 .i.xfS 12 Wbs. Svendsen­Jackson. corr. 1987. 1 2 . . . .i.e4 White's weakened king position must surely favour Black) 7 . . . lOxe4 8 .i.gS?! [8 lllgS is possible. 8 . . . .i.xc3+ 9 bxc3 0-0 10 lOxe4 dxe4 1 1 .xe4! tOe6 {if 1 l . . .l%e8 12 Ac4+ Wh8 1 3 .i.e6} 1 2 .i.c4+ Wh8 1 3 .i.e6 White might just be able to claim an edge) 8 . . . .t.xc3+ 9 bxc3 .d7! 10 lOd2 [ 10 ft) 0-0 1 1 fxg7 1I'xg7 1 2 .i.d2 .i.g4 hands Black a dangerous initiative) 1O . . . 0-0!? I I lOxe4 .xfS! 1 2 lllft)+ gxf6 1 3 .i.h6 :d8 14 .e7 .eS+ I S .xeS fxeS. Landgraf-Stummer. corr. 1990. Black's superior pawn structure out­weighs White's bishop pair) 6 . . . dxe4 7 tOe3 .i.b4 8 .i.d2 .l.xc3 (8 . . . 0-0!? 9 1I'c4+ Wh8 10 .xb4 �6 I I 1I'cs exf3 1 2 0-0-0 is also better for White; 8 . . . .i.xfS? 9 WbS+) 9 .i.xc3 .i.xfS (9 . . .• dS!? 10 :dl 1I'xfS 1 1 It:xi4 [ 1 1 .i.xft) gxft) 12

:ci4 is possibly even better] I I . . . .. CS 12 ""5+ "xbS 1 3 lOxbS with a clear advantage in the end­game. Destrebecq-Kozlov. corr. 1992) 10 lOd4 Ad1 1 1 0-0-0 with advantage. Destrebecq-MalmstrOm. corr. 1991 . Also. 4 . . . lOffi 5 d3 dS transposes to the above.

5 lt)d4 The odd move 5 lDgl ?! was first

played by Borrmann. S . . . dS 6 lDc3 (6 d3 .i.xf5 [6 . . . exd3 7 "xe7+ lOxe7 S Axd3 Axf5 is a simple al­ternative] 7 dxe4 "xe4 s lDc3 Ab4 9 .t.d2 .xe2+ 10 Axe2 lDffi. quite level. lackson-<>ren. corr. 1994/91) 6 . . . c6 1 �S+ "t7 S .xt7+ �t7 9 f3 (9 d3 offers a little plus) 9 . . . exf3 10 lOxf3 Axf5. Black has no prob­lems. Borrmann-Destrebecq. corr. 1990.

5 •.• lDc6! Aiming for a speedy develop­

ment. Other moves are:

a) s . . . lDffi!? 6 d3 ! (the most test­ing: 6 lDc3 is inaccurate. 6 . . . dS! [6 . . . cS?! 7 lDdbS dS S d3! trans­poses to S . . . cS; no one has taken up my idea of 6 . . . lDc6!? here. which is similar to s . . . lDc6 except that White can no longer play .5+: 7 lOxc6 [7 lDdbS? dS and .. . a6 ;. transposes into the Logunov-Kozlov game in the main line. DOte cto move six] 7 . . . dxc6 S g4 hS ! 9 gS lOdS 10

3 exf5 189

lOxe4 Axf5 I I d3 0-0-0 with very good strategical compensation for the pawn. indeed. As any King's Gambit player will know. White's backward f-pawn is quite useless] 7 d3 .eS! transposes below) 6 . . . dS (or 6 . . . exd3?! 1 "xe7+ Axe7 S Axd3 0-0 9 0-0 with advantage. Bray<andeias. Portugal 1997; 6 ... cS? 1 lDbS dS S Af4 [S lDlc3 is also good, transposing into S ... cS?!. but this is completely clear] S . . . lDa6 9 lDd6+ WdS 10 dxe4 IOxe4 1 1 lOxe4 "xe4 12 .xe4 dxe4 1 3 lDc3 Axf5 14 0-0-0+ WeS I S Axa6 bxa6 16 lOxe4 Axe4 17 IIhe I leading to a swift end. Clarke-Evans. corr. 1994/95; 6 . . . lDc6!? 7 lOxc6 dxc6 S AgS [S dxe4 is critical] S . . . Axf5 9 lOt3 exd3 10 .xe7+ Axe7 I I Axd3 Axd3 1 2 cxd3 0-0-0 is fairly level. Goncalves-Krustkains. corr. 19S5) 7 dxe4! (7 lDc3 "eS! s lDe6 Axe6 9 fxe6 lDc6 [9 . . . Ab4! 10 .t.d2 lDc6 is right. with approximate equality] 1 0 dxe4 lOd4? [ IO .. . Ab4 is best, a.sain] 1 1 "dl?! [ 1 1 f4! "xe6 12 �bS lDxe2 1 3 lOxc7+ Wt7 14 lOxe6 lOxc I I S lDgs+] 1 1 . ..dxe4 [it will be no surprise to the astute reader that 1 1 . . .�b4 is again best!] 12 lDe2?! 0-0-0 and Black has the better chances. Garcia Martinez­Busom. Barcelona 1995) 1 . . .• xe4 (7 ... lOxe4!?) S lDc3 Ab4 9 Ad2 .xe2+ 10 Axe2 0-0 I I 0-0 cS!? 12 lDe6 Axe6 13 fxe6 lieS. Bitter­Legouhy. Val Maubuze 1 990. and now 14 Af3 lDc6 ( 14 . . . Axc3? I S Axc3 /l)e4 16 Ilad I ) I S life I lOd4 16 1lad1 maintains a plus.

b) S .. .• eS?! . Black can ill afford the luxury of another queen move. but it is surprisingly popular. 6 lDbS lDc6? (objectively. this is losing. as is 6 . . . a6?? 7 d4 .dS S lOXc1+ Elburg-Kott, corr. 1991 ; 6 . . . lDa6! is the only try. c7 must be defended 1

1 90 3 exf5

O! [7 d4?! 'lxfS S 0 d5 9 lLxI2.

lDffi lO &4?! _g6 1 1 _SS .�SS 1 2 lDxe4 Wh4+ 1 3 Wf2 Wxf2+ Destrebecq-Downey, training game 1993 :j:] 7 . . . dS S fxe4 dxe4 9 g4 hS [9 . . . .i.e7?! fails to cause the least irritation: 10 .i.g2 .i.h4+ I I Wd I lDffi?! 1 2 d4! "e7 1 3 gS winning more material] lO lOlc3 bxg4 I I d4 .xfS 12 lOxe4 .i.e6? [ 1 2 . . . WdS has to be played] 1 3 lOg3 winning, Malmstrom-Reinke, COrT. 1994/95, as 1 3 . . .• dS 14 .i.g2 .c4 1 5 .i.xb7). Now we have:

b l ) 7 lOlc3 lOf6 S f4! .xfS (or the invariably pl�yed S . . ... xf4 trans­posing to 1 f4? Wxf4 S lOlc3 1Offi?, above) 9 lDxc7+ (9 \,4 lOxg4 10 .i.h3 is also convincing) 9 . . . WdS lO lDxaS dS with some vague practical chances for Black, although objec­tivel� speaking he is quite lost! l O . . . lDd4 I I "c4.

b2) 7 f4? is worse, 7 . . ... xf4 S lDlc3 lDffi? (White is also winning after S . . .• xfS? 9 �xc7+ WdS lO lDxaS lDd4, Black is relying on this counter-shot, but it amounts to little, 1 1 .c4! .i.cs [ l l . . .lDxc2+?? 1 2 Wdl lDxa l 1 3 .c7+ 1 -0, Svendsen­Schirmer, COrT. 1992] 1 2 b4 lDxc2+ 1 3 Wdl lDxb4 14 lDxe4 and Black is crushed, Destrebecq-Borrmann, COrT. 1990; but S . . . WdS! avoids los­ing a rook, 9 lDdS .xfS lO lObxc7 :bS and Black is doing alright!) 9 d3 (alternatively, 9 d4!? .xfS lO lDxc7+ WdS I I lOxaS d5? [ 1 l . . .lDxd4 1 2 "d2 .i.cs 1 3 lDdl :eS 14 �3! lOO+ 1 5 pO .i.xe3 16 .d6 would at least have given White some moves to make, al­though Black's attack seems to have fizzled-out, anyway] 1 2 .i.e3, Black is quite lost, Destrebecq-Jackson, COrT. 1 990) 9 . . ... eS (9 . . .• xfS? 10 lDxc1+ WdS I l lDxas lDd4 12 dxe4 .eS [ 1 2 . . . lDxe4? 1 3 .xe4! lDxc2+

14 Wdl "xe4 15 lDxe4 lDxal 16 .i.f4 1 -0 Landgraf-Spiegel, COrT. 1990, the white knight will escape, but not the black one] 1 3 "d3 .i.cs 14 g3 d5 1 5 .i.f4 Black can resign with a clear conscience, GHrd­Malee, COrT. 19S9) lO lDxe4 lDxe4 1 1 Wxe4 and Black is 'only' a pawn down with a rotten position, not bad for this line! Elburg-Miiller, COrT. 1991 .

b3) However the unplayed 7 d4! just wins on the spot, and renders the rest of this note quite irTelevant! i.e. 7 . . . lDxd4 (7 . . . exd3 S lDxc7+ WdS 9 "xeS lDxeS lO lDxaS dxc2 I l lDe3) S f4 lDxe2 9 fxeS lOxcl lO lDxc7+ WdS I I lDxaS and, as both knights are trapped, White will find himself an exchange to the good, 1 1 . . .b6 1 2 lDe3 .i.b1 1 3 :xcl .i.xaS 14 1LI1 etc.

c) S . . . cS?! 6 lObs (6 �S+?! WdS 7 lObS transposes; 6 .c4?! is too artificial, 6 . . . 1.Oa6 7 lOb3 lOffi S lDe3 lDe7 [S . . . lOb4! 9 Wdl dS] 9 .i.e2?! [9 "a4] 9 . . . dS 10 "a4+ .i.d7 and White has failed to deal with Black's imposing centre, and is worse, Niveau-Rogalski) 6 . . . dS 7 lOlc3 ! (best, 7 �S+?! is a waste of time, 7 . . . WdS S 1Olc3 lOffi 9 "gS a6 lO lOa3 bS I 1 lDab I lDe6 and Black's space advantage is almost decisive, 12 g4? �S 1 3 .i.e2, BOhm-Diepstraten, Hilversum 1993, 13 ... d4, winning a piece; 7 d3 White plays to break-up the black centre, 7 . . . a6 S lOSc3 [if S .i.gS, Swaffield­Bullockus, COrT. 1 972, then S ...• xgS! 9 lDe7+ Wf7 lO lDxaS "cl+ 1 1 .dl .xb2 1 2 lLxI2 exd3 13 .i.xd3 "eS+ with plenty of play] S . . . exd3 9 cxd3 lOffi 10 .i.gS /l}bd7 1 1 lDd2 bS [ 1 1 . . .d4? 1 2 lDce4 with advantage, Bajovit-Caroff, Metz Open 19S5] 1 2 0-0-0, Destrebecq­Nyman, COrT. 199 1 , 1 2 . . . Wf7 1 3 :el

.xe2 14 .i.xe2 .i.d6 with fair play for the pawn) 7 .. . lOa; 8 d3 a6 9 dxe4! axbS (or 9 .. . dxe4? 10 .i.f4! axbS I I lDxbS lDdS? 1 2 .i.d6 .17 13 .xe4+ �d8 14 0-0-0 1-0, Destrebecq-Strautins, corr. 1991) 10 eS d4?! (Destrebecq analyses 10 .. . b4 I I exa; "xe2+ 12 lDxe2 .i.xfS 13 lOf4! 1Dc6 14 lDxdS 0-0-0 I S lDe3 with advantage; 1O . . . lObd7 might be the best try, I I .i.f4 :as 12 exf6 lDxa; with a white advan­tage) 1 1 exa;?! ( 1 1 lOxbS! is very strong, l l . . .lDdS 12 lDd6+ �d7 [ 12 . . . �d8? 1 3 .i.gS] 1 3 lOI7! Ilg8 14 .i.gS winning) 1 1 . . .• xe2+ 12 lDxe2 .i.xfS 1 3 lOg3 .i.xc2 14 .i.xbS+ 1Dc6 I S fxg7 .i.xg7, Destrebecq-Krustkalns, corr. 1 99 1 . and Black has almost sufficient compensation.

6 fl}sc6 6 lObS?! dS 7 lOIc3 (on 7 d4

.i.xfS 8 .i.f4 Ilc8 9 lOIc3 lOa; 10 0-0-0 a6, the knight is obliged to retreat, and Black already has the upper hand, De Jong-<>ren, corr. 1994/9S; 7 �S+ �d8 achieves lit­tle) 7 ... lOa; 8 d3 a6 9 lDa3 It)d4 10 .e3 lOxfS White is in trouble, Logunov-Kozlov, corr. 1993; 6 �S+ �d8 7 lOxc6+ dxc6 transposes.

6 ••• dsc6

' .... 5+ This uses valuable time, but:

J uf5 191

a) 7 d3?! seems too acquiescent: 7 . . . .i.xfS 8 dxe4 .i.xe4!? (8 . . .• xe4 9 .xe4+ .*.xe4 equalises immedi­ately, 10 c3 .i.cs 1 1 lDd2 lOa; 12 .i.e2 0-0-0 1 3 lDxe4 lOxe4 14 0-0, Oren-Reinke, corr. 1994197, when 14 .. . �b8 I S .*.0 llhe8 16 .*.f4 as favours Black) 9 c3 (9 1Dc3?! .i.xc2 10 .*.e3 .i.g6 I I Ild I lOa; leaves White without sufficient compensa­tion for the pawn, Jackson-GUrd, corr. 1994/97) 9 . . . 0-0-0 10 .*.e3 lOa; 1 1 lDci2 (if I I .i.xa7 then Black obtains plenty of play by 1 1 . . . lle8 1 2 .*.e3 .*.xb 1 1 3 Ilxb 1 lDds) 1 1 . . . .*.g6 12 h3 �8 1 3 .0 lDe4 14 lDxe4 .*.xe4 IS .e2 .17 with a nice position, Destrebecq-Gaard, corr. I992.

b) 7 1Dc3 .i.xfS 8 g4! .i.g6!? (8 . . . .*.e6 9 .*.g2 lOa; 10 gS lDds I I .xe4 0-0-0 may be superior) 9 h4?! (9 .i.g2 lOa; [or 9 . . . 0-0-0 10 .i.xe4 .17 1 1 d3 .t.b4] 10 gS lDds 1 1 .*.xe4 .*.xe4 1 2 �e4 is critical) 9 . . . h6 (9 . . . hS!?) 10 .*.g2 lOa; I I gS lDds and Black is fine, Jensen­Clarke, corr. 1992, as, this time 1 2 .i.xe4 .i.xe4 13 lDxe4 is followed by 13 . . . hxgS 14 lDxgS lDh4.

' ••• �d8 8 lDc3 8 .*.e2 lOa; 9 .gS h6 10 .e3

.*.xfS I I 0-0 lDds 1 2 "d4 .d6 13 d3 lOb4! (an ending would suit Black fine) 14 .c3 exd3 I S cxd3 the black king looks less exposed now I S . . . .*.e7 16 a3 ( 16 .xg7? .*.a; and . . . 1Dc2 wins material) 16 . . . lDds with good play, Psomiddis-Hector, Katerini 1992.

8 • • • lOf6 9 .as .e5! 10 .*.c4 10 .*.e2 h6 1 I .g3 .xfS?!

( l l .. .• xg3 12 hxg3 hfS is simple, and good) 12 d3 .*.d6 1 3 dxe4 ( 1 3 .xg7! 1lh7 14 dxe4) 1 3 . . . lDxe4 14 -.M+ lOa; IS .*.d3 lle8+ 16 lDe2 .dS 1 7 0-0 .*.fS, play is balanced, MalmstrOm-Heap, corr. I 994/9S.

1 92 3 exf5

10 ••• b6 1 1 .13 .xf5?! Latvian Gaulbit players like to

keep queens on the board, but, as in the previous note, 1 1 . . ... xg3 1 2 hxg3 .i.xf5 is objectively strongest, when Black may even be better.

11 d3 .i.d6 13 .xa7! Ilb7 14 dxe4

Malmstrom-Grobe, corr. 1 994/95, 14 ... llxg7 I S exf5 1lxg2, White has an edge.

C 4 lOd4

This momentarily defends the f5-pawn, but on d4 the knight is a target for the black pieces.

4 •••• f6 This strange queen move is the

most popular, and leads to interest­ing positions, but the mundane 4 . . . /.l)f6 may be stronger:

a) 5 d3 cS (5 . . .• e7 6 "e2 trans­poses into 4 .e2) 6 /.l)b5 (the 6 �3 of Schiro-Krumins, USA 1973, is best countered by 6 ... exd3 7 .i.xd3 d5 8 .i.b5+ Wf1 9 .i.e2 [9 0-0 .i.xf5 is possible, as 9 . . . c4? 10 .i.xc4] 9 . . . .i.xf5 10 0-0 1Oc6 1 1 .i.f4 .i.d6 the game is level, Jaclcson­Stummer, corr. 1994/95; 6 1t)e2!? may be the best square, 6 . . . dS?! [6 . . . exd3 7 .xd3 d5 is simpler] 7 dxe4 /.l)xe4, Baas-Diepstraten, Hilversum 1993, 8 /.l)f4 /.l)ffi 9 .i.b5+ 1Oc6 to .e2+ is a little

awkward) 6 . . . a6 7 /.l)5c3 d5? (7 • . exd3 8 .i.xd3 d5 gives Black a strong, mobile centre, e.g. 9 g4?! c4! and . . . d4) 8 dxe4 dxe4 9 .xd8+, Downey-Heap, corr. 1992.

b) 5 .i.c4? is the sort of 'obvious' move that White plalers often choose, after 5 . . . c5 6 �3 dS 7 .i.b5+ (following 7 .i.fl?! .i.xf5 Black already had a clear advantage in Kugler-Farwing, Hamburg 1959) 7 . . . lDc6 (7 . . . Wf7! is possibly even stronger, as . . . c4 is threatened, cutting off the bishop's retreat) 8 .i.xc6+ bxc6 9 g4?! as 1 0 d4 exd3 I I cxd3 a4 1 2 /.l)3d2 .e7+, Daikeler-Sneiders, corr. 1987, and now White had to play 1 3 Wfl to stay in the game.

c) 5 .i.e2, the check on h5 will at least cause Black's king to lose his castting rights, 5 . . . .i.c5!? 6 .i.h5+ (or 6 lllb3 .i.b6 7 O-O?! [7 .i.h5+ Wf8 causes Black more inconven­ience] 7 . . . 0-0 8 d4 d5 9 .i.e3 .i.xf5 Black's position is more pleasant, Rabson-Lee, London 1983) 6 . . . WfB (6 . . . g6? 7 fxg6 0-0, Simon-Eberth, Felsonyarad 1985, is far too ambitious, 8 /.l)f5! and if 8 . . . dS 9 g7 wins the exchange) 7 /.l)b3 .i.b6, unclear.

d) 5 f3?! cS?! (5 . . . .i.c5 6 /.l)b3 .e7 appears promising, as 7 /.l)xc5 is answered by the Zwichenzug 7 . . . exf3+) 6 /.l)b3 dS 7 g4?, Sawyer­Hubbard, corr. 1968, can be coun­tered by, amongst others, 7 . . . c4 (or 7 . . . 1Oc6 threatening . . . c4) 8 /.l)d4 .i.c5 when White is in trouble, e.g. 9 1t)e6?1 .i.xe6 10 fxe6 /.l)xg4 .

e) 5 1Oc3 cS 6 �3 dS 7 d3 exd3 8 .xd3 (8 cxd3 .i.xf5 9 .i.e2 is already better for Black, Rantanen­Molander, Finland 1993) 8 . . .• e7+ 9 .e2 .i.xf5 10 .xe7+ .i.xe7 and Black has 110 problems, Schroder­Hufschild. Strelasund 1997.

S c3 The strongest reply, maintaining

the knight in the centre. a) The queen check is certainly

too risky here: S �S+?! g6 6 fxg6 hxg6 7 .dS 1De7 8 .xe4 (Betin§ @ves an amusing line: 8 Wc4 dS 9 Wc3 s'g7 10 lObS .f7!? l l lOxc7+ Wd8 12 .cS 1Dd7! 1 3 .d6 s'eS 14 We6! .g7 I S lOxdS [ I S lOxa8? lOfS wins the queen!] I S . . . lOxdS 16 _xdS s'xb2 17 s'xb2 .xb2 18 .as+? [ 1 8 .gS+ We8 19 .xg6+ draws] 1 8 . . . b6 19 .c3 .c l+ 20 We2 b6+ winning) 8 . . . llh4 9 g4 dS lO .d3 Ilxg4 1 1 c3 lObc6 1 2 .A.e2 IDes 1 3 We3 1le4 ( 1 3 . . . c S 1 4 s'xg4 .A.xg4 -+) 1 4 .g3 .A.g4 I S A .A.xe2? ( l S . . .• a6! wins on the spot) 16 lDxe2, Evans-Grivainis, Munich 19S8, and eventually the American Grandmaster managed to play him­self out of the hole he'd gotten into.

b) However, first S lObS lDa6 and then 6 Whs+ is possible: 6 . . . g6 7 fxg6 hxg6 8 .e2 dS 9 d3 .A.f5 lO dxe4 dxe4 I I g4 .A.e6 1 2 _xe4 0-0-0 1 3 gS, probably best, ( 1 3 lOxa7+? Wb8 14 s'xa6 .A.dS I S .a4 .A.xh I 16 .A.e2 .A.cs 1 7 s'e3 .A.xe3 1 8 fxe3 Ilxh2 0- 1 , Hansen-Pape, Denmark 1986; 1 3 s'gS? .xgS 14 _xe6+ Wb8 i s also wrong) 1 3 . . . s'dS! 14 gxf6 ( 14 .K4+ "CS I S _xf5+ gxf5 16 Ilg1 lle8+ 17 s'e2 Ilxe2+ 1 8 Wxe2 .i.c4+ is also

3 exfJ J 93

unclear) 14 . . . .A.xe4 I S Ilgl lDxf6 16 1OIc3?! 1Ob4 and White is in trou­ble, Romanenko-Dreibergs, USA 19S2.

S ••• CS!? a) S . . • IDe7?! 6 d3 (6 _g4!? dS 7

d3 1Dxf5?! [7 . . . .A.xf5 8 1Dxf5 .xf5 is not too bad for Black] 8 dxe4 1Dd6, Dimmel�Brunold, Germany 1988, when, 9 Whs+ g6 10 .xdS c6 1 1 .gS leaves White well in charge) 6 . . . 1Dxf5 7 dxe4 1Dxd4 8 cxd4 .A.b4+ 9 1Dc3, Jackson­Downey, corr. 1 987, and White has a strong centre and an extra pawn!

b) S . . . 1Dc6?! is quite common: 6 1Dxc6 (6 lObS .eS 7 _e2 Wd8 8 lOSa3 .xf5 led to a mutually diffi­cult position in Downey-Krantz, corr. 1992) 6 . . . dxc6 7 Whs+ g6 (this gives more practical chances than 7 . . .• f7 8 .xf7+ Wxf7 9 g4 hS 10 s'c4+ Wf6 I I d3 ! exd3 1 2 f3 with advantage Downey-Melchor, corr. 1992) 8 fxg6 hxg6 9 _e2 .A.e6 lO .xe4 0-0-0 I I d4 s'd6? (following the better 1 l . . .lle8 1 2 .A.e3 .A.d6 Black has little co�sation for the pawn) 1 2 .A.gS! _xgS 1 3 _xe6+ Wb8 with advantage Downey­Elburg, corr. 1990.

6 lL)bS .eS 6 . . .• xf5!? is difficult to believe,

nevertheless: 7 1Dc7+ Wd8 8 lDxa8 lOf6 9 S,bS? (9 d4) 9 . . . a6 lO .A.e2 bS 1 1 1Ob6 .A.b7 1 2 a4 Wc7 1 3 as c4 14 b4 cxb3 I S _xb3 .A.cs 16 0-0 1Dc6 1 7 _a2?, White should do something about Black's growing initiative, 1 7 . . . IDeS 1 8 .A.a3 lOf3+!, ripping open the white king's de­fences, Downey-Diepstraten, corr. 1993.

7 .A.e2 7 .e2! lDa6 8 g4 1Of6 9 d3 is

very good for White, Downey­Borrmann, corr. 1988.

7 •••• 6

1 94 3 exf5

7 . . . lDf6 8 d4 exd3 9 .xd3 a6?? (9 . . . dS) lO f4 1 -0, Jackson-Borrmann, COR. 1 988, Black will lose a rook; 7 ... lDa6 8 d4 exd3 9 .xd3 dS is a reasonable line.

8 d4 exd3 8 . . .• xfS?? 9 1Dc7+ 1 -0, Downey­

Grobe, COR. 1989. 9 f4 .u2+ 10 .u2+ du2 1 1

Cik7+ �d8 1 2 lDxa8 bS 1 3 lDb6 .iob7 14 84 �c7 IS 8S .ioJ.g2 16 :gl .ion

. . . seems very playable for Black.

o 4 lDgl!?

A bizarre idea of the German player Bucker, played with the in­tention of hanging on to his extra f-pawn. Black is forced to play a King's Gambit accepted with col­ours reversed, and with the e-pawn slight!y further advanced.

4 ••• l.i)f6 I suppose that any move played in

the King's Gambit is also quite playable here, too. Some examples:

a) 4 . . .• fti?! and: a I ) S 1I'hS+! g6 6 fxg6 hxg6 7

.e2 (7 .dS? runs into trouble: 7 . . . :hS! [7 . . . :h7!? is also interest­ing, 8 .xg8?! {8 1Dc3!?} 8 . . . :t7, trapping the queen, 9 lDh3 dS 10 .iobS+ c6, Diepstraten, I I .i.e2 .i.e6! Schiller] 8 .xg8 [8 Wb3 might be the safest, 8 . . . dS 9 d3 .iocS, Schiller;

8 .c4 dS 9 .xc7 1Dc6 10 .i.bS .i.cs it is clear that the white pieces have strayed too far from home, 12 is un­defendable, I I lDe2 { I I .g3 :gS 1 2 � J:[xg2 is also good for Black; I I f3 ),b6 1 2 .g3 :gS 1 3 � J:[xg2 likewise} 1 l . ...x12+ 1 2 �dl lDge7 1 3 .g3 .fti 14 .el .i.d7 I S :n .eS 16 h3 0-0-0 and the white pieces make for a comical picture, Pohl-Wegelin, corr. 1990] 8 . . . dS-forcing White to jettison material in order to extricate his queetr-9 g4 .i.xg4 [9 .. . .i.e6? 10 Wxf8+ .xf8 I I gxhS] 10 .i.h3 .i.xh3 I I lDxh3 liJc6! 12 lDf4 else .. . lDe7 1 2 . . .• xf4 13 .xg6+ .t7 14 .xf7+ �t7 I S d3 exd3 16 cxd3, Kron.sraf-Magee, corr. 1990, 16 . . . l.i)b4 -/+) 7 . . . l:hS (7 . . . l:h7?! turned into a fiasc:o: 8 1Dc3 .eS? 9 lDf3 .as lO d3 .i.b4 I I .i.d2 dS 1 2 dxe4 +/- Magee-Nolden, corr. 1991 ; but 7 . . . dS 8 d3 also favours White) 8 1Dc3 dS 9 d3 .iob4 10 .i.d2 :eS I I dxe4 dxe4 12 0-0-0 .fS? 1 3 .c4 1 -0, Magee-Melchor, corr. 1990.

a2) S d3 is also strong, S . . .• xfS (or S . . . dS?! 6 Whs+ .t7 7 .xf7+ �t7 8 dxe4 dxe4 9 .i.c4+ �e8 lO lDe2 .i.xfS I I lDbc3 c6?! 1 2 lDg3 +/- Stummer-Nolden, COR. 1991) 6 dxe4 (6 .e2 wins a pawn, see (b) 6 . . .• xe4+ 7 .i.e2, Melchor-Jensen, . corr. 1993 transposing into 4 .. .• gS.

a3) S g4 hS 6 1Dc3 (6 d3 exd3 7 .i.xd3 hxg4 8 .xg4 lDc6 9 1Dc3?! [9 .g6+] 9 ... lDeS 10 .e4 1&7 I I .i.f4 1Oxd3+ 12 .xd3, Hayward­Gaard, corr. 1990, 1 2 .. . d6 equal) 6 .. .• eS (6 .. . hxg4 7 .xg4 l:h4 8 .g6+ .xg6 9 fxg6 .i.b4 lO .i.c4 lDh6 I I lLlds .i.d6 and the open h­file gives Black some compensation for the pawn, Hayward-Krantz, COR . 1992) 7 .e2 hxg4 8 lDxe4, Hayward-Tiemann, corr. 1990, and after 8 . . . lDc6 9 .i.g2 lild4 lO f4 .e7

Black will recapture the fS-pawn, when his knight will be well-placed.

b) 4 . . .• gS S d3 (Black is well �aced to meet S g4: S . . . hS! 6 d3 .xg4 [6 . . .• e1?! 7 .i.e3 hxg4 8 .xg4 lOh6 9 "xe4?! {9 "86+ IOn lO 1Oc3} 9 . . . dS lO "xe7+ .i.xe7 1 1 1Oc3 c6 and Black has some play for the pawn, Hayward-Grobe, corr. 1989] 7 .xg4 hxg4 8 dxe4 dS! 9 .i.f4 [9 exdS .i.xfS lO 1Oc3 is critical] 9 . . . dxe4 lO .i.xc7 .i.xfS, unclear, Magee-Senkhaud, corr. 1990) S . . .• xfS 6 dxe4 (goil!8 for quick develo�ment, but 6 .e2!? [Schiller] 6 . . . lD£6 7 lDd2 dS 8 0 seems to win a pawn! 8 . . . e3 !? might be Black's best try, e.g. 9 "xe3+ .i.e7 10 g3!? 1Oc6 1 1 .i.h3 1I'hs 1 2 .i.xc8 Lc8 with free play in return for the pawn) 6 . . .• xe4+ (6 . . ... n? offers little in return for the pawn: 7 -*.e2 �cS 8 100 d6 9 0-0 10£6 10 eS with advantage, Knorr-Schmidt, corr. 1992) 7 -*.e2 IOf6 (White has a tidy lead in development, so Black should hasten to castle, 7 . . . dS 8 100 1Oc6? is too slow: 9 0-0 10£6 l O 1Oc3 .fS 1 1 :e I �e7 12 .i.bS .i.d7 13 .i.gS �n 14 .i.d3 .g4 I S :xe7+! �e7 16 IOxdS+ 1 -0, Magee-Pape, corr. 1991 ; note that 7 . . .• xg2?? loses to 8 .i.o "g6 9 �hS) 8 100 -*.cS (Otherwise, 8 . . . �b4+!? is a clever idea, attempt­ing to deprive White of the c3 square, 9 c3 [9 .i.d2 0-0 10 0-0 .i.xd2 1 1 .xd2 c6 1 2 1Oc3 .e8 1 3 :fel dS 14 .i.d3 .d8 � Melchor­Jensen, corr. 1993, soon drawn] 9 . . . .i.cS 10 b4 .i.b6 1 1 0-0 0-0 12 a4 c6 1 3 1Obd2 .g6 14 IOc4 .i.c7 I S 1Od6 1Oe4, Schirmer-Muller, corr. 1986, 16 1Oxe4 "xe4 1 7 .i.d3 with an edge) 9 0-0 0-0 l O 1Oc3 "fS 1 1 -*.d3 (the crude 1 1 .i.c4+ �h8 12 lOgS!?, Bussmeyer-Stamer, corr. 1988, is best countered by 1 2 . . . b6!

3 exf5 195

[ I 2 . . . lOg4? 1 3 IOn+ Ln 14 .i.xn lOxf2 IS .0] 1 3 1Ot7+ [ 1 3 lObS?! dS ! 14 IOxc7 dxc4 I S lOxa8 1086 traps the a8-knight; 1 3 .i.d3!?] 1 3 ... Ln 14 .i.xn .i.a6 which re­gains the exchange) 1 1 . . .1fhs 1 2 .i.c4+ �h8 (but White's slight de­velopment advantage does not count for much) 1 3 .i.e3 ( 1 3 :el?! , Magem Badals-Knox,London 1987, 1 3 ... lOg4 forces 14 -*.e3 lOxe3 I S fxe3 with Black for 'preference as 14 IOe4 allows 14 . . . lDxf2 IS 1Oxf2 .i.xf2+ 16 �f2 "cS+ 1 7 .i.e3 .xc4) 1 3 . . . d6 ( I3 . . . c6 was seen in Leeners-Elburg, corr. 1984) 14 lObS!? IOg4 IS :el? (better IS .i.xcs "xcS 16 .d4) I S . . . lOxe3 16 fxe3 .i.b6 and e3 is a serious weakness, Borrmann-Leisebein, corr. 1986.

c) or 4 . . . .i.e7 S 1fhS+ �m, unclear,

d) but if 4 . . . .i.cS?! S 1fhS+ �m 6 £6 looks awkward.

5 14 Again, by analogy with the King's

Gambit, this must be the move to play. Alternatively: S d3 .e7! (this idea of Kozlov's seems to work bet­ter than S . . . exd3?! 6 .i.xd3 dS 7 .e2+ .i.e7 8 100 0-0 9 0-0 cS lO c4! �, Hergert-Frenzel, corr. 1986; S . . . dS 6 dxe4 [Stefan Bucker prefers 6 g4!? exd3 {6 .. .• e7 7 d4 trans­poses to S . . .• e7} 7 .i.xd3 .e7+

1 96 3 exf5

{7 . . . .*.cS !? 8 g�? We7+ 9 �f1 1bc4 10 �S+ �d8 is good for Black} 8 �fl 1Oc6 {Black hastens to castle queenside, this is simpler than 8 . . . hS !? 9 gS 1bc4} 9 13 .*.d7 10 lDc3 d4 I I Ibc4 lOes 1 2 lOxffi+ "xft) 1 3 .*.c4 0-0-0, Black has good compensation, as the white kingside is loose, Augustin-Schmidt, corr. 1989] 6 . . ... e7 [6 . . . dxc4?! 7 .xd8+ �d8 8 1Oc3 .i.xfS 9 lOge2 lObd7 10 109) { IO lOd4! gains the bishop pair} 1O . • • .*.g6 1 1 .*.gS .*.b4 12 0-0-0 .*.xc3 1 3 bxc3 �e7 14 .*.c4, Nikolic-Raty, Bad Niendorf 1989, when 14 . . . h6 1 5 .*.f4 1Ob6 16 .*.b3 :'c8 is fine for Black] 7 1Oc3 .*.xfS 8 .*.gS dxc4 9 .*.c4 lObd7 10 "e2 0-0-0 I I 0-0-0, Gaard-Clarke, corr. 199 1 , and now any sensible queen move, avoiding lOds, equalises-1 1 . . ... eS for instance) 6 dxe4 (Bucker's preference, 6 d4! is best, 6 . . . dS 7 g4 hS!? [but 7 . . . e3 !? is also interesting, e.&. 8 .*.xe3 .c4 9 .13 "xg4] 8 gS lllg8 very aesthetic­the perfect answer to White's fourth move! 9 .*.h3 [9 1Oc3 may be stronger, 9 . . . c6 10 .*.h3 W£1 I 1 f3!? .*.xfS 12 fxc4 dxc4 which is un­clear] 9 . . ... £7 1 0 c4!?, Jensen­Kozlov, corr. 1989, and now I think Black is better after 1O • • • dxc4 I I g6 "ft) 12 Wa4+ Wc6; 6 .*.e2 dS 7 g4 exd3 8 .xd3 lOxg4 9 .xd5 lOft) 10 "d3 1Obd7 [ 10 . . . &6 is more active still] 1 1 .*.gS lOcs 12 "13 .*.d7 1 3 .*.xft) Wxft) 14 "e3+ .*.e7 1 5 1Oc3 .*.00 16 .*.hS+ g6 1 7 fxg6 hxg6 1 8 .*.13 0-0-0 19 .*.xc6 Wxc6 and Black has good play, Malmstrom­Svendsen, corr. 1990) 6 . . .• xc4+ 7 .*.e2 �cS 8 lOf3 0-0 9 1Oc3 "xfS 10 0-0 dS 1 1 .*.d3 ( 1 1 lOd4?! .*.xd4 12 .xd4 1Oc6 ;, Svendsen-Kozlov, corr. 1987; 1 1 .*.gS c6 1 2 .d2 [ 1 2 .*.d3 1bc4 13 We l :C8] 12 . . . 1bc4 13

lDxc4 dxc4 14 .*.c4+ �h8 1 5 lDh4 WeS and the offside white knight and active black pieces offer Black excellent play, Pape-Kozlov, corr. 1990) 1 l . . .1i'hS 12 .*.g5?, Tiemann­Svendsen, corr. 1990, 12 . . . lOg4! and the threat of capturing on 13 leaves White in a terrible mess.

5 •• ..*.«:5 By far and away the most popular

move here, but possibly not the best. The others:

a) S . . . hS ! 6 gS lOg4 7 d4 (7 d3 e3 [7 . . . exd3 transposes] 8 .*.xe3 lDxe3 9 fxe3 WxgS 10 Wf3 dS I 1 .*.h3 is similar, Lomo-Sersch, Gausdal 2000; if 7 h3 Black can play a 'reversed whip� : 7 . . . lDxf2!? 8 �f2 "xgS 9 d4 WxfS+ 10 �e 1 .*.e7 which looks like it could be fun) 7 . . . e3 !? (Not 7 . . . d5? 8 h3 lDxf2 9 · �f2 .*.xfS 10 .*.e3 +/- Hayward­Tiemann, corr. 1990; but 7 . . . exd3 8 .*.xd3 .*.cS 9 We2+ � 10 g6+ �f8 I I lOh3 [ 1 1 lOf3!? �f2 1 2 .*.gS We8] 1 l .. .1Ih4 12 Wf3 lDes 13 Wg2 [ 1 3 .c4!?] 1 3 . . . lDxd3+ [ l 3 . . . lObc6! leads to an unclear p0-sition] 14 cxd3 d6 1 5 lOgS, Pape­Clarke, corr. 1990, and again I s . . . 1Oc6 is reasonable) 8 he3 (8 lOb3!? is plausible, and avoids los­ing the gS-pawn) 8 .. . lDxe3 9 We2 (9

fxe3 .xgS 10 .f3 is also possible, but the text exchanges the queens) 9 .. .• xgS 10 .xe3+ .xe3+?! ( 10 . . . .i.e7 has got to be better, why improve the white structure?) I I fxe3 dS 1 2 .i.d3 .i.b4+ 13 c3 .i.d6 14 1Of3 0-0 I S iOh4 and Black's compensation for the pawn is inade­quate, Canal Oliveras-Svendsen, corr. I 997.

b) S . . . h6?! seems rather passive, 6 .i.g2 (or 6 d3 .e1?! [6 . . . exd3 7 .i.xd3 .e7+ 8 � b6!? is more combative] 7 dxe4 [7 d4 is not bad, either] 7 . . .• xe4+ 8 .e2 dS 9 1Oc3 .i.b4 10 .i.d2 .i.xc3 1 1 .i.xc3 1Oc6 [ 1 1 . . .0-0 is better] 1 2 f3 .xe2+ 1 3 lOxe2 with advantage, Hayward­Zschom, corr. 1991 ) 6 . . . 1Oc6 7 1Oc3 dS 8 f3! exf3 9 .xf3 with advan­tage, Stummer-Landgraf, corr. 199 1 .

6 g5! Just like a reversed Muzio Gam­

bit! Other moves: a) 6 d4? exd3 7 .i.xd3 .e7+! 8

� (unfortunately for White, if he puts a piece in the way he loses his g-pawn) 8 . . . hS 9 @ l.&4 10 .i.xe4 (10 W?! dS 1 1 "f3 .eS 12 1Oc3 .i.xf5 13 IOf4? .i.g4 0- 1 , Wittmann­Nyffeler, corr. 1990) 1O . . .• xe4 I I .e2 .xe2+ 1 2 1Oxe2 dS 1 3 f6 gxf6 14 g6 (after 14 gxf6?!, instead of 14 . . . .i.h3+, Borrmann-Svendsen,

3 exf5 1 97

corr. 1988, there is 14 . . . 1Od7 I S .i.gS Wfl 1 6 IOf4 c6 and. after Black recovers the f6-pawn, he will have a clear edge, as he has fewer pawn islands) 14 . . . .i.f5 I S IOf4 1Od7 16 c3 0-0-0 with an advantage in development, and chances of win­ning the g6-pawn, Leeners-Sinke, corr. 1982.

b) 6 1Oc3 0-0 7 gS lOc8 (7 ... dS!? 8 gxf6 .xf6 might be worth a shot) 8 .i.c4+?! (8 lOxe4 .i.b6 9 .f3 is more to the point) 8 . . . Wh8 9 d4 exd3 (9 . . . .i.b4 is superior) 10 .i.xd3 dS I I f6, Stummer-Svendsen, corr . 1990, 1 1 . . .1Oxf6! 12 gxf6 .xf6 13 .e2 .i.xf2+ 14 Wdl c6 with two pawns and an attack for the piece.

6 ••• 0-0 Black has no choice but to offer

his knight as 6 . . . IOdS? 7 1Whs+ Wf8 8 .i.c4 c6 9 1Oc3 is pretty miserable. However, 6 . . . 1Oc6?! might be a bet­ter way of doing it, controlling d4, if White had to capture, 7 gxf6 (unfor­tunately, White also has the prag­matic response 7 1Oc3!) 7 . . .• xf6 8 d3 dS 9 dxe4 "'4 1O .d2 dxe4?! ( lO . . .• xe4+!? 1 1 .e2 .i.xf5) I 1 .i.bS?? ( 1 1 .f4! administers a cold shower to Black's thoughts of at­tack!) 1 1 . . .e3 12 .e2 .xf2+ 1 3 Wdl �xis Melchor-Magee, corr. 199 1 , White is lost.

7 d4!

/ 98 3 exf5

This fine tactical point guarantees White the advan18ge; the obvious 7 gxf6 gives Black just what he's looking for: 7 . . .• xf6 8 .i.h3 dS 9 llX:3? (it is time for 9 d4! .xd4 [9 .. . .i.xd4 10 c3 .i.xf2+ I I �f2 .i.xfS is tempting, but 1 2 �g3! .g6+ 1 3 �h4 defends, amazingly enough] 1 0 .xd4 .i.xd4 I I 1&2 with advantage) 9 . . . .i.xfS 10 d4?! exd3 I I .i.xfS _xfS 12 .xd3 .xf2+ 1 3 �dl .fl+ 14 .xfl lbfl+ I S �e2 lbgl -/+ Hayward­Elburg, corr. 1 990.

7 ••• exd3 The one drawback of having the

e-pawn on e4, instead of eS, is that Black must capture on d3, not d4, 7 . . . .i.e7? 8 gxf6 .i.xf6 9 llX:3 dS 10 .i.h3 llX:6 I I .i.e3 +/- Magee­Spiegel, corr. 1 99 1 .

8 gxf6 .xf6 White has no need to fear

8 . . ... e8+?! 9 .i.e2! dxe2 (9 . . .• e4 10 "xd3! "xh l I I .g3 Ilxf6 12 .i.f3 traps the black queen, 1 2 . . . .i.d6 1 3 .i.xhl .i.xg3 14 hxg3 when White's two pieces are s,!perior to the black rook) 10 "dS+ WfT I I .xcS b6?! 12 "e7 "xf6 1 3 .xf6 +/­Hayward-Magee, corr. 1 99 1 ; 8 . . . lle8+? 9 .l.e2! "xf6 (9 . . . dxe2? 10 _dS+ �h8 I I fxg7+ �g7 1 2 "xcS is even worse) 10 .xd3 d6, Taylor-Hayward, corr. 1 987, I I 1Wb3+ �h8 1 2 .i.e3.

9 .i.xd3 d5 9 . . . lle8+? 10 �2 "'4 I I 0-0 b6

12 lOg3 .i.b7 1 3 Whs puts an end to

Black's hopes, Magee-Svendsen, corr. 1991 .

1 0 _13 c6 l l lDe2 1 1 llX:3!? aiming to castle 10n8.: 1 l ••• lDd7 12 lbg3?! lDe5 13 We2

.i.d7 14 0-0 :ad

15 .dl I S .i.e3?! �! 16 ""S .xhS 1 7

lDxhS .i.xe3 (Black is not even worse after 1 7 . . . lDxd3 ! 1 8 cxd3 d4 19 .i.d2 IlxfS 20 lOg3 1lf3 as the white pieces are misplaced, and the d3-paWjr is lost) 1 8 fxe3 lDxd3 19 cxd3 :Xe3 with some drawing chances, Magee-Harper, corr. 1 99 1 .

15···16? If I S . . . Wh4 then 16 �g2, but

I S . . . lDxd3!? 16 .xd3 hfS 17 lDxCS "xfS 1 8 _xfS lbfS is �ayable, e.g. 19 llX:3 d4 20 lOb I lle2.

16 .i.b6 1lf7 17 .i.e3 lOxd3 18 .xd3 .i.xe3 19 fxe3 _xb2 20 lDc3 with advantage to White, Magee­Kozlov, corr. friendly game 19911 .

1 1

A very natural response from White, but it would appear at first sight that Black is well placed to counter this. Indeed he is, but there is more than a drop of poison in this line for the unwary second player, as in the main variations White of­fers a dangerous piece sacrifice.

1 e4 e5 1 �f3 f5 3 d4

3 ••• fxe4 There seems little reason to search

for alternatives at this stage, as the text is obvioUs and good. For the sake of completeness, though, here they are:

a) 3 . . . d6 again transposes into the Philidor, followed by e.g. 4 .i.c4 (or 4 CbteS fxe4 etc.) 4 . . . exd4! .

b) 3 . . . exd4?! 4 lOxd4! (White could also consider playing a sort of 'reversed Falkbeer Gambit' with 4 eS!?, but not 4 .i.c4 which worked out badly in Tiemann-Downey, corr. 1990: 4 . . . fxe4 S �S .i.b4+! 6 c3 dxc3 7 bxc3 .f6! 8 .dS 1.Oe7) 4 . . . � S lOxf5 .xe4+ 6 1.Oe3 �f6 7 .i.d3 .c6?! (7 ... �4!?) 8 0-0 dS 9 �3 �7 10 I.OexdS �xdS 1 1

3 d4

�S+ g6 1 2 .xdS .xdS 1 3 lOxdS with advantage, Melchor­Diepstraten, corr. 1991 .

4 �xe5 Again, there is no advantage to be

gained from spurning this move: a) 4 .i.gS?! has been played sys­

tematically by the correspondence player Diepstraten, but the exchange of a pair of minor pieces only aids Black: 4 . . . .i.e7 (4 . . . /t)f6 S .i.xf6?! [S lOxeS transposes to the mainline] S . . . gxf6? [S . . .• xf6 6 lOxeS d6 is good for Black, and is examined in the note to White's main move six] 6 lOxeS! fxeS 7 �S+ gave White a devastating attack in Diepstraten­Falk) S �xe7 .xe7 (S . . . it)xe7 is also possible, 6 lOxeS d6 7 �4 0-0 8 1.Oe3 �f5 and if anyone is better, it is Black, Diepstraten-Elburg. corr. 1990) 6 lOxeS /t)f6 7 �3 d6 8 � (8 �S? is overly aggressive, 8 . . . lOxdS 9 �S+ g6 10 �xg6 .f7 1 1 .i.e2, occurred in a 1985_.same of Diepstraten, and now 1 1 . . .• xg6 12 .xdS .f5 won easily) 8 . . . dS (fmt 8 .. . 0-0 9 .i.e2 dS might be even

200 3 d4

more accurate) 9 lOe3 (9 lOe5 lLlbd7 10 lLlg4 c6 I I £e2 lLlxg4 1 2 £xg4 0-0 1 3 0-0 lLlffi ;, Diepstraten­Budovskis, corr. 1977/9) 9 . . . £e6 10 lLla4 lLlbd7 = Diepstraten-Grivainis, corr. 1977179.

b) 4 lLlfd2!? the 'Zemitis Varia­tion' 4 . . .• e7! (Black's choices are

somewhat limited as 4 . . . exd4?, and 4 . . . d5? are both met by 5 �5+, al­though 4 . . . d6 5 lLIxe4 lLlffi 6 £g5 £e7 is also fme) 5 d5 (5 �5+ g6 6 .xe5 .xe5 7 dxe5 .i.g7 [7 . . . e3 !? is interesting, 8 fxe3 .i.g7 9 £<:4 lLle6 10 lLle3 lDxe5 I I .i.b3 lLlffi 12 0-0 d6 with the superior structure, Zemitis-GrivainislHayward, corr. 1997] 8 lLlxe4 £xe5 9 lLlb<:3 lLle6 10 .i.e3 lLlge7 1 1 0-0-0 a6 and in this type of endgame Black's extra <:entre pawn should not be underes­timated, Zemitis-Strautins, corr. 1999; if 5 lLIxe4 then 5 . . . d5 ! [5 . . . exd4 6 £d3 {but not 6 .xd4? d5! -+} 6 . . . d5? 7 �5+] 6 lLlg5 lLle6! with a great position, Alberts­Bravo, corr. 1979/80, if 7 dxe5 .xe5+ 8 £e2 lLlffi 9 0-0 £<:5 ;) 5 . . . lLlffi 6 lLle3 .17 7 lLlexe4!? (7 lLldxe4 £b4) 7 . . . £e7 8 £d3?! lLIxd5 with advantage, Zemitis­Grivainis, corr. 1970/72.

c) 4 lLlg5?! exd4 (4 ... d5 5 dxe5 reached a Philidor position with a tempo more for Black in Lartigue­Alberts, corr. 197 1 , whilst in the game Spigel-Sawyer, corr. 1988, White tried the enterprising 5 lLIxe4!? dxe4 6 �5+ �d7 7 dxe5!? and went on to win. but 7 . . . g6 8 .g4+ �e8 9 .xe4 £g7 certainly favours Black. Otherwise, 7 .5+ �e8 8 �5+ �d7 9 .5+ �c6!? [9 . . . �e8 draws] 10 .xe4+ �b6 I I £e3 exd4 12 .i.xd4+ <:5 13 .i.e3 lLle6 and White has some play for the piece, Alberts-Trobatto, corr. 1979) 5 .xd4 (5 lLlxe4 lLle6 +)

5 . . . lLlffi (5 . . . d5 6 .e5+ lOe7 7 £e2 is undear, Redolfi-Atars, corr. 1973) 6 lL1xe4 .e7 7 lLlb<:3 lLle6 +.

4 ••• lLlf6 As in many other lines of the Lat­

vian, Black <:an also develop his queen to ffi: 4 . . .• ffi? but, for once, it is not especially good, S £<:4 lOe7 (5 . . . c6? 6 £t7+! �d8 [6 ... �e7 7 �5 with the threat of £g5] 7 .i.xg8 lbg8? 8 .i.g5 1 -0, �nCc::haud-Evans, corr. 1992, as

8 . . .• xg5 9 lLlt7+; and this tactic works well against other moves, as

well: 5 . . . lLle6? 6 £f7+ [6 lLlf7 .xd4 7 .xd4 lLIxd4 is also good, although not 8 �d2?? b5! undear, Goedhart-Ten Hove, corr. 1982. but 8 lLIxh8 lLlxc2+ 9 �dl lLIxal 10 £xg8 winning] 6 . . . �d8 (6 . . . �e7 7 lLle3!) 7 £xg8 lLIxe5 (again, 7 . . . lbg8?? 8 £g5) 8 dxe5 .86 9 .d5) 6 lLle3 d6 7 lLlf7 £e6 8 .i.xe6 .xe6 9 lLIxh8 with advantage Saleen-Sticker, Baden-Baden 1987.

5 £g5 Black is ready to expel the white

king's knight from e5, how best to react? The text move prepares a sharp piece sacrifice. Other moves:

a) 5 £<:4 is a sensible move, pr0-voking . . . d5, and thus maintaining the knight on e5. White has a dear plan, he will continue with <:4 put­ting pressure on d5, but Black sets

up his centre with gain of tempo. S . .. dS 6 .t.b3 (6 . . . .t.d6 7 .i.gS [7 0-0 lDbd7 8 lOxd7?! .i.xd7 9 f3 c6 10 .i.gS .c7 ; Prokopcuk-Ozimok. USSR 1969] 7 . . . c6 8 0-0 0-0 9 1Dc3?! [9 c4!] 9 . . .• c7 10 .t.f4. Mason-Pollock. London 1 887. 10 ... lDb<i7 ;) 6 ... .i.e6 7 j.gS .i.e7 8 0-0 0-0 9 lLxI2 (9 f3-Kercs) 9 . . . 1Dc6!? (perhaps 9 . . . lDbd7!? try­ing to exchange a couple o(pieces. or 9 . . . cS!? to dxcS .c7) 10 lOxc6 = Stockholm-Riga. corr. 1934/36.

b) 5 .i.e2 is very popular. but not very good. Black chases the eS-knight with gain of tempo. and completes his development easily. S . . . d6 (Black can avoid the piece sac with S . . . .i.e7!? as 6 0-0 [6 .i.hS+ g6 7 lOxg6?! {or 7 .i.e2 d6 8 lDg4. unclear. Grozshans-Spiegel. corr. 1992} 7 . . . hxg6 8 .t.xg6+ Wf8 leads nowhere; 6 .i.gS transposes to the note to Black's fifth move. S . . . d6] 6 . . . d6 7 lDg4 0-0 8 c4 lDxg4 9 .i.xg4 1Dc6 10 .i.e3 .i.f6 is level. Reefschlaeger--Pirrot. Germany 1996) 6 lDg4 (6 .i.hS+?! 5ai:rifices a piece. 6 . . . g6 7 lOxg6 bxg6 8 .i.xg6+ �e7 [8 . . . Wd7 might be simpler. 9 .i.gS dS to .i.fS+ We7 I I .g4 .i.xf5 12 .xfS .t.g7. consolidating. Marchio-Lutz, Hofbeim 1995] 9 .t.gS .t.h6 to .t.xf6+ Wxf6 I I .t.xe4 Wg7 1 2 .d3 1Dc6 ; Alherts-Ortiz, corr. 1967; 6 1Dc4 is an alternative. 6 . . . dS 7 lDes .i.d6 8 .i.f4 .i.e6 9 0-0 0-0 10 .i.g3 cS I I 1Dc3 1Dc6 and Black is already on top. Tomay­Delgado. Catalonia 1 994) 6 . . . .i.e7 7 1Dc3 dS and here Hector has faced two moves: 8 0-0 (8 lDes 0-0 9 .i.gS c6 to 0-0 .i.fS I I g4? weakening. 1 l . . ..i.e6 12 j.e3 lilfd7 13 f4 exf3 14 .i.xf3 lDxeS I S dxeS lDd7 + 16 .t.e2 lIxfl + 1 7 Wxfl lDxeS 18 .i.xa?? b6 19 .d4 lDd7 0- 1 . Moullimard-Hector. Clermont-

3 d4 201

Ferrand 1989) 8 ... 0-0 9 f4!? lOxg4 to .i.xg4 1Dc6 I I .t.e3 .t.f6 12 .i.xc8 lIxc8 13 g4!? lDaS! 14 b3 cS I S gS .i.xd4 16 .i.xd4 cxd4 17 .xd4 ]IfS 18 :adl 1Dc6 19 W12 lDb4 20 lDxe4 lIxc2 2 1 .d4 WaS 22 ]112 ]lxfl 23 lDxfl � 24 Wd3 lIxf4. Black seems to be get­ting on top. 25 :a I ! d4 26 Wc4+ ]In 27 lOe4 d3 28 g6! bxg6 29 .c8+ Wh7 30 �3+ Wg8 3 1 Wc8+ Wh7 In-In. Emms-Hector. Matalas­canas 1989.

c) 5 1Dc3 d6 (again. S . . . .t.e7 is possible. 6 .i.gS with transposition to a later note) 6 lDc4 (6 .i.gS relocates to the main line) 6 . . . dS (6 . . . .i.fS 7 .i.e2 .i.e7 8 0-0 0-0 =

Luna-Atars. corr. 1973, or 6 .. . .i.e7. . . . 0-0 =) 7 lDes (7 lDe3 c6 8 .i.e2 .t.d6 9 0-0 0-0 is pleasant for Black, Evans-Kozlov. corr. 1994/96) 7 . . . .t.d6 8 .i.e2 0-0 9 f4 exf3 10 .i.xf3 c6?! (Lein suggests to . . . lDc6! instead. I I lOxc6 [ 1 1 .t.f4 lIe8] l l . . .bxc6 12 O-O .i.fS) I I 0-0 Wc7?! ( l l . . ..i.xeS!? 1 2 dxeS lDe8? fails to 13 lOxdS! , but 12 . . . lDfd7 might be possible. aiming for the blockading square e6 by 13 Wd4 lDa6-c7) 12 .i.f4 .i.e6 i, Clarke-Zschom, corr . 1990, and not 1 2 . . . lDfd7? 1 3 lOxdS! cxdS 14 .t.xdS+ Wh8 I 5 �S win­ning quickly, Kaminskas-Erler. corr. 1995.

d) 5 lDg4 dS 6 lOxf6+ .xf6 7 Whs+ .n 8 .xt7+ Wxn =,

Marting-Pollock. New York 1 889. S ••• d6 Should Black wish to avoid the

coming complications. then he might try S . . . .t.e7!?: 6 1Dc3 (6 .t.xf6?! does not work here. 6 . . . .t.xf6 7 �S+ g6 8 �g6 bxg6 9 Wxg6+ Wf8 to .t.c4 We7 ; as 00 is no longer a problem I I 1Dc3 .g7 12 .xe4 ]lh4 1 3 f4 c6 0-1 , Magee­Svencisen. corr. 1991 ; and 6 .i.e2

202 3 d4

whilst common. is insipid: 6 .. . d6 7 tOe4 which transposes to S . . . d6 6 tOc4) 6 . . . dS (for 6 . . . d6? 7 -*.xf6 sce later) 7 f3 exf3 8 .xf3 c6 9 -*.c:l3 0-0 10 ()-(H) � Stummer-Svendsen, corr. 1991 .

Now, White has to make an im­portant choice. For some time the move of Pupol's, 6 tOe3 held the sway, but recently 6 llXI2 has be­come popular, avoiding the ex­change of queens. So, we have:

A 6 �! B 6 li)dl!?

Should White want to avoid the risk involved with either of these, then he can choose 6 tOc4, but. al­though it is very common in actual play, it is quite innocuous: 6 . . . .*.e7 (6 . . . -*.f5 7 &3 -*.g6 8 -*.c4 �bd7 9 0-0 c6 10 J.b3 .c7 I 1 c4 .*.e7 1 2 tOe3 0-0 � Booij-Westerinen, Dieren 1988) 7 -*.e2 (7 .*.xf6?! J.xf6 8 �S+ g6 9 .dS Navarro­Hammar, corr. 197 1 , 9 . . . tOe6! 10 c3 .e7 ;; 7 tOe3 keeps control of dS, and avoids the coming knight ma­noeuvre, but allows 7 . . . 0-0 8 .*.e2 .e8!? 9 0-0 bS 10 ltle3 c6 1 1 f3 dS! 1 2 fxe4?! b4! ; ChristofTerson­PetrofT, Munich 1936) 7 . . . 0-0 8 0-0 �! (exchanging dark squared bishops gives Black more room, and the king's knight will find a power­ful square on f4. Black can also

brin� his queen to the kingside with 8 . . .• e8!? 9 ltle3! [9 1Oc3 bS as per the Petroft' game, above] 9 . . .• g6 10 -*.h4 c6 I I c4 �6 12 tOe3 with a slight advantage to White, Drozdov-Nadanian, Krasnodar 1997) 9 -*.xe7 .xe7 10 1Oc3 �f4 1 I ltle3 c6 12 dS?! cS ; Carlton­Sim, England 1967.

6 .*.xf6?! provides Black with the bishop pair, 'free of charge', so to speak. 6 . . .• xf6 7 �g4 .g6 8 ltle3 J.e7 9 tOe3 c6 10 .*.e2 0-0 I I 0-0 dS already favours Black. Schueller­Bukacek. corr. 1989.

A 6 �

6 ••• dxe5 Black might just as well take this

piece: a) 6 . . . .*.e7? 7 .*.xf6 .*.xf6 8 �S+

g6 9 �g6 bxg6 10 .xg6+ Wd7 ( IO . . . WfB is also ho.peless, I I .*.c4 dS [ 1 1 . ...e7? 12 lOtS] 12 �dS .i.e6 1 3 �f6 -*.xc4 14 dS .e7 I S ()-(H) threatening to bring a white rook to the kingside, and winning quickly, Papp-Varga, Hungary 1980) I I � :t'8 12 .*.e2 ( 12 .f5+ We8 13 .xe4+ doesn't hold out any hope for Black. either) 12 . . . �6? ( 12 . . . cS?! 1 3 dxcS .*.xb2?? 14 .xd6+ is crushing, Blackbume-Pollock; 12 . . . .*.xd4 is a better chance, though 13 0-0-0 -*.eS 14 �7+ is still pretty grim) 13

.fS+ WC(; 14 lDe7+ 1 -0, it is mate in one, Rechel-Drill, Berlin 1998.

b) 6 . . . .*.fS? 7 .*.c4! (7 g4? dxeS 8 gxfS exd4 9 lOxc4 .e7 ; Salnins­Tomson, corr., or even 9 . . . .*.b4+ 10 c3 .e7) 7 . . . dxeS 8 dxeS .xdl+ 9 lhdl lObd7 (9 .. . lOfd7 10 lDds lDa6 1 1 .*.xa6 bxa6 12 lOxc7+ Wf7 13 lDxa8 .*.b4+ 14 c3 .*.as I S b4 with advantage; 9 . . . h6? 10 exffi! [ 10 hffi gxffi 1 1 lDds might lead to no more than a draw after I I . . .lIh7 12 lDxffi+ We7 1 3 lOg8+) 1O . . . hxgS 1 1 lDds Wd8 1 2 1Ob6+ lDd7 1 3 lOxd7! with advantage, Camiol-Destrebecq, corr. 1984) 10 exffi lOxffi 1 1 0-0 c6 12 life I with a clear advantage, 8m«haud-Niemand, corr. 1994/95.

c) Destrebecq's 6 . . . c6!? might just be playable, 7 lOxe4 dxeS 8 .*.xffi gxffi 9 1ihS+ Wd7 (9 . . . We7!? may be stronger, 10 dxeS .as+ 1 1 c3 .xeS 1 2 1rh4, Canal Oliveras­Niemand, corr. 1999, and now 12 . . . .*.h6 would stop White from castling long) 10 0-0-0 ( 10 dxeS Wc7 1 1 exffi b6 with a piece for three pawns, Elburg-Des�becq, corr. 1991) 1O . . . Wc7 1 1 dxeS .e8! with the tactical point 1 2 .xe8 .*.h6+ 13 Wb I IIxe8 and if anyone is better, it is Black, Kravitz­Ruggeri Laderchi, email 1998

:., dIeS .Idl+ I IIIdl

1 ••• h6

3 d4 203

a) Black has an alternative in 8 . . . .*.g4?! :

a l ) 9 1ld2! h6 (9 . . . lObd7? 10 lObS 0-0-0 1 1 exffi gxffi 12 .i.f4 .*.cS 13 .*.xr;7 with advantage Schlenker­Stwnmer, corr. 1991 ; but 9 . . . lOfd7!? might be Black's best bet, 10 lOds 1Ob6 I l lDxc7+ WfT 12 h3 [ 1 2 .*.e3 lDcti! 13 lOxa8 { 1 3 .i.xb6? IIc8 equal} 1 3 . . . lOxa8 with .sood chances) 12 . . . .i.e6 13 lOxa8 '&a8 14 1ld4 1Dc6 I S lhc4 .*.fS 16 e6+? [a blunder, 16 lIe2 had to be played, but Black is not without chances here) 16 . . . Wg6 17 e7 .i.xe7 0-1 , Waldner-Downey, USA 1992) 10 .*.xffi ( 10 exffi!? i s untried, although play will be similar, e.g. 1O . . . hxgS 1 1 lOds WfT [ I I . . .Wd8 12 fxg7 .*.xg7 13 1Ob6+ lDd7 14 .*.e2!) 12 lOxc7 1Dc6 13 lOxa8 .i.b4 14 c3 .*.as I S b4 lOxb4 16 cxb4 .i.xb4 17 .*.c4+ Wg6 1 8 a3 with some advan­tage) 10 . . . gxffi 1 1 lDds WfT!? ( 1 1 . . .lDd7!? has been more success­ful: 1 2 lDxc7+ Wd8 13 lOxa8 fxeS 14 h3 .i.e6 IS .*.bS Wc8 16 0-0 lOffi 17 IIfd I ;t but Black won, Heemsoth-Stamer, corr. 1984 or 12 f3?! 0-0-0 13 fxg4 fxeS 14 .*.c4 .*.cS Melchor-Grobe, corr. 1989/9 1 ) 1 2 lOxc7 .i.b4 13 c3 .*.as 14 .*.c4+ Wg6 I S lOxa8 (in the original game in this line, de Wit-de Zeeuwe, corr. 1989, White contin­ued I S 1ld6!? lDd7 16 lOxa8 lha8 17 .i.dS lDxeS 1 8 .i.xb7 IIb8 19 .i.xc4+ Wg7 20 :'6?! .i.b6 21 b3?! lIe8 22 0-0 .i.e2 and Black had gained the ascendancy) IS . . . 1Dc6 16 h3! ( 16 b4?! lOxb4 17 cxb4 .*.xb4 18 exffi IId8 19 0-0 .*.xd2 20 1Dc7 <i>xffi with a level position, Tiemann-Elburg, corr. 1991) 16 . . . lOxeS (what else? 16 . . . .i.fS 17 g4 lDxeS 1 8 gxfS+ <i>xfS 19 .*.e2 [ 19 lIdS! is even stronger) 19 . . . lha8 20 IIgl leaves Black

204 3 d4

almost zero compensation for the exchange Melchor-�nechaud, corr. 1990, and others) 17 .t.d5 .t.f5 I S .t.xb7 lLki3+ 19 WfI IIbS 20 .t.d5, Melchor-Tiemann, corr. training game 1993, and now 20 . . . J.xc3 2 1 lIe2 lOf4 may cause White a few slight problems.

a2) 9 exf6 is another option, White will obtain two pawns and the bishop pair in return for the ex­change, 9 . . . J.xd I 10 Wxd I lLki7 (1O . . . gxf6? I I J.xf6 IIgS 12 lOxe4 lLki7? [ 1 2 . . . lOc6] 13 J.c4 1Ig6 14 lie I with advantage �nechaud­Svendsen, corr. 1991192; 10 . . . lOc6?! I I lOxe4 h6 1 2 J.h4 ()-()-{)+ 13 Wcl lieS 14 f3 a6 I S J.c4 g5 16 .t.g3 lOe5 17 J.b3 and the black pieces are unable to find good squares, Pape-MalmstrOm, corr. 1994195) I I lOXe4 gxf6 (Black can also flick in I I . . .h6 12 J.h4 first) 1 2 lOxf6+ lOxf6 1 3 J.xf6 IIgS 14 J.b5+ c6 I S lie 1+ Wd7 1 6 J.fI , �nechaud­Gaard, corr., 16 . . . lIeS and Black shouldn't lose this, although care has to be taken to stop the kingside pawns.

a3) 9 J.e2? is a blunder, though: 9 . . . J.xe2 10 lOxe2 h6 I I J.xf6 gxf6 +, Groen-Diepstraten, Hilversum 19S3.

b) S . . . lObd7?! 9 exf6 lOxf6 10 J.xf6 gxf6 I I lLki5 WfT 12 lOxc7 is also favourable to White.

9 J.xf6 pl6 10 lOd5 Wd7 Taking the unusual step of walk­

ing into a double check, but 10 . . . WfT? 1 1 lOxc7 .t.e6 12 lOxaS J.b4+ 13 c3 .*.as 14 b4 favours White, Schreyer-�echaud, corr. 1992.

I l lOb6+ Picking up the as rook. Whether

or not Black can keep the as-knight trapped is crucial for the assessment of this line.

a) I I e6+?! is interesting, but can

now be considered doubtful: I I . . .WdS ! ! (Black gives White a second chance to use his discovered check! The older move I l . . .Wc6 is also reasonable: 12 e7 .t.g7 13 .t.b5+ [ 1 3 lOb4+!? Wb6 { 13 . . . �c5!? takes the king into uncharted, and very murky, waters: 14 J:ld5+! Wxb4 IS c3+ Wa4 16 J.e2 and Black seems to be in some trouble, but after 16 . . . b5! 1 7 .t.xb5+ waS I S .t.c6+ Wb6 19 J.xaS lLki7 Black is OK! } 14 lLlds+ repeats] 13 . . . Wxb5 14 lOxc7+ Wc6 I S eS=.+ IIxeS 16 lOxeS .t.f8 1 7 lOxf6 [though tempt­ing, 1 7 lidS? only succeeds in trap­ping the rook: 17 . . . lLki7 I S lOxf6 J.b4+ 19 c3 J.a5 winning, Melchor-Oren, corr. 1997] 17 ... J.f5 IS 0-0 J.g7!? [ I S . . �7 19 lOxd7 J.xd7 may be even simpler, the two bishops should certainly be more than a match for a rook] 19 lLlds .t.h 7, with reasonable play for Black, Melchor-Heap, corr. 1993) 12 .t.c4 ( 12 lOb6+?! .t.d6 13 lOns .t.xe6 must be good for Black, the white knight will never escape from its prison) 12 . . . .t.xe6 13 lOf4+ .t.d7 14 lOg6 .t.g7 I S lOxhS J.xhS 16 O-O?! (even after the superior 16 .t.d5 lOc6 17 J.xe4 We7 Black is better) 16 . . . f5 and Black has a solid material advantage, �nechaud­Clarke, corr. 1994/95.

b) Perhaps the rare 1 1 �t'6+!? is White's best: 1 1 . . .We6!? (attacking the key eS-pawn, and probably stronger than 1 1 . . .Wc6?! when the king is misplaced, 12 �xc4 .i.e6 13 f4 [ 1 3 a4!? is also awkward, 13 . . . &6 14 .i.e2 Wb6 I S f4 �b4 16 c3 lOcIs Goerlinger-Stamer, corr. 19S2, when 17 fS! .i.xfS I S as+ Wc6 19 .i.O lines up dangerous threats along the h i -aS diagonal] 1 3 . . . .i.e7 14 .i.e2 lOcI7 Schopp­meyer - Traut, East Germany 19S7, and I S �3! threatens both .i.bS+, and .i.O+, and looks like it should win) 12 llXIs ( 1 2 �c4!? � [ 1 2 .. .'�xeS!? 1 3 .i.d3 .i.g7 14 0-0 leaves the black king out-on-a-Iimb, where can it escape to?] 1 3 f4 Wf5!?, attacking with the active king! If Black doesn't try this then White has three connected pawns, and good minor pieces, for his sacri­fice) 12 . . . Ah7 and now Lein gives 1 3 lLlf4+ WfS 14 g3 ( 14 :dS Ah8!) 14 . . . �6 I S .i.h3+ WgS which looks fine for Black, but protecting the important eS-pawn by 1 3 f4! is much more worrying, 1 3 . . . exO 14 gxO �6 ( 14 . . . WxeS? IS f4+ We6 16 .i.h3+ wins) I S f4 threatening .i.h3+, so I S . . . tOb4!? 16 �f6, but it looks dangerous for Black.

1 I ••. Wc6 Il �xa8 Not 12 �cS? .i.cs and . . . :xcS. 1l ... fxe5! Black wants to follow-up with

.. . .i.d6, solidifying his position. 12 . . . b6 is possible, although there

is a suspicion that White will gamer too many pawns: 1 3 .i.c4 Wb7 14 �b6! ( 14 �c7 Wxc7 IS ext'6 lOcI7 16 f7 .i.g7 1 7 c3 �S with some play, Stumme .... Magcc, corr. 1991) 14 . . . Wxb6 IS ext'6 l&i7 16 f1 .i.g7 17 0-0 with a worse black structure than above, and the three extra white pawns proved too

3 d4 205

important in Stamer-Kozlov corr. I 994I9S; 12 . . . .i.e6?? 13 .i.bS+ 1-0, SCnechaud-Sadeghi, corr. 1991/92.

13 .i.c4 .i.d6

14 .i.d5+ 14 a4!? is an idea, menacing

.i.bS+, and asking Black how he will complete his development, 14 . . . .i.g4!? (this seems to work tac­tically, the only sensible alternative is 14 . . . Wd7 intending to bring the black king to the safe e7-square, e.g. I S .i.dS We7 16 .i.xe4 1Oa6 [fol­lowed by . . . c6, cS-bishop moves, then . . . LaS] 1 7 0-0 c6) I S .i.dS+ (if I S ILi2 then I S . . . Wd7 looks right) I S . . . Wd7 16 O! ( 16 .i.xb7? .i.xdl 1 7 Wxdl �6) 16 . . . exO 17 gxO .i.f5 I S :gl .i.xc2? lettin&. the knight out, ( I S . . . �6 19 :g7+ fi:Je7 20 .i.xb7 :bS is level) 19 :g7+ WdS 20 :c I J.xa4 2 1 lfu.c7 and victory is not far off, Sos-Peters, corr. 1982 .

14 ••• Wd7 15 .i.xe4 Destrebecq's IS c4 is best an­

swered by I S . . . �6! 16 .i.xc4 b6 and .. . .i.b7xa8.

15 ••• c6 This is successful, but I S . . . 1&6

and IS . . . We7 are also possible. 16 0-0 We7 17 I:td3 �6 18 :fdl

:d8 19 :93 M 10 .i.b7 The tactic 20 :g7+ leads no­

where: 20 . . . WfS 2 1 :h7 lfu.c4 22

206 3 d4

lIhS+ We7 23 lhdS WxdS 24 0 We7! 25 fxe4 b5 and . . . �b7xaS.

10 ••• lOe6 11 c3 lO15 II b4?! This combination turns out badly

for White. but if 22 �c2 �e6 and . . . lhaS.

1l ••• lOxh7 13 l1g7+ �6 14 lIxb7 Wg6

Trawing the rook. 15 lIxd6+ lIxd6 16 Jlb8 �e6 17

lIb8 l1d7 Making sure that the knight never

leaves as; now the active black king decides the game.

18 .4 Wf6 19 b4 Jlg7 30 as �d5 31 13 Wf5 31 lId8 We4 33 lIea b5 34 f4? (34 Wfl ) 34 •• .lIxg3+ 35 Wbl Wxf4 36 lI18+ WI4 37 lIn 0-1

Piefrusiak-Hector. Swedish ch 1984.

B 6 lOdl!?

Krantz's speciality. which has re­cently been much played by Sireta. This is the logical development of the previous variation. White bra­zenly leaves his knight en prise once again. but this time Black is unable to exchange the queens.

6 ••• du5 The acid test of the sacrifice is al­

ways its acceptance. and in this case avoidance brings its own set of problems:

a) 6 . . . �fS? 7 �c4! dxe5 S dxe5 lObd7 9 exffi gxffi 10 .i.f4 lOb6 1 1 .e2 .e7 1 2 0-0-0 0-0-0 13 lOb3 �g7 14 .i.a6!. threatening .i.xb7. with a strong attack, Krantz­Tiemann. corr. 1991 .

b) 6 . . . �e7?! 7 �xffi (this is not as effective as in A but nevertheless White gets a very dangerous initia­tive) 7 . . . �xffi S -..5+ g6 9 lOxg6 bxg6 10 .xg6+ Wf8! ( I0 . . . Wd1?! 1 1 lOxe4 �e7 12 1Oc5+!? dxc5 13 dxc5 �xc5 14 0-0-0+ �d6 15 �b5+ 1Oc6 16 .g7+ .e7 17 .xhS! .g5+ I S Wbl .xb5 19 lIhe I liJe7 20 .g7 with a continu­ing attack. Lundkvist-Schreyer. corr. 1984) 1 1 �c4 .e7 12 0-0-0 Diepstraten-JacOon. corr. 1994/95. when 12 . . .• g7 100ks safest.

7 dxe5 .d5! Since the first edition. Krantz's

suggestion of 7 . . . e3 !? has received a number of tries. S fxe3 .d5 9 �xffi (9 exffi .xg5 10 iOe4? would then encounter 1O . . .• xe3+) 9 . . . gxffi 10 exffi �d6!? (there is absolutely no consensus here as to Black' s best. 10 . . .• g5 1 1 .0 �c5 12 n+ WfS 13 0-0-0 1Oc6. unclear. Sireta­Grobe. corr. 1994/96. and after 10 . . .• e5 1 1 lOc4! .xffi 12 "'5+ WdS 13 0-0-0+ �d7 14 �d3 Black had regained one pawn. but the at­tack along the d-file is a problem. Sireta-Melchor. corr. 1994195.

however, Sireta's suggestion, lO . . . lbd7!? is also possible, 1 1 .Ac4 .xg2 1 2 -...5+ [ 1 2 ti+ �e7 1 3 .0 .xO 14 1Ox0 lOb6 I S .i.b3 .Ae6 is clearly better for Black] 12 . . .• g6 13 f7+ �d8 not allowing White to settle) 1 1 .Ac4 .eS 12 f7+ �e7 13 .0 lOc6 14 0-0-0 .Ae6 IS :ltfl :afS and White's attack is insufficient, Sireta-Kozlov, corr. 1994/95.

8 exf6 Better than 8 .Axffi?! gxffi 9

1IhS+ .f7 lO lOxe4 .Ae7 ( 10 .. . lbd7) 1 1 .Ae2, Diepstraten­SCnechaud, corr. 1995, when 1 1 . . . (H) is simplest.

8 .•.• xgS , lilu4

, •••• eS This allows White to maintain

some sort of attack. 9 . . .• aS+!? is simple and effective: 10 c3 and now 10 . . . .Ae6 (not 10 . . . .Af5? 1 1 fxg7 .Axg7 12 1Whs+ �e7 1 3 .gS+ � 14 g4 lOc6, Krantz-Svendsen, corr. 1989, when I S .xfS+ is certainly very good for White) 1 1 f4! (prepa� ing gS for the knight, others are in­ferior, 1 1 .e2 gxffi!? 12 lOxffi+ [ 12 lbd6+ �e7] 12 . . . �e7 -1+; 1 1 .Ad3 lbd7 12 (H) [ 12 fxg7 .Axg7 1 3 (H) 0-0-0, objectively speaking, White has only two pawns for his piece and nothing else, not even a lead in development, therefore he should

3 d4 207

lose] 12 ... 0-(H) 13 b4 .dS 14 .e2 lOxffi is hopeless for White, Solinas­Viola, Arzignano Magistrale 1997) I I . . .pffi! (the idea is that 1 1 . . .• f5?! allows 12 .Ad3 .xf4 13 .e2 .Ag4 14 fxg7 .i.xg7 I S lOffi+ �d8 16 .xg4 .xffi 17 0-0-0 with a strong attack) 12 lOxffi+ �e7 ( 12 . . . �ti 13 lOe4 �e7 seems simple enough) 13 lOe4 lbd7! ( l 3 . . . �g7 14 .id3 lOc6 IS lOgS when, instead of I S . . . h6? 16 lOxe6 �e6 1 7 .g4+ �f7 1 8 .i.c4+ �f8 19 0-0-0 with a winning light­squared attack, Sireta-Krantz, corr. 1994/95, there is I S . . . .i.g8) 14 lOgS 1Wb6 1 5 �e2 l:le8 16 .d2 �d8 sneaking away to safety on the queenside, Sireta-Budovskis, corr. 1994/96.

10 17+ A tempo-gaining pawn sac, but 10

.O!? has also been played: lO . . . .i.e6 1 1 0-0-0 lOc6 12 �a6!? lbd4 13 J:lxd4! .xd4 14 .i.xb7 l:lb8 I S .i.c6+ �d7 16 ti+?? ( 16 .i.xd7+ .xd7 1 7 lOgS is far from clear) 16 . . . �d8 0-1 , Terblanche-Stamer, corr. 1993 as 17 lOc3 .i.xc6 18 .xc6 �d6 leaves Black a rook up.

10 ••• �xf7 10 . . . �e7?! offers no obvious ad­

vantage: 1 1 .i.d3 lOc6 (the 1 l . . .�ti 12 .i.c4+ of Sireta­Downey, corr. 1994/96, transposes to lO . . . �f7) 1 2 (H) .i.e6 13 l:lel :d8 14 h4 .dS? IS lOgS :d6 16 c4 the e-fiIe pressure proving decisive in Sireta-Stamer, corr. 1994/97.

1 1 .i.c4+ �e6 This now seems best, although I

still believe that 1 1 . . .�g6!? is worth a try: 12 (H) lOc6 13 .0 ( 13 lOg3?! �cS 14 l:le l?? .i.xf2+! 'the boot is on the other foot' ! I S �f2 .cS+ 16 l:le3 l:lfS+ 17 �g 1 .xe3+ winning, Evans-Doyle, corr. 1993; I think that 13 �d3! is D,,'"St, 13 . . . �h6

208 3 d4

[ 1 3 . . . �t7!? 14 .i.c4+ �g6 might be a draw] 14 .d2+ gS I S f4 with some attack) 1 3 . . . �e7 14 �d3 �h6 I S g4 gS 16 �g3 .f4 17 .dS �xg4 1 8 :ae 1 :ad8 0- 1 . Melchor­Diepstraten. COlT. 1991 .

However, bringing the king to the open e-file is dodgy, 1 1 .. .�e8?! 1 2 0-0 .xb2? ( l 2 . . . �e7 1 3 lIe l is less clear) 1 3 .dS �7 14 �gS .a; I S �f1 lIg8 16 lIfe l+ �e7 1 7 �6+ cxd6 1 8 .xg8+, regaining material with a strong attack, Sireta Niemand, COlT. 1994/95.

ll .fJ+

1l···�18 At the time of my fmt book on

the Latvian. it was not clear which move was strongest, but now impor­tant new resources have been found in the main line, the text is clearly best. 1 2 . . . �e8?! is too difficult for Black: 1 3 �xe6 .xe6 14 0-0-0 �e7 I S libel (training its sights on the �r black king) I S . . . �6 ( l S . . . Wh6+ is similar, but leaves the queen offside, 16 �b 1 �6 17 �S Jld8 18 � hd1+ 19 .xdl .xh2 20 g3 ! with threats, Krantz­Melchor, COlT. 1989) 16 �S (to come to the key square e6, but 16 �3 has also been effective: 16 . . . �6+?! [ 16 . . .• t7] 1 7 �bl .xh2? 1 8 �S Jld8 19 J:td3 lId7 20 .g4 1 -0, Huizer-Owens, COlT. 1996;

16 �6+?! is worse, as although the black queen is lost, Black has plenty of -pieces as compensation, 16 . . .• xd6 1 7 J:txd6 cxd6 1 8 .g3 �fS unclear) 16 . . .• xa2 1 7 � .a1+ 1 8 �d2 �b4+ 19 �e2! (an improvement on 19 �e3?! from the same two protagonists, ending in a draw) 19 . . .. a6+ �

S.xb2?! 20

�c7+ �e7 2 1 + �e8 22 �b4 .xb4 23 �f1 + �7 24 �S+ with advanta�e, Krantz­Elburg, COlT. 1991) 20 Jld3 .i.d6 2 1 �f1 �S? (2 l .. .�7) 22 heS �xeS 23 �S+ �e7 24 .xeS � 25 �c7 .al+ 26 �e2 1 -0 Sireta­Vitols, COlT. 1994/96.

13 .d3!

Lining up menaces along the a2-g8 diagonal. 1 3 .i.xe6+?! allow­ed Black to fend off the attack in Svendsen-Destrebecq, COlT. 1991 : 1 3 . . .• xe6 14 0-0-0 �6 IS �gS? .xa2 16 JldS .a1+ 17 �d2 �b4+ 18 �d3 .a6+ 19 c4 lIfS +.

13 ••• llX6! This now seems to be best, al­

though 13 . . . �e7 is also a satisfac­tory reply here, 14 f4 (a key deflection) 14 . . .• f5! ( l4 . . .• aS+ I S c3 � 16 0-0-0 �6? [ 16 . . .• c6!? 17 �S .xcS 18 �xe6+ �f8 might just hold on] 1 7 :he 1 Jld8 [ l 7 . . . �f8 18 �gS] 1 8 .xd8+! .i.xd8 19 lIxd8+ Wf7 20 �gS+ 1-0,

Krantz-Svendsen, corr. 1 992) I S ltlgS .i.b4+! (to force the c-pawn to 'undefend' the queen, if I S . . .• xd3 then 16 .i.xe6+ WfS 1 7 cxd3 gives White a small edge, Krantz­Svendsen, corr. 1 992) 16 c3 .xd3 1 7 "'xd3 .i.dS I S cxb4 ltlc6 1 9 0-0 h6 and Black was almost equal, Sireta-Heap, corr. 1 994/97.

1 3 . . . c6!1 is another sound idea, 14 0-0-0 (the point is that 14 f41! is now met � 14 . . . .i.xc4 I S .xc4+ .dS) 14 . . .t&6 I S f41! ( I S lIhel is stronger) I S . . .• xf4+ 16 �bl .i.xc4 1 7 .xc4+ .t7 I S .d4 1Ob4 19 b3 ltldS consolidating, Sireta-Downey, corr. 1994/96.

14 f4

14 •• ..i.b4+!

3 d4 209

To force c3, for the same reason as before. Black was forced to part with his queen after 14 . . .• fS!1 I S ltlcS! lieS 16 .xfS .i.xc4+ 1 7 �d2! ( 1 7 �1 g6 I S .gS lIe2+ [ I S . . . .i.e7?? 19 lIhel �g7 20 lIxe7+ lIxe7 2 1 lIel with advan­tage, in Elburg-Destrcbecq, corr. training game 1993] 19 � ltld4+ 20 �g3 lllfS+ 2 1 �h3 .i.e7 22 .g4 lIe3+ 23 g3 .i.e2 wins!) 1 7 . . . g6 I S .g4 .i.xcs 19 Jlae I J:tdS+ 20 �c1 the undeveloped hS-rook is a problem, Sireta-Elburg, corr. 1994196.

14 .. .• aS+1 I S � ! ! (to free the el-square for a rook) I S . . . lIeS 16 ltlgS ltlds 1 7 lOxe6 lOxe6 1 8 lie I cM7 19 lieS � 20 .fS+ �gS 2 1 .xe6+ lIxe6 22 .i.xe6 mate, Krantz-Svendsen, corr. 1991 .

15 cl .15 1 6 csb4 lId8 1 7 .e1 ltld4! 18 lIdl "! 19 .i.se6 .se6 lO �

The queen versus two rooks posi­tion is terrible for White, although there is little choice, as after 20 .d3 ltlc6 he drops material.

lO ••• ltls:e1 11 lIsd8+ �e7 11 lIsb8 ltld4

With a winning position for Black, Sireta-De long. corr. I 994/9S.

12 3 ltJc3 Mlotkowski's variation

I e4 e5 1 �rJ f5 3 llX3

White sensibly continues his de­velopment. Although such a move can hardly be critical for the evalua­tion of the Latvian Gambit. never­theless Black has to take care. The move was first analysed by the American Mlotkowski in the BCM of 19 16. Essentially, Black has two main replies:

A 3 ••• �r6 B 3 ••• rxe4

There are a couple of popular al­ternatives that really belong to books on other openings:

a) 3 . . . d6, once again, a Philidor's Defence is reached 4 d4 fxe4 S 1Oxe4 �fti (S . . . dS 6 lOxeS dxe4 7 W'hS+ g6 8 �xg6, unclear, is the principal variation) 6 1Oxfti+ (6 .i.gS!? .i.e7 7 1Oxffi+ gxfti?! [but 7 . . . .i.xfti 8 .i.xfti gxfti =] 8 .th6

:g8 9 .i.d3 e4 10 .i.xe4 f5 1 1 .i.dS :g6 1 2 .d2 c6 1 3 .i.b3 :xg2 14 0-0-0 didn't look too good for Black in Melchor-Downey, corr. 1993) 6 . . . gxfti! (always capture towards the centre! 6 . . .• xfti? 7 .i.gS .e6 8 dxeS �7 9 .i.d3 .te7 10 0-0 with advantage Krantz-RiUenhouse, corr. 1993) 7 dxeS dxeS! 8 �! (the endgame after 8 .xd8+ Wxd8 9 .i.e3 .tg4 10 h3 .te6 I I 0-0-0+ �7 12 �bl .td6 1 3 � �e7, is very pleasant for Black, Sebastian­Hector, Aviles 1989) 8 . . . .i.e6 9 .0 .i.dS? (9 . . . c6 10 � .i.g7 I I .i.e3 0-0 unclear) 10 � .i.e7 1 1 .te3 c6 12 0-0-0 �7 13 IlxdS! ! cxdS 14 W'hS+ �f8 IS .tbS with a powerful attack, Krantz-Downey, corr. 1991 .

b) 3 . . . �6 The Three Knights' Game 4 d4 fxe4 (4 . . . exd4?! speeds up White's development: S 1Oxd4 IOxd4 6 .xd4 d6 7 .tc4 �fti 8 .i.gS .i.e7 9 0-0-0 with advantage, Svendsen-Melchor, corr. 1988) S �xeS �fti 6 .i.c4 .e7 (a rather clumsy move to have to make, but 6 . . . dS?! meets 7 �xdS! �xdS 8 W'hS+ g6 9 IOxg6 bxg6 [9 . . . �fti?? 10 .i.f7+! Wxf7 I 1 �S+ leads to mate] 10 .xg6+ with a strong aIL tack) 7 .tgS!? (7 .i.f4) 7 . . . �xeS 8 dxeS .xeS 9 .i.xfti gxfti 10 0-0 c6? I I 1Oxe4 dS 12 �3 .i.e6 1 3 LI .d6 14 �xdS! �f7 IS W'hS+ �g7 16 1lxe6 .xe6 1 7 �3 1 -0, Jensen­Melchor, corr. 1992.

A l ... lilf6

4 .A.c4 .

The most direct, but there is also: a) 4 exfS e4 this will probably re-

semble variations in Chapter 10, al­though some options are restricted by the inclusion of lilc3 and .. . lilffi (Anyway, better to avoid 4 . . . d6?! S d4 e4 6 lilh4 .A.e7 7 g4! [7 .A.gS 0-0 8 g4, Scherbakov-Lomakin, Novo-1ruznetsk 1997, is not so strong, 8 . . . dS! with reasonable .play] 7 . . . lilxg4 8 .xg4 .A.xh4 9 WhS+ �7 10 llg1 with a clear advantage love-Van Willigen corr. 1 998; De­strebecq suggests 4 . . .• e7!? but it is unnatural, S .A.c4 00 6 0-0 d� [6 .. . dS? is too weakening, 7 lIel �bd7 8 .A.b3 e4 9 d3 and Black has severe problems along the e-file, Oren-Canal Oliveras, corr. 1997] 7 d4 [7 lIel .A.xfS 8 d4 lilbd7 {8 . . . e4 9. dS! } 9 dxeS lilxeS is also better for White, Wittmann-Gunderam, corr. 1965] 7 . . . e4 8 dS! cS?! 9 .A.d3 blowing the e-file open, Oren­Kozlov, corr. 1994/97) S lilgS! (S liles and S lild4 transpose into Chapter 10, and S .e2 likewise, e.g. S . . . dS 6 d3 .A.b4 7 dxe4 O-O!? 8 .A.d2 .A.xc3 9 .A.xc3 lilxe4 ; Landgraf-Magee, corr. 1991 , whilst S lilh4 is a playable alternative: S .. . dS [S . . . .A.b4] 6 d4 .A.e7 7 .A.gS [7 g4] 7 . . . 0-0 8 g4 [8 g3 has also been tried, the fS-pawn can be defended

3 lilc3 Mlolkowsk; 's var;al;on 21 1

by .A.h3, 8 . . . h6?! {8 . . . lilc6} 9 .A.f4 lOc6 10 lilg6 with advantage, Spiegel-Magee, corr. 1990] 8 .. . ltlc6 [but a good King's Gambit plan would be 8 . . . OO!, and ... lile8] 9 �bS .d6 10 a3 with some advantage, Gaard-Behrendorf, corr. 1990192) S . . . dS 6 d3 h6?! (6 . . . .A.b4!? 7 dxe4 .e7 might be playable) 7 lile6 .A.xe6 8 fxe6 .A.b4 9 .A.e2 (9 dxe4! 0-0 [or 9 . . . d4 10 a3] 10 eS lDe4 1 1 e7 with advantage) 9 . . . 0-0 10 0-0 1le8 1 1 dxe4 .A.xc3 1 2 bxc3 lilxe4 1 3 .A.O lilc6 �, Melchor-lensen, corr. 199O.

b) 4 lilxeS?! d6 S lilO (S lilc4 fxe4 6 d4 .A.e7 is very comfortable for Black, as well) S . . . fxe4 6 lild4 dS 7 d3 .A.b4 (7 .. . cS looks like a good move, 8 lilb3 exd3 B 8 a3 .A.xc3+ 9 bxc3 0-0 10 .A.e2 cS 1 1 lilb3 lilc6 12 0-0 b6 ; Nolden­Stummer, corr. 1991 .

c ) 4 d4 fxe4 S dxeS!? ( S lilxeS d6 [but Black can also try S . . . .A.b4!?] 6 .A.gS! transposes to the 3 d4 varia­tion; S lilgS exd4 6 .xd4 00 7 lilgxe4 dS 8 lilxffi+ .xffi 9 .xft) gxffi= Tudor-Milner Barry, Hastings 19S2IS3, by transposition) S . . . exO 6 .xO .e7! (6 . . . �e7? 7 .A.c4! dS 8 exft) with advantage, Gaard­Tiemann, corr. 1990) 7 .A.f4 (7 .e2 lilg8 8 lilds .d8 should be good for Black, despite his lack of develop­ment) 7 . . . d6 8 0-0-0 dxeS 9 .A.xeS .xeS 10 .A.c4 GUrd-Kozlov, corr. 1989, and now 10 . . .• gS+ will win at a canter.

4 ••• fu4 a) The best reply to 4 ... lilc6 is

probably S d4 (S lOgS?! dS) with the continuation S . . . .A.b4!? (S . . . exd4 6 lilxd4 lilxd4 7 .xd4 fxe4 8 .A.gS c6 9 0-0-0 dS 10 lilxe4 with advantage Svendsen-Gaard, corr. 1992, is a lit­tle too easy for White) 6 lilxeS (6 dxeS! lilxe4 7 0-0) 6 .. .• e7 7 0-0

2 J 2 3 �3 Mlotkowski 's variation

.i.xc3 8 bxc3 fxe4 9 .i.g5?! �xe5 10 dxe5 .xe5 1 1 .i.xffi gxf6 12 .i.d5 Wd8. and Black managed to hang on to his ill-gotten gains. Svencisen-Krantz, corr. 1 990.

b) 4 . . . �xe4!? has received some attention since I flfSt wrote about the Latvian: 5 �e5 (or 5 �e4 fxe4 6 �e5 d5!? [or even 6 . . .• f6] 7 �5+ g6 8 �g6 bxg6 9 .xg6+ tran�sing to the main line) 5 . . .• e7 6 .i.f7+!? (6 d4 �xc3 7 bxc3 d6 8 .i.a3 cS ;) 6 . . . Wd8 7 �! .xe5 8 d4 .d6 9 .i.f4 .c6 10 .i.xc7+ .xc7 1 1 �c7 Wxc7 is very promlSmg for White. Niemand-Destrebecq. corr. 1999.

5 �xe5 d5 The alternative 5 . . .• e1?! is under

a cloud at the moment: 6 d4! (dis­covered by M.Johnson. and much stronger than 6 �g4 c6 7 �xf6+ gxf6 unclear; or 6 �f7? d5) with:

a) 6 . . . �6 7 .i.f4 (more incisive than 7 .i.t7+ Wd8 8 fuc6+?! bxc6 9 .i.b3 d5 10 .i.f4 .f7 1 1 0-0 .i.d6 ;t; Sturnmer-sen�haud. corr. 1991 ) 7 . . . d6 8 .i.t7+ Wd8 9 .i.b3 .i.e6 10 d5! �e5 1 1 dxe6 �g6 1 2 .i.g5 h6 1 3 � .e8 14 .i.xf6+ gxf6 1 5 �f6 .e7 16 �e4 with a big plus. Johnson-Clarke. corr. 1 989/90.

b) 6 . . . d6 7 .i.f7+!? Wd8 8 .i.g5 c6 9 .i.b3 .i.e6 10 0-0 .i.xb3 1 1 axb3 Wc7 12 .d2! .e6 1 3 � �bd7 14 f3 exf3 15 lIael .f7 16 lbf3 with advantage. Krantz-Grivainis. corr. 1991 .

c) 6 . . . exd3? 7 0-0 dxc2 8 .d2 � 9 lie I .i.e7 10 .g5 d5 1 1 .xg7 lIfS 1 2 lOd3 .d6 1 3 .i.f4 1 -0. Strelis-Alloin. corr. 1992.

d) 6 . . . c6 7 .i.g5 d6 (7 . . . d5?! 8 .i.xd5 �d5 9 �d5 with advantagc-Johnson) 8 .i.f7+ Wd8 9 .i.b3! with advantage Krantz­Grivainis. corr. 1 99 1 .

6 �xd5!

Any retreat would be tantamount to an admission of defeat.

6 ••• �xd5 Black has no real choice. as

6 . . . .i.e6 7 �f4 (7 � .xf6 8 .i.xe6 .xe5 9 .i.c8 is also winning easily. Krantz-Svendsen. corr. 1994195) 7 . . .• d6 8 �e6 .xe5 9 0-0 .i.d6 10 g3 should be winning. Borsdorff-Callinan. corr. 1968.

7 "'5+ 16 8 �xg6 bXl6 A.sain Black must face the music.

8 .. .Il�f6?! seems to lose: 9 .e5+ .i.e7 10 �h8 �6 1 1 .i.f7+ WfS 1 2 .g5 .i.d6 1 3 .i.b3! (the best way of extracting the knight, 1 3 �6+? We7 14 .g7 .is 1 5 .xiS+ WxfS 16 d3 �b4 with some chances. Franck-Druke. corr. 1989) 1 3 . . .• e7 14 �t7 .i.c5 ( l4 . . . �g4?! 1 5 h3 .xg5 16 �g5 �f6 17 0-0 .i.f5 1 8 f3 ! winning. Krantz-Druke. corr. 1990) 1 5 �h6 � and now as well as 16 0-0. Strelis-Destrebecq. corr. 1991 . 16 �g8 �c2+ 1 7 Wdl is im­mediately decisive.

9 .xl6+ The coming attack resembles that

of chapter eight. 9 .xh8?! .f6 should be OK for

Black. 10 .xf6 ( 10 �7?! .i.e6 1 1 d3?! .i.b4+ 12 WfI? [ 1 2 We2] 1 2 . . . .i.c5. soon winning. Arguelles­Barros. corr. 197 1 ) 1O • . . �xf6 1 1 d3 exd3 12 .i.xd3 Wf7 1 3 0-0 .i.d6 14 .i.d2 �6 and the black knights are quite the equal of the rook and pawns. Leeners-Alderden. co�. 1982.

9,..Wd7 If instead 9 . . . We7!? then 10 d3!?

(10 .g5+ We8 1 1 .e5+ .e7 12 .xh8 �f4 might give some counterplay) 10 . . . e3 1 1 Wg5+ •

Niemand-Canal Oliveras. corr. 1998. and now I I . . .We6!? might be worth trying.

10 .i.xd5

White has an interesting �ssibil­ity here: 10 .f5+ WOO I 1 .i.xdS+ .xd5 (or I l . . .Wb6!?) 1 2 .ffi+ .i.e6 1 3 .xh8 lLlci7 unclear. .

IO •••• e7 1O . . . c6 is currently more popular.

I I .i.xe4 Wc7 (the Icing reaches a relatively 'safe' square) 1 2 d4 .i.d6 1 3 .i.g5 .f8 14 h4 .i.g4 with a sharp material imbalance. Melchor­Destrebecq. corr. 1997).

I I .se4 With four pawns for the piece.

White can consider playing the end­game here.

a) 1 1 d3?! exd3+ 1 2 .i.e3 dxc2 (Black can 80 into another ending: 12 . . . c6! 1 3 .i.b3 d2+ [ 1 3 . . . dxc2] 14 Wxd2 .d6+) 1 3 .xc2 .e5 14 0-0-0 .i.d6 'h-'h. Melchor­Krongraf. corr. 1 99 1 . again some­what premature one feels, as both sides have reason to continue.

b) 1 1 .i.xe4 :tt6 1 2 .f5+ Wd8 1 3 .f.3 1&6 14 d3 lLlci4 1 5 .dl lIg6 is unclear. MalmstrOm-Collins. corr. 1999.

1 1 • • .Ilh4 Il Wu7+ .i.u7 IJ g3 lIb5 14 .i.e4 1Oc6 15 f4 .t.f6 16 cl Wd6 17 d3 .i.d7 18 .i.e3 &De7 19 0-0-0 c6 In-In. Evans-Koser. corr. 1993. although there is plenty of play left at the end.

B 3 ••• fse4

3 �3 M/otkowsk; 's variation 2 J 3

4 �u5 The natural move 4 �e4?! is sig­

nificantly worse: 4 . . . �ffi (4 . . . dS is also good. as 5 �e5? loses a piece to 5 . . .• e7. and 5 1&3 e4 6 1Od4?! �ffi 7 .i.e2?! cS ! 8 .i.b5+? Delavenne-Destrebecq. France 1977. 8 . . . Wt7 winning a piece) 5 �ffi+ .xffi 6 d3 .i.c5 7 .i.e2 0-0 8 0-0 �6 9 c3 d5 10 .i.g5 .g6 1 1 .i.h4 h6 1 2 .i.g3 .i.d6 ; Danielsson­Hector. Umea 1990.

4 •••• (6 Once again. the obvious move

4 . . . �ffi?! brings problems. as 5 �g4! wins a pawn (5 .i.c4 is not bad either and transposes into A; 5 d4 d6 [but 5 . . . .i.b4 6 .i.e2 d6 is safer] 6 .i.g5 transposes into Chap­ter 1 1 ) 5 . . . lOc6 (there is f\Othing bet­ter. 5 . . .• e7 6 �f6+ gxR; 7 .g4 with advantage. Spiegel-Landgraf, corr. 1 99 1 ; 5 . . . .i.b4 6 i.tlxffi+ .xffi 7 �e4 .e5 8 .e2 with advantage Destrebecq-Kozlov. corr. 1982186; and 5 . . . .i.e7 6 �ffi+ .i.xffi 7 �e4 0-0 8 .i.e2 [8 �ffi+ .xffi 9 .i.c4+] 8 . . . d5 9 �ffi+ .xffi 10 0-0 �6 1 1 c3 d4 1 2 d3 � Tiemann-Elburg. COP". 1990) 6 �ffi+ (6 .i.c4 is also good) 6 . . .• xffi 7 �5+! (7 �e4? loses in this instance to 7 . . .• e5 8 .e2 lLlci4) 7 . . . g6 8 �xe4? (as Franz Destrebecq points out, 8 .dS! wins the e-pawn under somewhat more favourable circumstances: 8 . . . � 9 .xe4+ .e7 10 .xe7+ .i.xe7 1 1 Wd l with advantage) 8 . . .• e7 9 ee2 lLlci4 10 .d3 .i.g7 1 1 c3 dS 1 2 cxd4 dxe4 ;. Schirmer-Koser. corr. 1993.

One other move worth noting is Doyle's idea 4 . . . 86?! with the possi­ble follow-up 5 .i.c4 (5 �xe4! .e7 6 d4 looks more wonying. i.e. 6 ... �6 7 .i.g5 .e6 [7 ... eb4+? 8 c3 exb2?? 9 1&4 ±] 8 f4 with advan­tage) 5 . . .• g5 6 fOf7 .xg2 7 lift c6 8 i.tlxh8 d5 9 .i.e2 .i.g7 10 d3 .xh2

2 J 4 3 �3 Mlotkowski 's variation

1 1 dxe4 d4 with a certain amount of compensation.

White faces another major choice: BI 5 r4!? B2 5 d4 Otherwise, we transpose directly

into chapter five if White chooses 5 llk4!, and another possibility is 5 lLlg4 although the knight is not pr­ticularly well placed here: 5 . . .• g6 (5 . . .• e6?! 6 lLle3 lLlffi 7 .i.c4 �, Evans-Schirmer, corr. 1992, is too awkward) 6 d3 (6 lLld5 .i.d6 7 d3 c6 8 lLlc3 .i.b4 9 .i.d2 d5, hitting the exposed knight on g4, Walker­Dravnieks, corr. 1955156, is fme for Black; 6 lLle3 lLlffi 7 d3 .i.b4 leaves Black better placed than in Chapter 3) 6 . . . d5 7 h3 .i.xg4?! (7 . . . .i.b4) 8 hxg4 exd3 9 .i.xd3 .e6+ 10 .e2 with an edge, 8orsdorfT-Strautins, corr. 1969.

BI 5 r4!?

This newish move has become very popular in the last few years, and has scored nearly I ()()o1o!

5 ... uD And not 5 ...• xf4? 6 �5+ g6 7

lbxg6 hxg6 8 .xh8, Niemand­Clarke, corr. 1999.

6 lLlJ:D .e6+ The queen is misplaced on the

open f-file, 6 .. . c6?! 7 d4 d5 8 .i.e2 .i.d6 9 0-0 .d8 10 lLlg5 lLlffi 1 1 .i.h5+ g6? 1 2 .e2+ �d7 1 3 .i.84+ 1-0, De Jong-Van de Velden, 1998.

7 .i.e2 lLlr6 8 0-0 .i.e7 9 d4 d5 10 lIel

10 .i.f4!? may be superior, 10 . . . c6 ( IO .. . �!?) I I .d2 0-0 12 Ilael gives White a useful lead in devel­opment, Oren-Grivainis/Hayward, corr. I997.

10 ••• 0-0 10 ...• d6 1 1 .i.b5+!? c6 12 .e2

cxb5 13 lbxb5 .d8 14 .i.f4 lLla6 I S c4 is dangerous, but Black managed to defend, Krantz-Budovskis, corr. 1994/95.

1 1 .i.c4!? .d6 12 lLlJ:d5 lLlJ:d5 13 .i.J:d5+ .J:d5 14 lIJ:e7

Krantz-Jackson, corr. 1994/95, when Black could have obtained a useful initiative by 14 . . . lLlc6! I S lIe3 ( I S Jlxc7?! .d6) 1 5 . . . .i.g4.

82 5 d4

5 ••• udJ

Both S .. . d6? 6 lDds .d8 7 �S+ g6 8 lOxg6 hxg6 9 .xh8 WfT 10 .i.c4 .i.e6 I I Wh7+ .i.g7 1 2 �f4! .i.xc4 13 .xg6+ WfS 14 dS 1 -0, Krantz-Hayward, corr. 1990, and S . . . .i.b4? 6 .g4 .i.xc3+ 7 bxc3 file7 8 .i.gS dS 9 .i.xft) .i.xg4 10 .i.xg7 1 -0 Diepstraaten-lzuel, corr., leave Black with difficult problems.

6 f4!? Somewhat sharper than the alter­

native 6 lOxd3. although this, too. has been successful in practice: 6 . . . d6 (6 . . .• e6+? is worse: 7 .i.e2 lOft) 8 0-0 .i.e7 9 :el g6 to .i.0 .fT I I .i.h6 d6 1 2 lObS, with advantage, Wundt-Dravnieks. corr. 1992; but 6 . . . c6!? 7 .i.e2 [7 .e2+ .i.e7 8 lOc4 .e6 9 .i.f4 dS 10 lDd6+ Wibe-Svedenborg. Norway 1978. gives an edge] 7 ... dS 8 0-0 [8 .i.f4!?] 8 ... .i.d6 9 .i.hS+ g6 10 :el + liJe7. Letbaili-Kosten, Reims 1994, is nothing special for White) 7 .e2+ .e6 8 lDdS .xe2+ 9 .i.xe2 �6 10 �3f4 Wd8 I I .i.d2! c6 1 2 .i.xa6 bxa6 13 .i.aS+ Wd7 14 �7 :b8 I S lOfe6 �ft) ( l S ... .i.e1?! 16 o-o-o .i.ft) 1 7 b3 We7 1 8 :hel with advantage. Downey-Svendsen. corr. 199 1 ) 16 0-0-0 dS 1 7 c4? ( 1 7 lOxfS+ :XfS 1 8 :he I avoids getting the knights trapped) 1 7 .. . .i.b4 1 8 .i.xb4 :Xb4 1 9 cxdS cxdS 20 Wb I Wd6 2 1 :hel :c4 22 :c l :xc l+

3 1Oc3 Mlotkowski 's variation 2 J 5

23 Wxc I file8 24 lOxe8+ :Xe8 0-1 . Svendsen-Clarke. corr. 1994/95 .

6 ..• dul There is no reason Black should­

n't take this pawn, but 6 .. . .i.b4 is also possible, 7 .i.xd3 (7 .xd3 is not so logical. Tiemann-Melchor, corr. 1990. continuing: 7 .. . d6 8 .5+ 1&6 9 lOxc6 .i.xc3+ 10 Wt2 a6 unclear) 7 .. . d6 8 O-O!? hc3 9 bxc3 file7 (I now feel that it is best not to take the piece. Keres gives 9 . . . dxeS?! 10 fxeS � (perhaps it is better to save this check by 10 . . .• e6!? stopping .i.c4, and de­fending the kingside: in particular fT, 1 1 h3 lDd7!? 12 �S+ g6 13 Wh4 with strong pressure for the piece] I I Whl lOe7 12 .i.a3 .e6 [ 12 . . . .i.e6?! 1 3 �S+ g6 14 Wh4 leaves Black in severe trouble. 14 . . . cS IS .ft) :g8 16 :abl 1-0. Niemand-Vitols. corr. I 994/9S] 13 �S+ g6 14 Wh4 with an annoying threat of infiltration on ft). Niemand-MalmstrOm, corr. 1994198) 10 .i.a3!? ( 10 �g4 .i.xg4 1 1 .xg4 �bc6 [ 1 1 . . .lDd7 12 .i.e3 .xc3 1 3 :abl 0-0 14 .i.xa7 with advantage, Littleboy-Alberts, corr. 1977n9] 12 :hI :b8 [ 12 . . .• xc3? 1 3 .i.b2 .cS+ 14 Whl .dS I S .i.xg7 1 -0, Svendsen-Magee, COlT. 1990] 1 3 .i.b2 0-0 14 :bel �fS with Black's superior structure off­setting the bishops, Nolden-Kozlov, corr. 1990) 10 . . . 0-0 1 1 :bl � 12 lOxc6 ( 12 .e2!?) 1 2 . . . lOxc6 13 �S g6 14 � :b8 ( l4 . . . lDd8! is worth serious consideration. menac­ing to trap the white queen) IS c4 .g7 16 Wh4 .i.fS 1 7 .i.b2 with advantage. Krantz-Kozlov. corr. 1992.

7 .xcl .i.b4 7 . . . 1&6 is an alternative, 8 00!?

.d6 9 ec4 �ge7 10 .i.c4 lOxdS I I lOxc6+ .e6 1 2 files .i.b4+

2 J 6 3 �3 Mlotkowski 's variation

( 1 2, ..Il�ffi!?) 1 3 .i.d2 .i.xd2+ 14 Wxd2 lDffi I S .d4 dS ( l S . . . cS !?) 16 Jlael , Krantz-Strautins, COlT. 1990, and now 16 . . . 0-0 is unclear.

8 .A.c4 8 .A.e3?! d6 9 0-0-0 .i.xc3 10

.xc3 dxeS 1 1 .i.bS+ liX6 12 fxeS

.g6 13 :Itn lDge7 14 :0 .A.e6 and Black eventually came out on top, Krantz-Heap, COlT. 1993.

8 ••• lDe7 9 0-0

9 •• ..i.sc3 10 bsc3 10 .A.f7+ misplaces the black

king, but does nothing for White's minor piece arrangement, I O .. .'�)d8 1 1 bxc3 d6 12 :dl ltXI7 1 3 lDxd7 • xf7!? 14 lLles .ffi and Black's

solid structure and pawn extra give him the advantage, PIlpe-Sireta, COlT. 1994196.

10 ••• d6 IO . .. dS !? 1 1 .i.d3 .i.f5 1 2 :bl b6

1 3 c4 0-0 14 .i.b2 .d6 I S :bdl gives White sufficient compensa­tion, Stamer-GUrd, COlT. 1994I9S.

l l lDfJ 1 1 .A.a3!? is more fun, and almost

certainly stronger, 1 1 . . . dxeS ( 1 l . . ..A.fS looks safer) 12 fxe5 _g6 ( 1 2 . . . �+ 13 Whl .id7 14 :adl is also unpleasant for Black) 13 .xg6+! and instead of 13 . . . hxg6? 14 ,j.f7+ winning, Melchor-Reinke, COlT. 1994/95, 1 3 . . . lDxg6 14 .A.f7+ Wd7 1 5 :fd I + Wc6 16 .i.d5+ may be playable for Black.

1 1 ••• lDbc6 11 .A.e3 .A.f5 12 .. . .A.e6 13 .A.d3 0-0-0 14 ICxI4

ltXiS!? is also playable, Stamer­Sireta, COlT. 1994196.

13 "'3 0-0-0 14 lDd4 lDaS IS "'S lDu4 16 .u4 dS

16 . .. c5 !? 17 .cS b6 .. . with Black for preference,

Stamer-Elburg. COlT. 1994195 .

13 Unusual third moves for White

We have already examined a large variety of possible third moves for White, but there are some players who disdain theoretical correctness and who always search for a differ­ent route. So, we have five less usual possibilities:

A 3 d3 B 3 b4 C 3 c4 o 3 g4 E 3 "el F 3 b3

The only common factor is that none of the moves otTer White any advantage!

A I e4 e5 l ll)O f5 3 d3

This is a solid, if passive, choice but it is the type of reply weaker players are likely to make over-the­board. So much so, that this move is played more often than all the other moves in this chapter put together!

3 ••• 1Oc6 Black can just as well play the

prudent 3 . . . d6 4 �3 ll)a; 5 g3 (in

the game Nicholas-Simons, England 199 1 , White took his passive start to its logical conclusion: 5 .i.d2?! .i.e7 6 .i.e2 0-0 7 h3 �6 8 00 f4 9 ll)g4 lLxI4 1 0 �f6+ .i.xa; 1 1 .i.g4 .i.h4 1 2 �S?! c6 1 3 �3 .a; 14 �2? f3 -+) S . . . .i.e7 6 .i.g2 (6 h4!? c6?! 7 .i.h3 fxe4 8 J.xc8 .xc8 9 dxe4 h6 10 .i.e3 bS 1 1 a3 as 12 b3 �6, unclear, Vandervort-Diemer, Holland 1983) 6 . . . �6 7 0-0 0-0 8 exf5 .i.xfS 9 ll)h4 .i.g4 10 f3 .i.e6 1 1 f4 dS!? 1 2 ll)f3 e4 13 dxe4 d4!? ( 1 3 . . . �e4 14 ll)xe4 dxe4 IS .e2 .i.dS is a good altematwe) 14 ll)gS .i.c4 I 5 1&2 .i.cs with a certain amount of compensation for the pawn, Davies-Vasyukov, Orested 1990.

4 � a) Giving up the centre by 4

exfS!? is possible, 4 . . . dS 5 .e2 (5 .i.gS .i.e7 6 .i.xe7 ll)gxe7 is favo� able to Black; and 5 g4 fails to hold on to the pawn after S . . . hS) S . . .• d6!? (if instead S . . . .i.d6 then 6 ll)xeS .i.xeS 7 f4 .i.xf5 8 fxeS .e7 leads to a rough equality, but the

218 Unusual t�ird moves/or White

pawn sac 5 . . . .*.xf5!? is interesting, e.g. 6 lOxe5 1lld4 7 .dl .fti 8 lDO .*.g4) 6 b3 .*.xf5 7 .*.b2 d4 8 ll)bd2 0-0-0 9 0-0-0 with chances to both sides, Geertse-Diepstraten, corr. 1982.

b) Whilst White was horribly mauled in the game Dousse-Hector, Geneva 1986, after 4 c3 lDft) 5 .*.g5?! d6 6 lDbd2 h6 7 .*.xft) .xft) 8 g3 g5! 9 exf5 .*.xf5 10 .e2 0-0-0 1 1 0-0-0 d5 12 .*.g2 .*.c5 13 lDb3 .*.b6 14 l%hfl as! I S d4? (a grue­some opening of the bl -h7 diago­nal !) 1 5 . . . .*.h7 16 ll)el? ( 16 lDXe5 lOxe5 17 dxe5 .f5 1 8 �d2 Ilhe8 19 Illd4 is the best chance) 16 . . . exd4 1 7 lOxd4 hd4 1 8 cxd4 lDb4 19 b3 1lhe8 20 �2 .c6+ 2 1 �d2 .a6 22 a3 lle2+ 0- 1 .

4 ••• lDf6

5 ed5! Now is the right moment for this

move. Others: a) 5 .*.e2 d5!? (5 . . . .*.b4 6 exf5 d6

7 .*.d2 .*.xf5 8 a3 .*.xc3 9 .*.xc3 0-0 is pleasant for Black, Hultin­Nyholm, corr. 1994) 6 .*.g5 .*.b4 7 exd5 .xd5 8 .*.xft) .*.xc3+ 9 bxc3 pft) 10 ll)h4 0-0 I I 0-0 ll)e7 12 c4 .d6 1 3 f4 lDg6 14 lOxg6 bxg6 I S fxe5 fxe5 ;, Bezilko-Hector, Cappelle-Ia-Grande 1988.

b) 5 .*.g5 .*.b4 6 .*.e2 d6 7 0-0 .*.xc3! 8 bxc3 fxc4 9 dxc4 .*.e6 10

lOd2 0-0 ; Svendsen-Destrebecq, corr. 1991 .

5 ••• d5!? This is OK. but S ... d6 is also pos­

sible, 6 d4 (6 lDh4!? .*.e7 7 g4 lOx&� 8 .xg4 .*.xh4 9 .xg7 .*.ft) 10 .g4 Illd4 1 1 �5+ �d7 is un­clear) 6 . . . exd4 7 lOxd4 lOxd4 8 .xd4 .*.xf5 9 .J.d3 .*.xd3 10 .xd3 c6 1 1 0-0 �e7 12 ll)e2 d5 1 3 Illd4 .d7= Lemer-Bareev, USSR 1986, by transposition .

6 d4! As Destrebecq points out, holding

on to the pawn by 6 lDh4!? allows the combination 6 . . . 1lld4!? 7 g4 lDxg4 8 .xg4 .xh4 9 .xh4 lDo+ 10 We2 lOxh4 1 1 lOxd5 .*.d6 equal, i.e. 1 2 ft) .*.g4+ 13 �e3 0-0 with goOd chances against the exposed white king. Nevertheless, 6 . . . .*.e7 looks good, 7 .*.g5 0-0 8 g3 h6 with good ylay, and 6 . . . d4 has its points, as 7 t&4 .*.b4+ forces 8 lOd2 (as 8 .*.d2? loses to 8 . . . lOxc4).

6 ••• exd4 6 . . . e4?! 7 ll)e5 .*.xf5 8 .*.b5 .A.d7

9 .*.p with advantage. 7 Illl:d4 lDI:d4 8 .l:d4 .*.1:15 9

.*.15 .J.n2 White's advantage is minimal af­

ter 9 . . . .*.e7 10 0-0-0 c6 1 1 .e5 .*.g6 12 Ile I , but exists.

10 Ilct .*.&6 1 1 .*.1:16 .1:16 12 .e3+

IZ ••• Wt7? A tactiaal miscal<:u1ation, Bla<:k is

fine following 1 2 . . .• e7 1 3 .xe7+ Wxe7! 14 �dS+ Wet7 I S �<:7 Ab4+ 16 Wet 1 :lad8 17 AbS+ We7+ and his bishops may even give him the advantage.

Il /Oxd5 Ab4+ 14 /Oxb4! :lW 15 h4+ n 16 0-0

The l'Oint. 16 •• .:lxe3 17 1xe3 A15 18 ... Winning a further pie<:e for the

queen, Taimanov-Zai<:hik, Lenin­grad 1989, Interestingly, this game, like at least one other in this sedion, Ktually started as a Dut<:h Defen<:e: 1 100 fS 2 d3 lOffi 3 e4 eS.

B l b4?!

A move mlK:h loved by the Fren<:h player SCn6chaud.

l ... fxe4

Unusual third moves for White 219

Alternatively, 3 . . . .i.xb4 4 lOxeS IOffi (4 . . .• e7 S �S+ g6 6 IOxg6 .xe4+ 7 h2 IOffi 8 -gS �S8 9 .xffi hxg6 10 £.3 .xgl 1 1 .eS+ Wd8 and White fOr<:e5 a perpetual, SCn6chaud-Clarke, <:orr. 1990) S A<:4 ee7 6 Ab2 d6 7 00 .xe4+ (7 .. . £.S?! 8 eS dS?! 9 0-0 IOg4 10 AxdS ..... 1 1 h3, beating off the attKk with a <:rushing advantage, SCn6chaud-Logunov, <:orr. 1990) 8 ee2 AaS 9 1OC3 .xe2+ ex<:hanging queens, when White's wmpensation for the pawn is not suffident, SCn6chaud-S<:l:!reYer, <:orr. 1990.

3 . . . d6 4 d3 /Offi S bS £e7 (pawn grabbing by S .. . fxe4!? 6 dxe4 �e4 7 A<:4 <:6 is perfe<:dy viable) 6 10<:3 0-0 7 Ab2 fairly level, SCn6chaud­JKkson, <:orr. 1994I9S.

4 IOxe5 1Of6

S el!? Quite in keeping with the spirit of

the variation, otherwise: S 1Og4? Axb4 (S ... dS is also good, 6 1Oxffi+ .xffi 7 10<:3 .17 [is this nec:essary? 7 . . . Axb4!? is tempting, as after 8 IOxdS .xa l 9 IOx<:7+ Wd8 10 lOxa8 b6 Bla<:k will win the aB-knight by .. . Ab7] 8 :lbl as 9 a3 £d6 10 43 0-0 1 1 0 axb4 12 axb4, SCn6chaud-M�ee, <:orr. 1990, when 12 . . . exd3 13 .xd3 Axb4 is strong, for 14 :lxb4 .e7+ wins the ex­<:hange) 6 Ab2 £e7 7 �ffi+ Axffi

220 Unusual third moves for White

8 .*.xf6 .xf6 9 1Oc3, 8enechaud­Stummer. corr. 1990, 9 . . . c6! 10 �S+ ( 10 lDxe4?? .eS 1 1 d3 dS) 1O . . . g6 1 1 �6 dS with a clear Black advantage. S d3? d6 -/+ SCnCchaud-Littleboy, corr. 1990.

5 ••• d6 S . . . dS 6 d4 .*.d6 7 f4?! exf3 8

pf3?! 0-0 9 :gl .*.fS 10 .*.d3 .i.xd3 1 1 .xd3, SCnechaud­Svendsen, corr. 1990, I I . . .�bd7 and Black enjoys an edge, owing to his superior structure.

6 �4 d5 7 � �6 8 b5 � 9 d4 esd3 10 .*.sd3 .*.c5 1 1 0-0 0-0 Il .*.al .*.sal 13 �sal .d6 14 IOacl

8enechaud-Budovskis, corr. 1 990, when Black can exploit the weak­ened white queenside by 14 . . . a6! I S bxa6 b6.

C 3 c4!?

A speciality of Leo Diepstraten, in fact hardly anyone else has ever played this odd move!

3 ••• &e4 The simplest, although other

moves are also satisfactory: a) 3 . . . � 4 d4 exd4 S lDxd4

.*.b4+ 6 1Oc3 lDxd4 7 .xd4 .e7 8 .i.d2 �f6 9 eS?! d6 and Black will win the e-pawn, Diepstraten-Krantz, corr. I996.

b) 3 ... d6 4 1Oc3 1Oc6 (4 . . . �f6?! S exfS .*.xfS 6 d4 t4?! 7 �h4 .*.d7 8 .e2 winning the e4-pawn, Kveinis­Antoshin, USSR 198 1 ) S d4 � 6 dS (6 exfS lDxd4 7 lOxd4 exd4 8 .xd4 .*.xfS is equal) 6 . . . fxe4 7 dxc6?! exf3 8 .xf3 .*.g4 9 .d3 bxc6 and Black has an extra pawn, DiC))Straten-Strautins, corr. 1998.

4 �u5 �f6 5 1q4 Played to weaken Black's defenc:e

ofe4. 5 ... �.6! 6 M � 7 �d6+

.d6 8 d4 esd3 9 .*.sd3 9 .*.e3 is more accurate. 9 ••• �sd3+ 10 .sd3 .*.b4 1 1 .*.dl

.e5+ Play is quite equal, Diepstraten­

Downey, corr. 1997.

D 3 g4?

Actually, looking at this move re­minds me of a game I played against Mike Basman at Manchester once. I, a mere whippersnaP,PCf at the time, after opening 1 e4, replied to Basman's l .. .gS?! with 2 f4!? Tony Miles, who happened to be passing by at that moment, immedi­ately burst into a loud sustained laughter. I can only imagine how he would react to this idea of SCnCchaud's!

3 • • • &e4 4 �u5 d6 5 �4 d5 6 � d4 7 �f5

Black is just making pawn moves, while White is just using his knight!

7 ••• �f6! 7 ... .i.xf5 8 gxf5 1ih4 ;

SCn6<:haud-Koser, corr. 1993. 8 d3 p! I have no idea what is happening

after 8 . . . .i.b4+ 9 cl dxc3 10 bxcl .i.xf5 I I gxf5 exd3!? 1 2 cxb4 .d4 ( 12 . . .• dS 1 3 :gl .eS+ is also wild) 13 .i.xd3 .xa I 14 .e2+ �7 I S (H), SCn6<:haud-Ustinov, corr. 1992.

9 dxe4?! pf5 10 uf5 SCn6<:haud-Evans, corr. 1995, and

now IO . . .• e7+ I I .i.e2 �g4 12 .xd4 :g8 is a rout.

3 ••• fse4 4 .se4 d6 5 d4

Unusual third moves for White 22 J

S .i.c4 �ffi 6 .e2 e4? (6 . . . .i.e7) 7 (H)?! (7 lDc::3 .e7 8 �gS) 7 . . . dS 8 .i.b3 .i.d6 9 d4? .i.g4 � Parker­Destrebecq, corr. 1979.

5 ••• �f6 6 .el e4 7 .i.g5 d5 8 �fdl .i.e7 9 c4

With chances to both sides, Viaggio-Carlsson, corr. 1978 .

F 3 b3?!

Similar to 3 b4. 3 ••• be4 3 . . . �ffi 4 �eS (4 exf5!?) 4 . . . d6 S

lDc::4 fxe4 6 .i.e2 .i.e7 (6 . . . dS) 7 (H) (H) 8 .i.b2 lDc::6 9 �3 .i.e6 10 f3 .e8 I I fxe4 �e4 1 2 c4 :Xfl+ 1 3 .xfl .16 =1= Fragola-Laffont, COlT. 1992.

4 �u5 �f6 5 .i.bl d5 S ... d6 6 lDc::4 dS. 6 c4?! d4 7 d3? .i.b4+ winning,

Lebrun-Matwienko. corr. 1973.

Index of Variations

Chapter I I e4 e5 l ltlf3 IS 3 IOnS .f6 4 d4 d6 S �4 fxe4 6 1Oc3 .&6 7 13 nf3 8 .xf3 s . . . IOa; S . . . .i.e7 S • • • 1Oc::6

9 10ds 9 lObs 9 dS 9 .i.d3 .g4

10 .xg4 to .f2 10 .e3+

Chapter 1 I e4 e5 l ltlf3 IS 3 IOxe5 .f6 4 d4 d6 S ltlc4 fxe4 6 1Oc3 .16 7 .i.f4 ltla;

7 1Oe3 7 �e3 7 10ds 7 .e2 7 dS 7 g3 7 h3

S ltle3 8 .d2

7 7

13 I S 16 1 7 I S 2 1 22 22 26

30 30 30 33 34 39 40 42 44 45 46

Chapter 3 I e4 eS l ltlf3 IS 3 ltlnS .f6 4 d4 d6 S iOc4 fxe4 6 1Oe3 47 6 . . . c6 7 �c4 47 6 . . . 1Oc::6 SO

Chapter 4 6 . . .lh6 6 . . .• dS 6 . . . dS 6 . . .• n 6 . . . hS?

I e4 e5 1 li)f3 IS 3 li)xe5 .(6 4 d4 d6 s li)c4 he4 6 hl SS SS 59 64 65 f61

Index 223

Cb�ter 5 I e4 e5 2 /L}f3 IS 3 /L}xe5 .16 4 ltk4 fxe4 5 � 68 S . . .• g6 70 S . . .• e6 72 s . . .• n 6 �3 c6 7S

7 /L}xe4 78 7 d3 exd3 8 .i.xd3 dS 9 (H) 83

9 . . . .i.cS 86 9 . . . .i.d6 91

10 :e1 �7 91 I 1 �xdS! ·92 I 1 �!? 96

Cb.pter 6 I e4 e5 2 /L}f3 IS 3 /L}xe5 100 3 . . . lDc6 lOO

4 �c6 100 4 �S+ 107 4 d4 1 12

3 . . .• e7 1 1 3 3 . . . d6 1 14 3 . . . /L}ffi?! l i S

Cupter 7 I e4 e5 2 /L}f3 IS 3 .i.c4 1 1 7 3 ••• fxe4 4 /L}xe5 .C5 5 d4 .X12 6 �5+ g6 7 .i.f7+ Wd8 8 .i.x16 .xbl+ 9 We2

9 . . .• xc l 124 9 . . . c6 10 �3 1 29

1O . . . /L}ffi 130 10 . . . �c7 1 33 10 .. . e3 1 3S

1 1 /L}t7+ �c7 1 36 12 .gS! 1 37 1 2 .i.xe3 1 37

Cb.pter 8 I e4 e5 2 /L}f3 IS 3 .i.c4 he4 4 /L}xe5 d5 140 5 "'5+ C6 6 /L}xg6

6 . . . hxg6 140 7 .xg6+ 140 7 .xh8 �t7 8 .d4 .i.e6 1 5 1

9 .i.b3 I S3 9 .i.e2 1 56

6 . . . /L}ffi 159

Cb.pier 9 I e4 e5 2 /L}f3 IS 3 .i.c4 165 3 . . . b5 165 3 . . . /L}ffi 1 7 1 3 . . . �6 1 73

224 Index

Cbapter 10 I e4 e5 l lLlf3 15 3 ex15 e4 4 lOC5 lLlffi 5 .i.e2 �

5 . . . d6 5 . . . .i.e7 5 . . . d5? 5 .. . .i.c5?!

4 .e2 .e7 4 lLld4 4 lLlgl !?

Cbapter 1 1 I e4 e5 l lLlf3 15 3 d4 fxe4 4 lLlxe5 lLlf6 5 .i.g5 d6

6 lLlc3 6 lLld2

Cbapter 11 I e4 e5 l lLlf3 15 3 lDc3 3 . . . lllffi 3 . . . fxe4

4 lLlxe5 .ffi 5 f4!? 5 d4

Cbapter 13 I e4 e5 l lLlf3 15 3 d3 3 b4?! 3 c4!? 3 &.4? 3 .e2!?

1 76 1 76 1 78 1 83 1 85 1 87 1 88 192 194

199 200 202 206

210 2 1 1 2 13

2 14 2 14

2 1 7 2 1 7 ·2 19 220 220 22 1