Decision-making skills of high-performance youth soccer players: Validating a video-based diagnostic instrument with a soccer-specific motor response

This is the Published version of the following publication

Murr, D, Larkin, Paul and Höner, O (2021) Decision-making skills of high-performance youth soccer players: Validating a video-based diagnostic instrument with a soccer-specific motor response. German Journal of Exerciseand Sport Research, 51 (1). pp. 102-111. ISSN 2509-3142

The publisher’s official version can be found at https://link.springer.com/article/10.1007%2Fs12662-020-00687-2Note that access to this version may require subscription.

Downloaded from VU Research Repository https://vuir.vu.edu.au/42775/

Main Article

Ger J Exerc Sport Res 2021 · 51:102–111https://doi.org/10.1007/s12662-020-00687-2Received: 14 August 2019Accepted: 15 October 2020Published online: 26 November 2020© The Author(s) 2020

Dennis Murr1 · Paul Larkin2 · Oliver Höner1

1 Institute of Sports Science, Eberhard Karls University, Tübingen, Germany2 Institute for Health and Sport, Victoria University, Melbourne, Australia

Decision-making skills of high-performance youth soccerplayersValidating a video-based diagnosticinstrument with a soccer-specific motorresponse

Introduction

In team sports like soccer, a multidimen-sional spectrumofperformance factors isrequired to performat the elite level. Thishas been acknowledged by Williams andReilly (2000) who developed a heuristicmodel for the categorization of soccertalent predictors. The model identifiespotential talent predictors across fourcore areas of sport science, includingphysical, physiological, psychologicaland sociological characteristics. Whilethere seems to be an emphasis on phys-iological and physical characteristics inresearch and practice (Johnston, Wattie,Schorer, & Baker, 2018; Wilson et al.,2016), recently, there has been increasedinterest in the psychological attributes,such as perceptual-cognitive factors(Mann, Dehghansai, & Baker, 2017).Researchers have highlighted the im-portance of perceptual-cognitive factorsfor skilled performance, with findingsshowing highly skilled players possesssuperior decision-making, anticipationand situational probability proficiencycompared to their less-skilled coun-terparts (e.g., Lex, Essig, Knoblauch,& Schack, 2015; Ward, Ericsson, &Williams, 2013).

The focus of the current study isdecision-making as the cognitive perfor-mance factor. Causer and Ford (2014)define decision-making in sports asa cognitive process in which one uses

knowledge about a (current) situationto select an appropriate decision, basedon one’s perceived ability to executea context-specific motor skill. Froma sporting perspective, the ability tomake the correct decision during com-plex game situations, under high gamepressure and time constraints is a keycomponent of in-game performance(Höner, Larkin, Leber, & Feichtinger,2020). Thus, decision-making has beenshown to be an important skill, withseveral cross-sectional studies assess-ing decision-making performance anddemonstrating that decision-makingskills discriminate between skilled andless-skilled players in team sports (e.g.,Diaz, Gonzalez, Garcia, & Mitchell,2011; Lorains, Ball, & MacMahon, 2013;Woods, Raynor, Bruce, & McDonald,2016). With respect to soccer, RuizPérez et al. (2014) found Spanish clubplayers with international experiencedemonstrated better decision-makingperformance in comparison to locallevel players. Further, Höner (2005)found youth national players had su-perior decision-making skills comparedto local youth players, and additionallyolder players (i.e., U17 age group) hada significant decision-making perfor-mance advantage over younger players(i.e., U15 age group). While researchershave used an expertise approach tohighlight superior performance of ex-pert/elite players over novice/nonelite

players (e.g., Ruiz Pérez et al., 2014) re-search is scarce within talent promotionprograms (e.g., regional association oryouth national teams; youth academies)with high-performance level (e.g., elite)players. This statement is emphasized ina recent meta-analysis which exploredcognitive functions measurements withperformance level as the moderator vari-able (Scharfen & Memmert, 2019). Theresults indicated, high-performance levelathletes had superior performance, witha small to medium effect size, comparedto low-performance level athletes. Whilethere is a plethora of studies examiningknown group differences, a potentialreason for the lack of research examin-ing homogenous samples could lie in thefact that it is more difficult to find largeeffect sizes for discriminating athletesof a similar ability (Bergkamp, Niessen,den Hartigh, Frencken, & Meijer, 2019).

Despite the fact researchers have usedcross-sectional approaches to discrimi-natebetweenhigh-and low-performancedecision-makers, Murr, Feichtinger,Larkin, O’Connor, and Höner (2018)highlighted in their systematic reviewthat there is a significant gap in theliterature concerning empirical evidencerelated to thepredictive valueofdecision-making assessments. According to thisreview, only one investigation has useda video-based assessment to examinethe predictive ability of decision-makingskills in soccer. O’Connor, Larkin, and

102 German Journal of Exercise and Sport Research 1 · 2021

Williams (2016) demonstrated a largesignificant effect in discriminating be-tween selected and nonselected playerswithin an Australian talent developmentprogram. However, it should be notedthe prognostic period for this studywas very short (i.e., the selection at theconclusion of the data collection), withfurther research required to investigatelonger prognostic periods. Despite thispositive result,more research isneeded todevelop valid and reliable measures thathave a strong predictive value to assistthe identification of high-performancelevel talent.

In addition to the lack of assessmentwithin high-performance level groups,a further restriction of the current de-cision-making literature is the limitedunderstanding of the potential per-formance differences between playingpositions. From a physical perspective,researchers have found physiologicaland anthropometric differences acrossplaying positions. For example, Rago,Pizzuto, and Raiola (2017) revealed mid-fielders and defenders completed morehigh-intensity running than forwards,and Boone, Vaeyens, Steyaert, VandenBossche, andBourgois (2012) found cen-tral defenders are taller and heavier thanmidfielders and wing defenders. There-fore, it is possible that due to specificgame-related roles of players in differentpositions (e.g., make goals, organizethe build-up), in addition to physicalcapabilities, playing position may alsoinfluence decision-making ability (De-prez et al., 2015; Pocock, Dicks,Thelwell,Chapman, & Barker, 2019). An initialinvestigation by Höner (2005) attemptedto address this issue and demonstratedthat midfielders make better decisionscompared to defenders and forwards.However, the results could be attributedto the situations used in the video sceneswhich presented more offensive deci-sions specific to midfield performance.Therefore, research is needed to addressthis limitation by attempting to measuredecision-making performance in dif-ferent game contexts, such as build-up(i.e., wing and central defense situations)and offensive (i.e., midfield and forwardsituations) game-based decisions.

Traditional diagnostic instrumentsused to examine decision-making, suchas video-based tests, provide advantagesin test execution or methodological con-trol and have demonstrated the abilityto discriminate skilled and less-skilledplayers (e.g., Roca, Williams, & Ford,2012). However, a limitation of previousinvestigations is the video presenta-tion stimuli which is often presentedfrom a third-person perspective (i.e.,television broadcast perspective; e.g.,O’Connor et al., 2016). It has been sug-gested however, when developing sport-specific assessments, that researchersconsider increasing the representative-ness of a task (Bonney, Berry, Ball, &Larkin, 2019). Therefore, there could bea benefit of presenting the video stimulusfrom a first-person perspective whichrepresents a more realistic perceptualenvironment compared to broadcastfootage.

A further limitation of the currentdiagnostic instruments is the method ofresponding to the video presentation,with participants providing a verbal,written or button response (e.g., Larkin,O’Connor, & Williams, 2016; Ward& Williams, 2003). Currently, it re-mains unclear if these nonsport-specificresponses correspond to real game situa-tions and performance (vanMaarseveen,Oudejans, Mann, & Savelsbergh, 2016).Overall, the extant research lacks studiesexamining decision-making competenceon video-based measures which incor-porate a sport-specific motor response.To address this limitation, Hagemann,Lotz, and Cañal-Bruland (2008) utilizeda video-based decision-making train-ing tool with a soccer-specific motorresponse. Using a similar visual-mo-tor response (i.e., players had to passthe ball against different targets in thevideo), Frýbort, Kokštejn, Musálek, andSüss (2016) investigated the influence ofvarying exercise intensity on decision-making time and accuracy. Althoughthese investigations provide the foun-dation for incorporating soccer-specificresponses within laboratory-based de-cision-making training, to date no di-agnostic instruments assess decision-making skills using a more represen-tative task (e.g., participants dribble

with the ball while watching the videostimuli and then execute their decisionby passing to a player in the video).Travassos et al. (2013) reaffirmed thisissue and emphasized that an expertiseeffect is more consistent if participantshave to execute sport-specific actions inexperimental studies. Hadlow, Panchuk,Mann, Portus, and Abernethy (2018)address these topics in a new classifi-cation framework to modify perceptualtraining in sport and emphasize the im-portance of high sport-specific stimulusand response correspondence.

Present study

With respect to their meta-analysis,Travassos et al. (2013) suggested fur-ther research is required to develop andexamine the impact of decision-mak-ing instruments which integrate sport-specific perceptual (e.g., first-personperspective) task constraints and sport-specific (e.g., passing or shooting) skillexecutions. Therefore, this study aims todevelop and evaluate a valid first-personperspective video-based diagnostic in-strument which incorporates a soccer-specific motor response to the decisionprocess and, thus, to provide a morerepresentative task.

To address this aim, three objectiveswere pursued: In order to ensure a scien-tific sound assessment, the reliability ofthe diagnostic instrument was evaluated(Objective I). The focus of this studywas on the criterion-related validity ofthe diagnostic instrument (Objective II).Here, both diagnostic and prognosticvalidation methods were used to ex-amine positive relationships betweenthe diagnostics’ results and appropriateperformance factors (i.e., age groupsin middle to late adolescence, playingtime in official matches of the currentseason, future draft in youth nationalteam) resulting in three hypotheses:

Diagnostic validity:

H1 Decision-making skills improveacross adolescent age groups (i.e., U16;U17; U19) Specifically:

German Journal of Exercise and Sport Research 1 · 2021 103

4 H1a: U17 soccer players demonstratebetter decision-making skills thanU16 players

4 H1b: U19 soccer players demonstratebetter decision-making skills thanU16 players

4 H1c: U19 soccer players demonstratebetter decision-making skills thanU17 players

H2 Players who played more minutesin official matches of the current sea-son demonstrate better decision-makingskills than players with less minutes.

Prognostic validity:

H3 Future youth national team play-ers demonstrate better decision-makingskills than nonselected players.

Finally, in an explorative analysis itwas examined whether decision-makingperformance is influenced by playing po-sition (Objective III).

Methods

Sample and design

The study sample consisted of 86 youthacademyplayers, born between 1996 and2001, from a professional German soc-cer club. The players competed in thehighest German youth league, and thusbelong to the top 1% of German playersfor their age groups (i.e., U16, U17, andU19). At the firstmeasurement point theplayers were 15–19 years of age (median[Mage]= 16.7 years, standard deviation[SDage]= 0.96) and were tested over a 3-year period (i.e., near the end of seasons2014/15 to 2016/17), resulting in threemeasurement points (T2015, T2016, andT2017). Over the 3-year period, a totalof 140 data points were collected. Due tothe nature of professional youth academyselection processes (i.e., some players aredeselected, and new players are recruitedto the academy)which resulted in not ev-ery player being tested at all three timepoints (. Table 1).

To examine the three hypothesizes,the study samplewas separated into threesubsamples. To assess diagnostic crite-rion-related validity (Objective II), po-tential differences between the indepen-

Abstract

Ger J Exerc Sport Res 2021 · 51:102–111 https://doi.org/10.1007/s12662-020-00687-2© The Author(s) 2020

D. Murr · P. Larkin · O. Höner

Decision-making skills of high-performance youth soccer players.Validating a video-based diagnostic instrument with a soccer-specific motor response

AbstractObjectives. This study aimed to developa valid video-based diagnostic instrumentthat assesses decision-making with a sport-specificmotor response.Methods. A total of 86 German youth aca-demy players (16.7± 0.9 years) viewed gamesituations projected on a large video screenand were required to make a decision bydribbling and passing to one of three targets(representing different decision options).The test included 48 clips separated intotwo categories: build-up (bu) and offensivedecisions (off). Criterion-related validity wastested based on age (i.e., U16, U17, andU19), playing status (i.e., minutes played inofficial matches of the current season) andin a prospective approach relating to futureyouth national team status (i.e., selected ornonselected). Finally, it was investigatedwhether decision-making competence wasinfluenced by playing position (i.e., defendersvs. midfielders vs. forwards).Results. Instrumental reliability demonstratedsatisfactory values for SCbu (r= 0.72), andlower for SCoff (r= 0.56). Results showedthe diagnostic instrument is suitable fordiscriminating between playing status (SCbu:Φ= 0.22, p< 0.01; SCoff: Φ= 0.14, p< 0.05)and between younger (U16) and older

players (U17>U16 in SCbu: Φ= 0.24 and SCoff:Φ= 0.39, p< 0.01; U19>U16 in SCbu: Φ= 0.41and SCoff: Φ= 0.46, p< 0.01); however, therewas no difference between U17 and U19players. Furthermore, the predictive value ofthe test indicates that future youth nationalteam players make better decisions withrespect to the build-up category (SCbu:Φ= 0.20; p< 0.05), whereas playing positiondid not significantly influence decision-making competence.Conclusion. Results indicate the video-baseddecision-making diagnostic instrument candiscriminate decision-making competencewithin a high-performance youth group.The outcomes associated with nationalyouth team participation demonstrate thepredictive value of the diagnostic instrument.This study provides initial evidence to suggesta new video-based diagnostic instrumentwith a soccer-specificmotor response can beused within a talent identification process toassist with assessment of decision-makingperformance.

KeywordsFootball · Talent identification · Adolescence ·Perceptual cognitive factors · Athleticperformance

dent variable age group (H1) were ex-amined splitting the sample into threeage categories (U16: N= 41, U17: N= 55and U19: N= 44; H1a–H1c). A furtheranalysis of the criterion-related validitywas conducted on the independent vari-able playing status (H2) that was deter-mined based on minutes played in of-ficial matches (i.e., U19/U17 GermanYouth Bundesliga and U16 Oberliga).Median split procedure was utilized toseparate players into two categories: firstteamregular (i.e., whoplaymoreminutesthan median; N= 70) and reserve play-ers (i.e., who play less minutes than themedian; N= 68). Further, to examine thepredictive value of the decision-makingdiagnostic instrument relative to futureyouth national team status (H3), data ofthe respective first measurement point

of all 86 players were captured. Play-ers who participated at least in one Ger-man youth national team training coursein subsequent seasons (i.e., 2015/2016;2016/2017; 2017/2018)were identified asselected (N= 16), with all other playersidentified as nonselected (N= 70). Fur-thermore, with respect to the explorativeanalysis (Objective III), players were sep-arated by playing positions as defined bytheir respective coaches (i.e., defenders[DF, N= 55]; midfielders [MF, N= 61];and forwards [FW, N= 24]).

Decision-making diagnosticinstrumentStimulus materials Decision-makingcompetence was assessed using a newlydeveloped soccer-specific video-baseddiagnostic instrument. To create the

104 German Journal of Exercise and Sport Research 1 · 2021

Table 1 Datapointscollectedwithrespectto age group andmeasurement timepointAgegroup

Measurement point Total

T2015 T2016 T2017

U16 13 12 16 41

U17 21 17 17 55

U19 15 14 15 44

Total 49 43 48 140

diagnostic instrument, more than 300dynamic on-ball decision-making situ-ations, recorded from the first-personperspective of the ball carrier at differentpositions on the pitch (i.e., forward,midfield, wing defense, central defense;. Fig. 1) were designed and filmed aspart of a pilot study (Dieze, 2015).The footage of each situation moved likea player in possessionof the ball andwerereviewed by a panel of expert coaches(i.e., one UEFA pro-level, one UEFA A-level and two UEFA B-level). During thereview process, a round table forum washeld, whereby the panel discussed theoutcome of each clip (i.e., best correctdecision), until 100% agreement wasreached. Following this detailed evalua-tion by the expert panel, 30 video sceneswere selected as the most realistic gamesituations and were thus used for thestudy (24 testing scenes; 6 familiarizationscenes).

To ensure a more comprehensive di-agnostic instrument, the 24 testing videoscenesweremirrored (i.e., the samescenepresented from the opposite direction),creating a final diagnostic instrument of48 game situations. Each situation wasthen classified as either offensive deci-sions (i.e., situations in forward andmid-field positions where players create scor-ingopportunities) ordecisions that occurin the build-up of a game (i.e., situationsin defensive positions such as wing andcentral defense where players start offen-sivemoves). For testing, the video sceneswere projected on a 2.76mwide× 1.50mhigh video screen and were presentedin four video blocks of 12 videos (i.e.,forward; midfield; wing defense; centraldefense) completed by all participants.During the test, each video scene was5–6s in duration and had three possiblepassing options; however, in the forwardcategory players had the choice between

two passing options and one option toshoot at goal. Between each trial, a blackscreen remained for 6 s. Each clip com-menced with the first frame of the videofrozen for 1.5 s, to orientate the partici-pants to the situation.

Procedure Each participant individuallycompleted the test in the same indoorroom at the youth academy over a 6-week period at each measurement point(i.e., T2015; T2016; T2017). At the com-mencement of each test, the participantwas provided with a standardized in-struction (e.g., “In each scene you startto dribble from one of five positions.”;“A short freeze frame at the start of eachvideowill help you toorientate to the cur-rent game situation.”; “During the drib-bling your decision is based on passingthe ball to one of three targets that rep-resent the position of your teammate.”).Prior to each testing block, three famil-iarization video scenes were presentedfor participants to become accustomedto the procedures and the respective cat-egory situation. Participants were thenable to ask any questions they had inrelation to the video or response pro-cedures. Following the familiarizationvideos, the 12 testing videos were pre-sented. For each trial, participants werepositioned with the ball on a startingposition which represented the locationon the pitch. As the video commencedthe participants started to dribble withtheballwhilewatching the evolvinggamesituationonthe screen. Prior tooverstep-pingalimitinglineorthevideosceneend-ing the participant provided a responseby passing to one of three possible tar-gets (i.e., supporting teammates in thevideo, . Fig. 2). For each trial, the firstauthor recorded the target (i.e., decision)the participants passed the ball to, whilea research assistant placed a ball on thenext starting position.

Dependent measures The decision-making accuracy for each video situa-tionwas assessed using a coding criterionof one for a correct decision (i.e., passingor shooting to the best option identifiedby the expert panel) and zero for anincorrect decision (i.e., passing or shoot-ing to an option not rated the best).

Only one correct option was determinedby the expert panel for each scenario.Any trial where the participant misseda target or did not decide before thescene ended was also coded as a zero.Subsequently, participants’ decisionsfrom 48 testing video situations wereused for calculating offensive (SCoff; i.e.,mean value of 24 corresponding scenes)and build-up (SCbu; i.e., mean value of24 corresponding scenes) decision accu-racy scores. Both scores (i.e., SCoff andSCbu) were converted to a percentage ofcorrect decisions. As the primary aimof the study was to determine decision-making accuracy, we did not calculate orassess speed of decision-making withinthe study.

Statistical analyses

Data were analyzed using SPSS ver-sion 24. Diagnostics’ instrumental relia-bility (Objective I) was examined usinga split-half procedure (odd-even-methodcorrected by Spearman–Brown formula;Lienert & Raatz, 2011) for players’ firstassessment results.

An examination of distributionalproperties revealed the decision-makingaccuracy scores SCoff and SCbu were notnormally distributed. A Mann–Whit-ney-U-test (as one-tailedhypothesis)wasused to determine differences in SCoff

and SCbu between age groups, playingstatus and future youth national teamstatus (Objective II, H1–H3). To ex-plore mean differences between playingpositions, nonparametric Kruskal–Wal-lis tests were conducted with post hocpairwise comparisons (Objective III).

As thenumberofmeasurementpointsmay influence decision-making scoresand consequently the results in regards tothe diagnostics’ validity (due to memo-rableorhabituationeffects), apriori anal-ysiswas conducted to determinewhetherthe number of measurement points cor-related with the independent variablesplaying status, age group, future youthnational team status, and playing posi-tion. The variables age group (rs= 0.18;p< 0.05) and playing position (rs= 0.22;p< 0.01) correlated with the number ofmeasurement points. Therefore, in a sec-ond step and as an additional analysis for

German Journal of Exercise and Sport Research 1 · 2021 105

Main Article

Fig. 18 Example of amidfield scene usedwithin the first-person perspec-tive video-based decision-making diagnostic instrument

Fig. 28 The structure of the first-personperspective video-baseddecision-making diagnostic instrument experimental set up

the effects of age groups andplayingposi-tion, the number of measurement pointsserved as a manifest covariate. This wasto assess whether the number of mea-surementpoints influencedtheresultpat-tern and statistical decisions with respectto age groups. As there is no statisti-cal procedure controlling for a covariatein nonnormal distributional properties,ANCOVAs (analyses of covariance) wereconducted in addition to Mann–Whit-ney-U-tests (post hoc analysis) to assessfor possible influences of the number ofmeasurement points on the results.

Effect sizes ω and Φ for nonparamet-ric tests were calculated and classified inaccordance to Cohen (1992). To deter-mine the size of a possible populationeffect (within objective II and III), sensi-tivity was calculated by post hoc poweranalyses using G*Power version 3.1.9.7.G*power provides sensitivity calcula-tions for mean comparisons betweentwo groups only based on Cohen’s d.The analyses determined the sensitivityfor objective II (α= 0.05, 1– β= 0.85,one-tailed), whereas H1 ranged from0.56≤ d≤ 0.60, with H2 equal to d= 0.47,and H3 equal to d= 0.77. Regarding ob-jective III, the power analyses (α= 0.05,1– β= 0.85, two-tailed) provided a rangeof 0.57≤ d≤ 0.76. Therefore, mediumeffect sizes could be detected within thepresent study. As the present study uti-lized Φ as an effect size, this correspondsto a range of Φ between 0.30 and 0.50.Due to the fact that only three measure-ment points are available for a very smallnumber of players, a longitudinal studywas omitted.

As H2 and H3 investigated a groupprediction, additional stepwise logisticregressions were conducted. The overallmodel fit was examined with the like-lihood ratio chi-square (X2) test. TheSCoff and SCbu regression coefficients aswell as the odds ratios eβ and their 95%confidence intervalswere calculatedwithreference to the respective selection cri-teria (i.e., “reserve player” for H2; “non-selected” for H3). Finally, individual se-lection probabilities were determined forplaying status and future youth nationalteam status on the basis of a player’s SCoff

and SCbu.

Results

Instrumental reliability demonstratedsatisfactory values for build-up scenes(r= 0.72), and lower for offensive scenes(r= 0.56; Objective I). . Table 2 presentsan overview of the decision-makingscores relative to age (H1) playing status(H2) and future youth national team sta-

tus (H3; Objective II). Mann–Whitney-U-tests identified significant differences(each p< 0.01) in decision-making com-petence between U17 and U16 (H1a;U17>U16 in SCbu: Z= 2.35, Φ=0.24and in SCoff: Z= 3.82, Φ=0.39), aswell as between U19 and U16 players(H1b; U19>U16, SCbu: Φ=0.41 andSCoff: Φ= 0.46). However, no significantdifferences were found between the com-parison of U19 and U17 players (H1c;U17>U19 in SCbu: Z= 0.04, p= 0.48 andU19>U17 in SCoff: Z= 0.69, p= 0.244).In relation to playing status (H2) first-team regular players performed signif-icantly better than reserve players inboth build-up and offensive situations,with low to moderate effect sizes (SCbu:Z= 2.57, Φ=0.22, p< 0.01 and SCoff:Z= 1.69, Φ=0.14, p< 0.05). With re-spect to the stepwise logistic regression,only build-up situations showed signif-icance and therefore remained in themodel (χ2(1)= 8.00, p< 0.01). The oddsratios eβ from the logistic regression

106 German Journal of Exercise and Sport Research 1 · 2021

Table 2 Descriptive and inferential statistics for criterion-related validitywith regards to age group (Objective II, H1), playing status (Objective II, H2),future youth national teamplayers (Objective II, H3), and the explorative analysis of playing position (Objective III)Variables Decision

accuracyscore (%)

Descriptivestatistics(M±SD)

Effect size

Post hoc analysis/Mann–Whitney U-test Kruskal–Wallistest

Age group U16 U17 U19 U16 vs. U17 U16 vs. U19 U17 vs. U19

(N= 41) (N= 55) (N= 44)

Φ

SCbu 66± 14 73± 14 71± 18 0.24** 0.41** 0.00

SCoff 63± 11 74± 13 75± 14 0.39** 0.46** 0.00

Playing status FTRP RP FTRP vs. RP

(N= 70) (N= 68)

Φ

SCbu 74± 13 67± 17 0.22**

SCoff 73± 13 69± 15 0.14*

Youth national team Selected Nonselected Selected vs.nonselected(N= 16) (N= 70)

Φ

SCbu 73± 11 65± 16 0.20*

SCoff 69± 11 66± 13 0.08

Playing positionaDF MF FW DF vs. MF DF vs. FW MF vs. FW

(N= 55) (N= 61) (N= 24)

Φ ω

SCbu 71± 16 72± 13 63± 17 0.00 0.21 0.25* 0.44#

SCoff 70± 16 73± 13 69± 9 0.07 0.07 0.15 0.15

FTRP First team regular player, RP Reserve player, DF Defenders, MF Midfielders, FW Forwards, SCbu Decision accuracy score for build-up, SCoff Decisionaccuracy score for offensive**p< 0.01, *p< 0.05; # p< 0.10aPlaying position correlated significantly with the number of measurement points. However, ANCOVAs (analysis of covariance) with the main factor playingposition and the covariate number of measurement points demonstrated no changes in the statistical decisions regarding the multiple group comparison.Thus, the results pattern for playing position is independent of the number of measurement points

model indicated that a one standarddeviation (SD= 0.153) increase in build-up score, improved the chance of beinga first-team regular player by a factor of1.65 (= (eβ)0.153 = 26.690.153).

With respect to the prognostic va-lidity (H3), Mann–Whitney-U-testdemonstrated that selected playershave a higher decision-making accu-racy than nonselected players in build-up (Z= 1.82, Φ= 0.20, p< 0.05) and of-fensive (Z= 0.78, p= 0.217) situations;however, only build-up situations pro-vided a significant difference. Analyzingthe prediction of future youth nationalteam status with regard to the logisticregression models led to the same re-sult as for the playing status variable.Only build-up score demonstrated sig-nificance and therefore remained in themodel (χ2(1)= 4.15, p< 0.05). A onestandard deviation increases in build-

up score improved the chance of beingselected for a future youth national teamby a factor of 1.89 (= (eβ)0.153 = 64.200.153).

Regarding playing position (Objec-tive III), the results for the differentsubsamples (i.e. DF, MF, FW) in bothbuild-up and offensive decision categoryranged in mean from 69–72% with onlyone exception (i.e., the result of 63%from FW deviated in their nonposition-specific build-up category compared toother playing positions distinctly). De-spite these facts, the Kruskal–Wallis testsdid not reveal significant differences inany decision-making competence (SCbu:H(2)= 5.26, p= 0.072; SCoff: H(2)= 1.81,p= 0.404). In terms of possible influ-ences caused by correlations betweenthe manifest variable, the number ofmeasurement points, and the indepen-dent variable playing position, ANCOVAdemonstrated no changes in the statis-

tical decisions regarding the multiplegroup comparisons of playing position.

Discussion

The aim of this study was to developa valid video-based decision-makingdiagnostic instrument which presentedfootage from a first-person perspectiveand incorporated a soccer-specific mo-tor response. A further strength of thestudy was the video stimuli were notlimited to one game situation, but ratherdifferent situations (i.e., build-up and of-fensive decisions). While there were nodifferences between playing positions,the study does provide a foundationfor future research. In particular, thediagnostic instrument was developed asameasure to discriminate decision-mak-ing competence within a group of youthhigh-performance level players. Results

German Journal of Exercise and Sport Research 1 · 2021 107

Main Article

showed thediagnostic instrument, whichincluded realistic soccer video-scenes incombination with a soccer-specific mo-tor response, is a suitable instrumentfor discriminating playing status andpartially for age (U19 and U17>U16).Further, it provides a more representa-tive assessment compared to previousstudies where the diagnostic instrumentwas limited by a lack of a soccer-specificresponse (e.g., Bennett, Novak, Pluss,Coutts, & Fransen, 2019). An additionalaim was to examine the predictive valueof the new video-based decision-makingdiagnostic instrument, with the findingsindicating that in parts of the study (i.e.,in build-up categories) future youth na-tional team players perform better in thedecision-making diagnostic instrument.

Using reliable diagnostic instrumentsis fundamental to scientific work. Theimplications associated with using a di-agnostic instrument with a lack of, orunknown, reliability is whether the par-ticipant performance differences are dueto random test error or actual per-formance changes of the participants(Gadotti, Vieira, & Magee, 2006). Incomparison to theassessmentofmanifestvariables (e.g., time, height, distance),the measurement of latent constructssuch as decision-making competenceposes a much larger challenge. Whileresearchers have highlighted the limitedreporting of reliability for new diagnos-tic instruments (Hadlow et al., 2018;Schweizer, Furley, Rost, & Barth, 2020),a key aim of the current study was tomeasure the reliability of the latent con-struct decision-making (Objective I).Here, an adequate level of reliabilityfor the build-up situations was found;however, the lower reliability for offen-sive situations has to be considered asa larger limitation. A potential reasonfor the lower reliability of the offensivesituations could be due to the restric-tion of range phenomenon within thehigh-performance level group (i.e., ho-mogeneous expert samples) the studywas conducted (Schweizer et al., 2020).Despite this limitation, the current studydoes provide a method for establishingthe reliability of the offensive decision-making situations. In addition, futurestudies may improve the reliability of

these situations by exchanging currentlower reliability clips with newly devel-oped higher reliability clips. Also addingmore items to the diagnostic instrumentcould be an approach for improvingthe reliability and even reduce the needfor large sample sizes (Schweizer et al.,2020).

This study advances current sport-based decision-making literature bydemonstrating the ability to developa decision-making diagnostic instru-ment in which the process of visualperception of in-game video-scenes (i.e.,first person perspective) is more chal-lenging by an additional soccer-specificmotor action (i.e., dribbling) and re-sponse (i.e. passing). Traditional video-based instrumentswhichassess decision-making performance generally presentfootage from a broadcast (i.e., televisionbroadcast) or third person perspectiveassociated with nonsoccer-specific de-cision response (i.e., written, verbal orbutton response). In this context, Mann,Farrow, Shuttleworth, Hopwood, andMacMahon (2009) indicated using a di-agnostic instrument from first person-perspective is more realistic. However,it was noted this may be more difficultfromadecision-making perspective thanusing footage from an aerial perspective,and therefore may provide more robustdiagnostic instruments. Furthermore, ithas been suggested the lack of sport-spe-cific responses may limit the correlationbetween video-based tests and actualin-game decision-making performance(e.g., van Maarseveen et al., 2016). Thisis further supported by Travassos et al.(2013) who indicated that perceptual-cognitive assessments need to considerthe task representativeness when de-veloping sport-specific diagnostic in-struments. By developing assessments,which consider representative task con-straints, such as presenting decision-making situations from a first-personperspective, combined with a sport-specific motor response, enhances thevalidity of the assessment measure.

While the current studydemonstratedthe ability to differentiate U16 and olderadolescent athletes (i.e., U17 and U19,objective II), developing a decision-making diagnostic instrument sensitive

enough to differentiate performance ofhigh-performing late adolescent athletesrequires further investigation. A possi-ble explanation as to why there was nosignificant difference between the U17and U19 players (H1c) could be due toat least two possible reasons. First, fromthe perspective of the U19 team, thebest players from the squad had alreadyplayed for the senior professional team(i.e., in either the first or reserve/secondteam) and therefore were not able toparticipate in the investigation. In rela-tion to the U17 group, two of the threeU17 cohort squads in this study werevery successful and participated in thefinals of the German U17 championship(i.e., the highest level of competition atthis age). Therefore, it seems feasiblethat despite the age difference with re-spect to U19 group, both groups were ofa very similar performance level (limit-ing the studies’ internal reliability withregard to discriminating between agegroups). Second, this finding supportsother studies to investigate age-relatedsoccer-specific performance skill differ-ences. For example, Huijgen, Elferink-Gemser, Post, and Visscher (2010) high-lighted there were no improvementsfrom ages 16 to 19 in dribbling andsprinting, which the authors attributedto the end of puberty. Furthermore,Beavan et al. (2019) revealed in a batteryof cognitive function tests (e.g., ViennaTest System, which measures inhibitionand cognitive flexibility) that there wereno differences between U19 and U17 agegroups. At this stage, further researchis needed to explore whether there arepeak developmental phases for cognitivefactors such as decision-making.

A strength of this study is that thesample consisted of participants fromone of the most successful youth socceracademies in Germany, thus, result-ing in a very homogeneous high-levelgroup. This is not a common approachused by researchers investigating soccer-specific decision-making skills as theyhave generally conducted cross-sectionalstudies comparing and identifying skilldifferences between groups of high-performance and low-performance levelparticipants (i.e., sub-elite; intermediate;novice participants; Diaz et al., 2011;

108 German Journal of Exercise and Sport Research 1 · 2021

Fig. 39 Selectionprobabilities ob-tained from the lo-gistic regressionfor reaching firstteam regular player(FTRP) and for draft-ing youth nationalteam (YNT) status

Roca et al., 2012; Scharfen & Memmert,2019). Therefore, while the findings ofthe current study indicated only a smallsignificant effect size for discriminat-ing playing status (H2), the possiblereason for this may be attributed to thehigh sensitivity when comparing athleteswithin a homogeneous high-level group.This supports previous findings by vanMaarseveen et al. (2016) who also failedto detect significant differences in per-ceptual-cognitive skills between talentedfemale soccer players of a comparablehighly skilled performance level (i.e.,national soccer talent team). Never-theless, the logistic regression modelsdemonstrated that players with gooddecision accuracy on the diagnostic in-strument will have a higher probabilityof being a first team regular player (SCbu:Odds Ratio= 1.65; . Fig. 3). Therefore,the results provide initial evidence tosuggest the new diagnostic instrumentcan potentially identify more skilledindividuals within a high-performancelevel youth group.

Further, in comparison to previousinvestigations which only assessed per-formance using a cross-sectional design(e.g., Höner, 2005), the current investi-gation implemented a prospective designto allow for a greater understanding ofpotential future soccer success (H3). Thelogistic regression model for SCbu indi-cated that players who performedwell onthe assessment improved the chance ofbeing selected (i.e., selected to participateinaGermanyouthnational teamtraining

course) by a factor of 1.89. The probabil-ity curve (. Fig. 3) increasedmore rapidlyfor higher SCbu values and therefore re-sults proved the diagnostics’ prognosticrelevance, as participants who performbetter on the assessment are also morelikely to be selected. This finding sup-ports other video-based research whichhasdemonstrated theability topredict se-lectionintohighperformingyouthsoccerprograms (O’Connor et al., 2016). Thisindicates the possibility for a video-baseddiagnostic instrument with a soccer-spe-cific motor response to be used withinthe talent identification process to assistwith the assessment of players’ decision-making performance. The nonparamet-ric analysis demonstrated a small signifi-cant effect size for SCbu and no significantdifference for SCoff between selected andnonselected future youth national teamplayers. However, this finding may havebeen limited by the relatively low num-ber of participants selected into nationalyouth programs (N = 16). Therefore, fu-ture research should consider increasingthe number of national youth players inthe sample to fully understand the po-tential decision-making skill differencesbetween these high performing players(i.e., selected; nonselected for youth na-tional team; Ackerman, 2014).

The present study extended the cur-rentresearchknowledgerelateddecision-making assessments by developing a di-agnostic instrument which incorporateddecision situations from defense, mid-field, and forward playing positions (Ob-

jective III). However, while assump-tions were based on previous investiga-tions which have demonstrated differ-ences based on playing positions (e.g.,Höner, 2005), thecurrentfindingsdidnotreveal a significant difference. Despitethis finding, the novel approach used inthe study (i.e., provide the players video-scenes from different playing positionson the pitch) may enable more detailedcomparisons between playing positionsin future. Hence, research should divideplaying positions in a stricter manner,for instance when considering the posi-tion of a forward, there may be specificdifferences between central forwards andwide forwards (i.e., wingers).

As this is one of the first studies todevelop adecision-makingdiagnostic in-strumentwhich linksvisualperceptionoffirst-person video scenes and soccer-spe-cific motor responses in the assessment,the findings should be considered withrespect to several limitations. First, theexecution-time of decision-making wasnot measured directly. Decision-makingwas only restricted by the length of thevideo scenes and a marked limiting line.Currently, there are different opinionsabout the importance of execution timewith respect to decision-making. Thereseems to be the belief that experiencedplayers make quicker and more accuratedecisions compared to less skilled play-ers (Vaeyens, Lenoir,Williams,Mazyn, &Philippaerts, 2007). Nevertheless, a fur-ther development of the diagnostic in-strument could include themeasurementof execution-time to determine whetherhighly skilled players decide faster thanless skilled players or whether they waiteven longer for the perfectmoment to ex-ecute the appropriate response. Second,the current sample was limited by the re-strictions of testing within a professionalsporting club environment. As such, sev-eral high-performing U19 athletes werenot included in the sample due to pro-fessional club commitments. As a result,it may be possible that the lack of signif-icant differences between U17 and U19players in the current sample may notfully describe the potential age-relateddifferences between these two groups.Therefore, future investigations shouldaim to sample all athletes from within an

German Journal of Exercise and Sport Research 1 · 2021 109

Main Article

age group to ensure a true representationof the performance group. Finally, sim-ilar to other studies (e.g., Höner, 2005),this study limited the number of options(i.e., three option per situation) and thesport-specific motor response (i.e., play-ers only had the options to pass or shoot).Future studies should also consider thedevelopment of diagnostic instrumentswhichincludeotherdecision-makingop-tions, such as dribbling. While observa-tional studies (e.g., recording decision-making skills in small-sided games) suchas Romeas, Guldner, and Faubert (2016)have attempted to assess the whole spec-trum of decision-making alternatives, itis a challenge developing such situationswithin a video stimuli. A possible alter-native maybe developing diagnostic in-struments that providemore graduationsof decisions (i.e., best option; second bestoption; etc.), rather than current binaryforms (e.g., Bennett et al., 2019). Further,off-the-balldecision-making indefensivegame situations such as positioning closeor far from the ball, which has not beenassessed in the literature, should also beconsidered in future research.

Conclusion

Sport-based decision-making is a cogni-tive process in which athletes use theirknowledge about a (current) situation toselect an appropriate decision based ontheir perceived ability to execute a sport-specific motor skill response (Causer &Ford, 2014). While decision-making isdefined by the ability to perceive appro-priate stimuli and execute a sport-specificskill response, traditional video-basedinstruments assess decision-making us-ing nonsport-specific responses such asverbal, written, or button responses (e.g.,Roca et al., 2012). This study addressedthis limitation by developing a validfirst-person perspective video-baseddiagnostic instrument which requiresa soccer-specific motor response (i.e.,participants dribble with the ball whilewatching a video stimulus and then exe-cute their decision by passing to a playerin the video). The results indicate thevideo-based decision-making diagnosticinstrument can discriminate decision-making competence within a high-per-

formance youth group, in particular forplaying status and to some extent age.Although the reliability of the diagnosticinstrument is satisfied for the build-upcategory, future research should aim forhigher reliability for all measured vari-ables (e.g., offensive decision), by usingthe guidelines proposed, for example,by Schweizer et al. (2020). Further, theresults demonstrate the predictive valueof the new video-based decision-makingdiagnostic instrument, especially thebuild-up situation for national youthteam selection. However, future investi-gations with a longer prognostic period(e.g., up to senior level) is of interest tostrengthen the results. With respect toplaying positions, no differences werefound between positions; however, fur-ther examination of playing position ondecision-making skills is warranted (e.g.,aremidfielder better decision-makers re-gardless on which area of the pitch thedecision-making skill is required?).

Considering these results, this studyprovides initial evidence to suggest a soc-cer-specific video-based diagnostic in-strumentcanbeusedwithintalent identi-ficationprocesses toassistwiththeassess-ment of players’ decision-making perfor-mance. Finally, future decision-makingassessment studies can use similar proce-dures to those employed in the current in-vestigation to develop a valid and reliablevideo-based decision-making diagnosticinstrument in other sports/domains. Inaddition, further exploration of the inte-gration of motor specific responses withthe video footage to create an even morerealistic diagnostic instrument may bepossible. One approach would be to usevirtual reality technology to allow bodyparts (e.g., foot) interactingwiththestim-ulus material or to measure a player’s insitu task constraints.

Corresponding address

© MarioHeilemann

DennisMurrInstitute of Sports Science,Eberhard Karls UniversityTübingen, Germanydennis.murr@uni-tuebingen.de

Funding. Open Access funding enabled and orga-nized by Projekt DEAL.

Compliance with ethicalguidelines

Conflict of interest. D.Murr, P. Larkin andO. Hönerdeclare that theyhave no competing interests.

All procedures performed in studies involvinghumanparticipantswere in accordancewith the ethical stan-dards of the 1964Helsinki declaration and its lateramendments or comparable ethical standards. Theethics department of the Faculty of Economics andSocial Sciences at theUniversity of Tübingen and theyouth academyof the professional soccer club ap-proved the implementation of this study. All playersand legal guardians (i.e., parents) provided informedconsent prior to participation in the study.

Open Access. This article is licensedunder a CreativeCommonsAttribution 4.0 International License,whichpermits use, sharing, adaptation, distribution and re-production in anymediumor format, as long as yougive appropriate credit to the original author(s) andthe source, provide a link to the Creative Commons li-cence, and indicate if changesweremade. The imagesor other third partymaterial in this article are includedin the article’s Creative Commons licence, unless in-dicatedotherwise in a credit line to thematerial. Ifmaterial is not included in the article’s Creative Com-mons licence and your intendeduse is not permittedby statutory regulation or exceeds the permitteduse,youwill need toobtain permissiondirectly from thecopyright holder. To viewa copyof this licence, visithttp://creativecommons.org/licenses/by/4.0/.

References

Ackerman, P. L. (2014). Nonsense, common sense,and science of expert performance: talent andindividual differences. Intelligence, 45, 6–17.https://doi.org/10.1016/j.intell.2013.04.009.

Beavan, A. F., Spielmann, J., Mayer, J., Skorski, S.,Meyer, T., & Fransen, J. (2019). Age-relateddifferences in executive functions within high-level youth soccer players. Brazilian Journal ofMotor Behavior,13(2), 64–75.

Bennett, K. J.M., Novak, A. R., Pluss, M.A., Coutts, A. J.,& Fransen, J. (2019). Assessing the validity ofa video-based decision-making assessment fortalent identification in youth soccer. Journal ofScience and Medicine in Sport, 22(6), 729–734.https://doi.org/10.1016/j.jsams.2018.12.011.

Bergkamp, T. L. G., Niessen,A. S.M., denHartigh, R. J. R.,Frencken,W.G. P., &Meijer, R. R. (2019). Method-ological issues in soccer talent identificationresearch. Sports Medicine. https://doi.org/10.1007/s40279-019-01113-w.

Bonney, N., Berry, J., Ball, K., & Larkin, P. (2019).Australian football skill-based assessments: aproposed model for future research. Frontiersin Psychology, 10, 429–429. https://doi.org/10.3389/fpsyg.2019.00429.

Boone, J., Vaeyens, R., Steyaert, A., Vanden Bossche,L., & Bourgois, J. (2012). Physical fitness ofelite Belgian soccer players by player position.Journal of Strength and Conditioning Research,

110 German Journal of Exercise and Sport Research 1 · 2021

26(8), 2051–2057. https://doi.org/10.1519/JSC.0b013e318239f84f.

Causer, J., & Ford, P. R. (2014). ‘Decisions, decisions,decisions’: transfer and specificity of decision-making skill between sports. CognitiveProcessing, 15(3), 385–389. https://doi.org/10.1007/s10339-014-0598-0.

Cohen, J. (1992). Statistical power analysis. CurrentDirections in Psychological Science,1(3), 98–101.

Deprez,D.,Fransen,J.,Boone,J.,Lenoir,M.,Philippaerts,R., & Vaeyens, R. (2015). Characteristics of high-level youth soccer players: variation by playingposition. Journal of Sports Sciences, 33(3),243–254. https://doi.org/10.1080/02640414.2014.934707.

Diaz, D. C.D., Gonzalez, V. S., Garcia, L. L., & Mitchell,S. (2011). Differences in decision-makingdevelopment between expert and noviceinvasion game players. Perceptual and motorskills,112(3), 871.

Dieze, S. (2015). Positionsspezifische Diagnostik derEntscheidungskompetenz im Fußball: Entwick-lung und erste Validierung einer Videotestbatterie[Position-specific diagnostics of decision-makingin soccer: developing and validation of a video-based test battery]. Tübingen: Eberhard KarlsUniversitätTübingen.

Frýbort, P., Kokštejn, J., Musálek, M., & Süss, V. (2016).Does physical loading affect the speed andaccuracy of tactical decision-making in elitejunior soccerplayers? Journal of Sports Science &Medicine,15(2), 320–326.

Gadotti, I., Vieira, E., & Magee, D. (2006). Importanceand clarification of measurement propertiesin rehabilitation. Brazilian Journal of PhysicalTherapy,10(2), 137–146.

Hadlow, S.M., Panchuk, D., Mann, D. L., Portus, M. R.,& Abernethy, B. (2018). Modified perceptualtraining insport: anewclassification framework.Journal of Science and Medicine in Sport, 21(9),950–958. https://doi.org/10.1016/j.jsams.2018.01.011.

Hagemann, N., Lotz, S., & Cañal-Bruland, R. (2008).Wahrnehmungs-Handlungs-Kopplung beimtaktischen Entscheidungstraining: Eine ex-ploratorische Studie [Perceptual-action-cou-pling in tactical decision-making training: anexplorative study]. E-Journal Bewegung undTraining,2, 17–27.

Höner, O. (2005). Entscheidungshandeln im SportspielFußball: Eine Analyse im Lichte der Rubikontheo-rie [Decision making in football—An analysis inthe context of the Rubicon theory]. Schorndorf:Hofmann.

Höner, O., Larkin, P., Leber, P., & Feichtinger, P.(2020). Talentauswahl und -entwicklungim Sport. In J. Schüler, M. Wegner &H. Plessner (Eds.), Sportpsychologie: Grundlagenund Anwendung [Textbook of sports psychology:basics and application] (pp. 499–530). Berlin,Heidelberg,NewYork: Springer.

Huijgen, B. C.H., Elferink-Gemser, M. T., Post, W., &Visscher, C. (2010). Development of dribbling intalented youth soccer players aged12–19 years:a longitudinal study. Journal of Sports Sciences,28(7), 689–698. https://doi.org/10.1080/02640411003645679.

Johnston, K., Wattie, N., Schorer, J., & Baker, J. (2018).Talent identification in sport: a systematicreview. Sports Medicine, 48(1), 97–109. https://doi.org/10.1007/s40279-017-0803-2.

Larkin, P., O’Connor, D., &Williams, A.M. (2016). Doesgrit influence sport-specific engagement andperceptual-cognitive expertise in elite youth

soccer? Journal of Applied Sport Psychology,28(2), 129–138. https://doi.org/10.1080/10413200.2015.1085922.

Lex, H., Essig, K., Knoblauch, A., & Schack, T.(2015). Cognitive representations and cognitiveprocessing of team-specific tactics in soccer.PLoS One, 10(2), e118219. https://doi.org/10.1371/journal.pone.0118219.

Lienert, G. A., & Raatz, U. (2011). Testaufbauund Testanalyse [Setup and analysis of tests].Weinheim: Beltz.

Lorains, M., Ball, K., &MacMahon, C. (2013). Expertisedifferences in a video decision-making task:speed influencesonperformance. Psychology ofSport and Exercise, 14(2), 293–297. https://doi.org/10.1016/j.psychsport.2012.11.004.

van Maarseveen, M. J. J., Oudejans, R. R.D., Mann,D. L., & Savelsbergh, G. J. P. (2016). Perceptual-cognitive skill and the in situ performanceof soccer players. The Quarterly Journal ofExperimental Psychology. https://doi.org/10.1080/17470218.2016.1255236.

Mann, D., Dehghansai, N., & Baker, J. (2017).Searching for the elusive gift: advances intalent identification in sport. Current Opinion inPsychology,16(SupplementC),128–133. https://doi.org/10.1016/j.copsyc.2017.04.016.

Mann, D., Farrow, D., Shuttleworth, R., Hopwood,M., & MacMahon, C. (2009). The influence ofviewing perspective on decision-making andvisual search behaviour in an invasive sport.International Journal of Sport Psychology, 40(4),546–564.

Murr, D., Feichtinger, P., Larkin, P., O’Connor, D.,& Höner, O. (2018). Psychological talentpredictors in youth soccer: a systematic reviewof the prognostic relevance of psychomotor,perceptual-cognitive and personality-relatedfactors. PLoS One, 13(10), e205337. https://doi.org/10.1371/journal.pone.0205337.

O’Connor, D., Larkin, P., & Williams, A.M. (2016).Talent identification and selection in elite youthfootball: anAustraliancontext. European Journalof Sport Science,16(7), 837–844. https://doi.org/10.1080/17461391.2016.1151945.

Pocock, C., Dicks, M., Thelwell, R. C., Chapman,M., & Barker, J. B. (2019). Using an imageryinterventiontotrainvisualexploratoryactivity ineliteacademyfootballplayers. Journal of AppliedSport Psychology,31(2), 218–234.

Rago, V., Pizzuto, F., & Raiola, G. (2017). Relationshipbetween intermittent endurance capacity andmatch performance according to the playingposition in sub-19 professional male footballplayers: preliminary results. Journal of PhysicalEducation and Sport,17(2), 688.

Roca, A., Williams, A.M., & Ford, P. R. (2012).Developmental activities and the acquisition ofsuperior anticipation and decision making insoccerplayers. Journal of Sports Sciences,30(15),1643–1652. https://doi.org/10.1080/02640414.2012.701761.

Romeas, T., Guldner, A., & Faubert, J. (2016). 3D-Multiple Object Tracking training task improvespassing decision-making accuracy in soccerplayers. Psychology of Sport & Exercise,22, 1–9.

Ruiz Pérez, L.M., Palomo Nieto, M., García Coll, V.,NaviaManzano, J. A., Miñano Espín, J., & Psotta,R. (2014). Self-perceptions of decision makingcompetence in Spanish football players. ActaGymnica,44(2), 77–83.

Scharfen,H. E.,&Memmert,D. (2019). Measurementofcognitive functions in experts andelite athletes:

a meta-analytic review. Applied CognitivePsychology,33(5), 843–860.

Schweizer, G., Furley, P., Rost, N., & Barth, K.(2020). Reliable measurement in sportpsychology: The case of performance outcomemeasures. Psychology of Sport and Exercise, 48,101663. https://doi.org/10.1016/j.psychsport.2020.101663.

Travassos, B., Araujo, D., Davids, K., O’hara, K., Leitão,J., & Cortinhas, A. (2013). Expertise effectson decision-making in sport are constrainedby requisite response behaviours—A meta-analysis. Psychology of Sport and Exercise, 14(2),211–219.

Vaeyens, R., Lenoir, M., Williams, A.M., Mazyn, L.,& Philippaerts, R.M. (2007). The effects oftask constraints on visual search behavior anddecision-making skill in youth soccer players.Journal of Sport and Exercise Psychology, 29(2),147–169.

Ward, P., & Williams, A.M. (2003). Perceptual andcognitive skill development in soccer: themultidimensionalnatureofexpertperformance.Journal of Sport & Exercise Psychology, 25(1),93–111.

Ward, P., Ericsson, K. A., & Williams, A.M. (2013).Complex perceptual-cognitive expertise ina simulated task environment. Journal ofCognitive Engineering and DecisionMaking,7(3),231–254.

Williams, A.M., & Reilly, T. (2000). Talent identificationand development in soccer. Journal of SportsSciences,18(9), 657–667.

Wilson, R. S., James, R. S., David, G., Hermann, E.,Morgan, O. J., Niehaus, A. C., et al. (2016).Multivariate analyses of individual variation insoccer skill as a tool for talent identification anddevelopment: utilising evolutionary theory insportsscience. Journalof Sports Sciences. https://doi.org/10.1080/02640414.2016.1151544.

Woods, C. T., Raynor, A. J., Bruce, L., & McDonald,Z. (2016). Discriminating talent-identifiedjunior Australian football players using a videodecision-makingtask. Journal of Sports Sciences,34(4), 342–347. https://doi.org/10.1080/02640414.2015.1053512.

German Journal of Exercise and Sport Research 1 · 2021 111