Full Terms & Conditions of access and use can be found athttps://www.tandfonline.com/action/journalInformation?journalCode=uqst20

Quest

ISSN: 0033-6297 (Print) 1543-2750 (Online) Journal homepage: https://www.tandfonline.com/loi/uqst20

Compromising Talent: Issues in Identifying andSelecting Talent in Sport

Joseph Baker, Jörg Schorer & Nick Wattie

To cite this article: Joseph Baker, Jörg Schorer & Nick Wattie (2018) CompromisingTalent: Issues in Identifying and Selecting Talent in Sport, Quest, 70:1, 48-63, DOI:10.1080/00336297.2017.1333438

To link to this article: https://doi.org/10.1080/00336297.2017.1333438

Published online: 20 Jul 2017.

Submit your article to this journal

Article views: 1375

View Crossmark data

Citing articles: 8 View citing articles

Compromising Talent: Issues in Identifying and SelectingTalent in SportJoseph Baker a, Jörg Schorer b, and Nick Wattie c

aKinesiology and Health Science, York University, Toronto, Ontario, Canada; bInstitute of Sport Science, Carlvon Ossietzky Universitat Oldenburg, Oldenburg, Germany; cFaculty of Health Sciences, University of OntarioInstitute of Technology, Oshawa, Ontario, Canada

ABSTRACTThe past few decades have seen a significant change in the deliveryof sport and in trends related to athlete development. However, thenotion of talent continues to play a critical role in most athletedevelopment models. In this brief review, we highlight concernswith the notion of talent and how it is conceptualized in highperformance sport systems. These include: the assumption that talentis a fixed capacity that can be identified early, the influence of talentbeliefs on athlete development, the different levels of risk for talentselection decisions, biases evident in approaches to athlete selection,the inadequacy of current statistical approaches, the problems withusing current performance to predict future outcomes, and howshort-term priorities and competition between sports for talentedathletes undermine the overall efficiency of athlete developmentsystems. These concerns form the basis for more focused discussionof avenues for future work in this field.

KEYWORDSAthlete; coaching; sportpsychology

The past few decades have seen a significant change in the delivery of sport and in trendsrelated to athlete development. One example is the increasing pervasiveness of “eliteacademies” that focus on the development of high performance athletes. At theseacademies, parents pay considerable funds so their child can develop in a highlytargeted, athlete-centered environment generally focused on early specialization(Baker, Cobley, & Fraser-Thomas, 2009), where sport involvement is focused on theperformance demands of a single sport with very little participation outside this activity(for more on these types of academies, see Campbell & Parcels, 2013). Although thereare culture- and sport-specific differences (see Bridge & Toms, 2013), the increasedgeneral emphasis on early specialization, among other changes in high performancesport systems, resulted in increasing weight placed on early identification of youth withthe potential to succeed in high performance sport environments (Baker, Cobley,Schorer, & Wattie, 2017; Cobley, Schorer, & Baker, 2012).

Unquestionably, the ethical issues regarding this increased attention to early talent identifica-tion are considerable. Ethical issues include the psychological and behavioral implications oftelling someone they are talented (or not talented), the extent to which practices are inclusive or

CONTACT Joseph Baker bakerj@yorku.ca York University, Kinesiology and Health Science, 4700 Keele St, Toronto,Ontario, M3J1P3 Canada.Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/uqst.

QUEST2018, VOL. 70, NO. 1, 48–63https://doi.org/10.1080/00336297.2017.1333438

© 2018 National Association for Kinesiology in Higher Education (NAKHE)

exclusive (see Dries, 2013 for a detailed discussion), and the consequences of focusing on earlytalent selection for the promotion of physical activity and healthy lifestyles (for reviews, seeBailey & Toms, 2011; Kerr & Stirling, 2013). Despite these ethical issues, early identification andselection is the modus operandi of high performance sport. Notions of talent (or the relatedconcept of giftedness; see Gagné, 2013) play a critical role in most athlete development models,though it remains an elusive quality. In 1998, Howe, Davidson, and Sloboda proposed adefinition where talent (a) originates in genetically transmitted structures and is therefore atleast partly innate, (b) involves some advanced indications of future potential, (c) shows up inadvanced indications that could be used to provide a basis for identifying who is most likely tosucceed, (d) is only found in aminority of people, and (e) is relatively domain-specific. Althoughthe notion of talent was ultimately dismissed byHowe and colleagues (1998), much has changedin the nearly 2 decades since their seminal review. For instance, sport science researchersidentified specific genes that may serve as early precursors to future potential (e.g., Ahmetovet al., 2015; Papadimitriou et al., 2016). The gene for Collagen, Type V, Alpha 1 (COL5A1),which encodes for a specific collagen protein relevant to flexibility of ligaments and tendons,appears to determine an individual’s predisposition to Achilles’ tendon (Mokone, Schwellnus,Noakes, & Collins, 2006) or anterior cruciate ligament injury (Posthumus et al., 2009).Considering the robust relationship between time spent in practice and attainment (see Baker& Young, 2014 for a review), this gene could affect the amount and intensity of trainingindividuals in certain sports can perform throughout their development, and therefore reflect(directly or indirectly) a measure of future potential.

Given the continued dominance of talent conceptions in coaching discussions and sportscience domains, prudence requires the continual evaluation of the evidence for talent. In thisarticle, we highlight several contemporary concerns with the notion of talent and how it isconceptualized in high performance sport systems. These concerns are not meant to encom-pass the entirety of issues related to talent identification and development in sport, but to formthe basis for more focused discussion of these and other issues. Certainly, there are otherreviews and discussions about issues in the identification and selection of athletes in highperformance sport (e.g., Davids, Araújo, Vilar, Renshaw, & Pinder, 2013; Li, Wang, & Pyun,2014; Vaeyens, Lenoir, Williams, & Philippaerts, 2008)—some of which discuss similar issuesto the current review.However, we believe the continued attention to these issues is important,especially since they do not seem to have improved much despite prior discussion. In thisarticle, we positioned our interactions with key stakeholders in sport (e.g., coaches, athletes,administrators) to identify issues that seem to be most salient to these groups as they try tonavigate and improve talent selection.Moreover, we highlighted several novel issues especiallyrelevant to those working in high performance sport. In order to stimulate further discussion,we presented these as propositions and noted some of the key supporting evidence.

Proposition 1: Early talent selection assumes talent is a fixed capacity thatcan be identified early

Selection decisions generally occur throughout development, although the number of selectionsteps and when they occur are sport-specific.With increasing attention to providing the optimalenvironments for athlete development, many sport systems focused on the identification andselection of talent during early phases of development. In one of the most extreme examples of

QUEST 49

this phenomenon, in 2011, a Dutch soccer club signed an 18-month-old to a 10-year “symbolic”contract—gaining considerable worldwide media coverage.

Unfortunately, early focus on talent selection is fraught with limitations, most signifi-cantly (at least in the current context), the assumption that early indicators are validpredictors of future potential. This assumption suggests talent is a fixed capacity that (a)can be identified early and (b) does not change over developmental time. Generally, theevidence in sport (and other areas, see Howe, Davidson, & Sloboda, 1998) supportingearly indicators of talent is weak (e.g., Barreiros & Fonseca, 2012; Brouwers, DeBosscher, & Sotiriadou, 2012), purportedly because indicators of early performancesuccess (e.g., height, weight) have little connection to the factors determining success atthe adult level (i.e., the groups are largely homogenous on these qualities).

The second suggestion—talent is a fixed capacity that does not change over time—hasalso been challenged (e.g., Simonton, 1999). In his emergenic and epigenetic model oftalent, Simonton (1999) proposed a more valid explanation of talent as a multi-facetedquality that reflects the relative contribution of physical, physiological, cognitive, anddispositional traits that facilitate or impede the acquisition of expertise in a domain. Inthis model, these qualities emerge at different rates for different individuals over thecourse of their development. Attempts to identify these qualities early are thereforeproblematic because they do not account for these changes. Similar discussions concern-ing the dynamic nature of talent selection and athlete development have occurred in sport(see Abbott, Button, Pepping, & Collins, 2005; Davids, et al., 2013), thus emphasizing thedifficulty of early identification of these qualities.

Proposition 2: Beliefs about talent matter

Research over the last 3 decades (e.g., Dweck, Chiu, & Hong, 1995; Wulf & Lewthwaite,2009) reveals that individuals’ attitudes about talent affect their motivation, behavior, andperformance. Popularized as a “Growth Mindset” (see Dweck, 1999), this research sug-gests people generally have inherent ability beliefs or acquirable skill development beliefs(Wulf & Lewthwaite, 2009) about the origin of their capabilities. More specifically, thesebeliefs reflect whether a person believes their performance reflects innate and unchange-able qualities or is the product of experiences and therefore changeable.

Viewing talent as a gift, something that someone innately has or does not have, is problematicfor a number of reasons. First, it ignores the salient issues raised in Proposition 1, most notablythe notion that talent is a fixed capacity that can be identified early and remains stable over time.Second, viewing abilities as innate can influence both coach and athlete behavior in significantways. Individuals who view abilities as having innate origins often attribute someone else’sperformance, good or bad, to their innate level of ability. Conversely, those who believe abilitiesare “developable” often attribute other’s performance to their level of effort (Dweck, 1999). Thesebeliefs may also relate to whether practitioners have an exclusive perspective on talent (i.e., aselect few have innate talent), or whether they see talent (or potential) in everyone and adopt amore inclusive approach to managing talent (see Dries, 2013). This has significant implicationsfor the predictive decisionsmade by coaches and scouts.When a coach or scoutmakes a decisionabout who has talent or has the potential for further development, they are ultimately making aprediction about a range of future sport outcomes. Importantly, research suggests ascribingrigidly to one theory or way of thinking is associated with decreased accuracy in predictions

50 J. BAKER ET AL.

(Tetlock, 2005). Whether believing that talent is innate or developable, believing too stringentlycan result in both Type I and II errors in talent identification and development (believing thattalent is there/can be developed when it is not/cannot be and believing that talent is not there/cannot be developed, when in fact it is there/can be developed).

Furthermore, the beliefs of others, like a coach, can influence athletes once they have beenselected for a team (Dweck, 2003; Jourden, Bandura, & Banfield, 1991; Wulf & Lewthwaite,2009). This is important at the level of the athlete; learners who believe abilities are innate andultimately unchangeable are more likely to react to failure and negative feedback with decreasedeffort, persistence, and increased negative emotions (Dweck, 1999; Dweck et al., 1995). Suchbeliefs may also increase the risk of cultivating learned helplessness (i.e., self-handicappingbehavior: Ommundsen, 2001), or forms of complacency wherein athletes believe their talentassures them future success and subsequently decrease their effort and motivation to improve(i.e., the crown prince syndrome: see Dries, 2013 for a discussion). These responses are notconducive to learning andperformance, and as a result, can compromise talent identification anddevelopment. Conversely, those who believe talent to be predominantly developable are morelikely to respond to poor performances with increased effort and persistence. While thesemotivational and behavioral responses may seem more positive, over-persistence may also beamaladaptive response and itself a form of helpless behavior (Dweck, 1999); believing that talentcan be developedwith enough effort or practicemaynot be healthy or realistic. In short, believingin either perspective too rigidly can be damaging (for amore detailed discussion of this topic, seeWattie & Baker, 2017).

In summary, labeling high performance youth athletes as talented can create the expecta-tion that they will be successful at later stages in the athlete development pathway becausetheir talent is innate (natural). Research clearly indicates that hard work and high qualitytraining are important to becoming a high performance athlete (see Baker & Young, 2014).Moreover, creating unrealistic expectations can be damaging to young athletes given the lowcorrelation between success at one level of sport and success at the next higher level ofcompetition (Barreiros & Fonseca, 2012). Similarly, creating the expectation that increasedeffort and persistence will inevitably result in success may also place unfair expectations onathletes. Practice is clearly necessary, but it may not be sufficient: genes, resources, and luck arealso likely important (Baker & Horton, 2004).

Proposition 3: Talent selection decisions have different levels of risk

Those involved with athlete selection are faced with the difficult job of identifying potentialwithout objective or valid measures of its existence. While genetic factors (e.g., the genes forinjury risk noted earlier) may emerge at some later point to provide some predictive value,this is not likely to occur anytime soon (see Loland, 2015; Webborn et al., 2015). As a result,coaches who are required to stream athletes into different levels of competition typicallyinfer potential from current levels of performance (Vaeyens et al., 2008). This is problematicand results in a risk matrix such as the one shown in Figure 1.

Due to the limitations of using performance as a proxy for potential, practitioners are atrisk in several areas. Athletes in the white boxes of Figure 1 are not generally a problembecause those who do not have high enough performance will be removed from the system,while those with superior performance will stay in. Conversely, athletes in the light grey boxesrepresent moderate risk because (a) they ultimately represent “average” performance, which

QUEST 51

will not be good enough to attain the levels required for true “elite” performance (i.e., averagewill not cut it), and (b) they have the potential to clog the system by taking spots from thosewith higher potential. The highest levels of risk are the other dark grey boxes, since they eitherrepresent the highest levels of potential that may be lost if coaches focus too much onperformance as an indicator of value (boxes 7 and 8), or reflect those with initial high levelsof performance, but low long-term potential (box 3). This latter group also has the potential totake spots that would be better suited to athletes with higher potential for future performance.Importantly, these risks are buffered by the demand for spaces in the athlete selection system;when there are fewer spaces, these risks are magnified.

This matrix highlights the importance of considering talent selection decisions fromdifferent levels of risk. Coaches, administrators, and scouts need to determine the level ofrisk for each athlete being evaluated and, perhaps more importantly, the type of error withwhich they are most comfortable. If the goal of a talent development system is to keep anypotential talent within the system, Type II errors (i.e., false negatives) are more costly thanType I errors (false positives), since the latter would result in talented athletes removedfrom the system. However, sporting systems in most countries do not have the resourcesto accept high rates of either types of error, and so stakeholders need a clear (and honest)position regarding what level and type of error their system will accept.

Proposition 4: Talent identification is not done on a level playing field

Previous research on the development of high performance athletes also indicatedthe presence of systematic biases. For example, the use of annual age groups in youthsport has been found to promote relative age effects, where those born early in the

Figure 1. A risk matrix for talent identification decisions.

52 J. BAKER ET AL.

selection year (i.e., the relatively older) are more likely to be selected to sport teamsand reach elite levels of play than relatively younger peers born later in the selectionyear. Over the last decade, researchers identified participation and attainmentinequalities resulting from annual age grouping procedures across varying formsand levels of sport participation (see Wattie, Schorer, & Baker, 2015 for a detailedreview). Generally, youth born early (e.g., first 3 months) in the selection year (i.e.,the relatively older) have selection and attainment advantages over their relativelyyounger peers born later in the selection year. Baker and colleagues (2010) observedthat over 35% of players in two amateur developmental ice hockey leagues were bornin the first 3 months of the selection year (January–March), while less than 10% wereborn in the final 3 months (October–December). In rugby league, researchers docu-mented the representation of athletes in the first 3 months of the selection year to behigh (60%) among 13 to 15 year olds (Till et al., 2010). Over-representations ofrelatively older players have been consistently observed in a variety of sports (forreviews, see Cobley, Baker, Wattie, & McKenna, 2009; Musch & Grondin, 2001).Intriguingly, a recent meta-analysis by Cobley and colleagues (2009) noted thatalthough over-representations of relative older players exist at recreational levels ofsport, they are most pronounced at more competitive levels of participation. Theseeffects can also increase in size within sport systems that use 2- or more year agebands (Schorer, Wattie, & Baker, 2013). This was recently demonstrated using datafrom the soccer Under 17 World Cup (Steingröver, Wattie, Baker, Helsen, & Schorer,2017) as well as in German basketball (Steingröver, Wattie, Helsen, Baker, & Schorer,2017). For example, in German basketball, the under 16 age-group in the nationalyouth league contains players who are nearly 16 years, as well as players who justturned 14. As a result, the relative age effects here are even larger than previouscomparisons of within-year effects (Steingröver, Wattie, Helsen, et al., 2017).1

There is also evidence that access to financial resources and socioeconomic status(SES) may limit an individual’s likelihood of becoming a high performance athlete. Moregenerally, data suggest household income is significantly related to sport participation.For example, while 58% of Canadian children from households that earn less than$40,000 participate in sport, 85% of children from households that earn over $80,000participate in sport (Heritage Canada, 2013). While informative, aggregate data such asthese do not provide information about biases related to high performance sportparticipants. Prior research in this area (Beamish, 1990, 1992) highlighted the consis-tency of SES limitations to participation in high performance sport, but the stability ofthis effect in contemporary samples is unknown.

Proposition 5: Multivariate approaches to talent selection may beproblematic

Over the past few years, there has been an increasing focus on the use of complexmultivariate tests for the identification of talent (e.g., Elferink-Gemser, Visscher,Lemmink, & Mulder, 2004, 2007). These kind of inferential statistics are based oncomparisons of group means and variances on a series of dependent variables/tests.For example, in German handball, technical and motor tests, psychological question-naires, actual game performance, among others, are taken into account when making

QUEST 53

talent selection decisions (Schorer et al., 2012). In this approach, a number of skills areconsidered for each athlete, and the “best” are considered for future development.However, this approach may not be as simple as it seems. For example, the timeframeof those making decisions about athletes’ future development is often short (e.g., whowill be best for next season?), resulting in decisions made from a short-term perspec-tive (see also Proposition 7). Furthermore, if we consider Figure 2, in which all threeathletes would have the same summative score but different profiles of performance,other important nuances emerge. While Athlete 1 has all elements at roughly the samelevel, Athlete 2 shows an exceptional score in skill 3 and lower performances thanAthlete 1 in all other skills. Finally, Athlete 3 shows half of his skills above Athlete 1and half below them. The critical question is, therefore, which player should bedeveloped? While many would argue that Athlete 1 is the best choice because shehas the superior overall “package,” in certain cases (e.g., team activities), the athletewith a clear area of exceptional skill (i.e., Athlete 2) might be better—especially ifcertain skills can be substituted by other members of a team or by subtle changes inteam tactics (see “compensation phenomenon” as discussed by Vaeyens et al., 2008). Ifthese values formed the basis of quarterback selection in the National Football League(NFL), for example, with skill 3 reflecting throwing speed and accuracy, the choice isobvious. In our example, the overall sum of the skills would have been 30 points, butquite often, athletes who are exceptional in one skill have a lower summative oraverage score, and are therefore not evident from strictly statistical approaches.

The contrasting approach commonly used by coaches in the field often reflects thedecision to see a player as talented because they “have something special” (i.e., they havesome specific skill or ability that could become important with time). Collectively, weknow very little about the efficacy or limitations of coaches’ decision making about talent.While coaches can be biased with regard to these decisions (e.g., see earlier discussion ofthe relative age effect), their in-field “hypothesis-testing” should be more seriously con-sidered in scientific studies to determine the value and cost of the multivariate “overall”performer versus the univariate “outlier.”

Figure 2. Hypothetical patterns of performance for three athletes.

54 J. BAKER ET AL.

Proposition 6: Selecting talents requires predicting the future of sport

Identifying and selecting talent involves predictions about which athletes have the bestpotential for future success. Typically, this includes the analysis of current experts todetermine which pattern of skills will be necessary for future elite performance. While thisapproach is a good way to understand current elite performances, elite performance inlater years might look appreciably different. A simple example comes from sprinting andthe influence of Usain Bolt. Until relatively recently (i.e., the pre-Bolt era), talent selectionin sprinting largely focused on step frequency, and being tall was seen as a drawback. Asnoted in a recent article in The Telegraph, biomechanics specialist Dr. Ian Bezodis noted“ten years ago, most people would have said that being above average height is adisadvantage, but what Bolt has done has shown it’s not as simple as that” (Thomas,2016). Now, there is evidence that both step length and step frequency explain eliteperformance in the 100-m sprint (e.g., Salo, Bezodis, Batterham, & Kerwin, 2011), andcurrent talent identification and selection evolved to focus on both of these outcomes.

Effective talent selection, therefore, requires accurate prediction of how sportswill change inthe future in order to anticipate how the skills and capabilities underpinning successfulperformance will evolve and/or change between selection and demonstration of elite skill.Arguably, this is much more difficult than it would initially appear. Sports do not follow apredictable trajectory (see Figure 13.2 in Baker, Wattie, & Schorer, 2015). In a recent analysis,Baker and colleagues (2015) examined changes over time in Olympic record performance forathletic and aquatic events, noting the considerable degree of variability in improvement acrossdifferent sports over time. Presumably, the factors that drive these improvements in perfor-mance also change in a dynamic andunpredictableway, whichwould clearly limit the accuracyof any predictions about future performance. Moreover, as the timeframe of predictionincreases, accuracy of predictions would almost certainly decrease, which raises obviousconcerns about talent selection in childhood or early adolescence (e.g., do early selectiondecisions remove talent from the athlete selection pool that might emerge later?).

Proposition 7: Short-term priorities undermine talent selection anddevelopment

Athlete development is a long-term process and one of the challenges in traditional athletedevelopment systems is the difficulty balancing short-term and long-term goals. Thisconflict is nicely demonstrated in many developmental leagues and organizations. Forexample, the Canadian Hockey League (CHL) describes itself as “the world’s largestdevelopment hockey league.” And yet, the reality of the teams within the CHL is thatthey are privately owned, for-profit organizations, dependent on the revenue generated bythe teams. As such, like any professional team, these “developmental organizations” facethe pressures of depending on the marketability of their athletes, and the success of theirteams. This conflicting reality plays itself out in a number of other sports at the highschool level and the collegiate level in the United States. In such systems, coaches retaintheir employment based on their team’s short-term success: their win-loss record or howtheir team does in a major tournament. As a result, teams and organizations cannotalways make talent selection decisions in terms of what is best for future athletedevelopment.

QUEST 55

Our concern is that the conflict between short-term and long-term goals decreases theefficiency of the entire “talent system.” For example, talent identification largely relies onthe use of static measurements of anthropometrics and/or performance on discreetphysical tasks (e.g., sprint speed, strength tests, and performance on drills that rewardstrength and speed: see Pearson, Naughton, & Torode, 2006). During this process, coachesand scouts are not selecting based on talent, nor potential talent. They are selecting basedon performance, and who will perform best in a relatively short timeframe. Currentperformance is not necessarily an indication of future potential or “talent” (e.g., seework by Barreiros, Côté, & Fonseca, 2014; Barreiros & Fonseca, 2012), and these practicescreate contexts that can result in biases toward advanced growth and maturation thatultimately decrease the efficiency of athlete development. A very pertinent question to askis whether coaches and selectors at developmental levels of sport actually equate currentperformance with future potential. We suspect they do. If that is the case, then we shouldnot be surprised when our talent identification and development systems are inaccurateand inefficient (Barreiros & Fonseca, 2012; Koz, Fraser-Thomas, & Baker, 2012). It wouldbe more accurate to describe these activities as performance identification rather thentalent identification. While performance identification might serve short-term goals (i.e.,winning games), they might ultimately undermine talent identification and the long-termdevelopment of talented athletes.

Another issue with short-term priorities is that it creates a context that might not place thecurrent or long-term needs of athletes at the forefront. For example, how many injuries andburnouts occur because the short-term goals of an organization take priority over the long-termdevelopment of the athlete? This distinction is also at the center of broader issues around thecommoditization of developmental athletes as a means to generate wealth currently discussedaround rights of “amateur” athletes (see Charney Lawyers, 2016; Hruby, 2016). Do such systems,which are facing serious allegations of exploitation, really have the current and long-term needsof athletes as priorities?

The challenge, which is considerable, is to incentivize athlete development systems in sucha way that long-term outcomes are valued more than short-term rewards. Such a transforma-tion will likely require changes to policy, and will require economic and scientific evidencethat supports the benefits of shifts toward a long-term focus. Furthermore, if talent identifica-tion practices are to improve beyond short-term performance, then coaches and scouts needto be equipped with measurement tools with better long-term predictive utility. Addressingthese significant challenges will likely require collaborative efforts between stakeholders, policymakers, practitioners, and scientists.

Proposition 8: Increased competition for talent between sports underminesathlete development

Another factor that may further compromise the potential of talent identification and devel-opment (TID) relates to the protectionist and isolationist approaches to talent identificationand development. Indeed, a common refrain is that sport organizations are increasinglydemanding youth specialize in one sport early during development. The motives for promot-ing early specialization may stem from desires to have athletes devote more time to trainingand skill development from competition for talented athletes, and/or from the increasingcommercialization of youth sport (Hyman, 2012). In these environments, sports essentially

56 J. BAKER ET AL.

compete against each other for the youth samples fromwhich to identify and develop athletes.Regardless of the participant pool size available to sport systems, talent identification anddevelopment may be constrained by such inter-organization and/or inter-sport competition.

Culturally popular sports, like ice hockey in Canada and soccer worldwide, have the benefit oflarge participant samples. Yet in such circumstances, athlete development programs and orga-nization within that sport compete against each other; organizations vie for the commitment ofyouth athletes to their specific developmental program. For example, some sports have entrydrafts, where adolescent athletes are selected by teams in a competitive process analogous toNorth American professional sports (see Ontario Hockey League, 2016). In other contexts,developmental sport organizations extensively scout youth and adolescent athletes and attemptto secure athletes’ commitment to their team. One important question to ask is whether theperformance needs of the teams in these contexts, driven by inter-team competition, help tocreate circumstances where short-term priorities supersede the long-term development ofathletes (see Proposition 7). Another question: Do developmental organizations in popularsports make decisions that compromise long-term athlete development because they have theimplicit knowledge that there are large recurring youth participant populations to draw from inthe future?

Many other sports are less fortunate and have to generate successful athletes frommuch smaller groups of available athletes. Here too, there is competition for prospectivetalent, but it is inter-sport competition. Conscious that there is a limited population ofprospective athletes to draw from, these circumstances may promote early specializationand isolationist practices. In these contexts, organizations may also continue to invest inathletes when there are indications that they have a low likelihood of reaching the highestlevels of achievement. This approach, where sports operate completely separate from eachother, limits the opportunities to maximize the productivity of TID in sport.

Talent transfer programs aim to identify and transfer high performance athletes whohave greater potential for success in another sport, such as the one developed by theAustralian Institute of Sport to transfer athletes from track and field, surf life-saving, andwater skiing to the target winter sport of skeleton (Bullock et al., 2009). These types ofprograms aim to mitigate some of the above limitations by capitalizing on two factors.First, not all athletes within a developmental system will go on to achieve the highest levelswithin their sport. Second, some athletes have “talent” in a domain that can be readilytransferred to other sports. Here, the key is that athletes, who are likely already high-levelperformers, transfer from one sport to another. As long as task, environment, and/orindividual constraints align in a way that makes the transfer feasible, such programs arethought to increase the productivity of overall athlete development. While there have beensuccessful talent transfer initiatives (see Rea & Lavallee, 2017), it is far from commonpractice within and between sport organizations, and there is still a great deal about theprocess that requires further investigation.

Moving the discussion forward

Our objective in this short discussion article was not to provide conclusive evidencefor the notion of talent, nor to advocate for its elimination from athlete developmentmodels. In truth, there is very little objective and rigorous evidence to build a case foreither conclusion. Nor have we discussed all of the possible multidisciplinary

QUEST 57

constraints that compromise talent identification. Rather, our objective was to high-light the need for continued discussion of this concept since it permeates all aspects ofathlete development—from decisions parents make about what sports to enroll theirchildren in, to the policies governments create that determine sport funding. In orderto move this discussion forward, we advocate continued examination of each of thepropositions noted above with specific reference to the following points.

(1) There is a strong need for greater research and evaluation in this field ofscience. A recent systematic review (Robinson, Wattie, Schorer, & Baker, 2016)noted a surprisingly limited evidence base for coaches to use to inform theirtalent selection decisions. This has clear implications for the validity of talentselection and limits our theoretical and conceptual understanding of talent(e.g., What is it? Can it be validly and reliably measured? How does it changeover time?).

(2) It is important to emphasize that talent selection is only one (albeit important) step inan athlete’s journey, and the quality of the talent development environment (TDE) isarguably more important to long-term success. One of the most important findings inlearning and development research over the past 25 years is the strong positiverelationship between time spent in high quality (i.e., deliberate) practice and improve-ments in skill (see Baker & Young, 2014 for a review). However, an often-overlookedelement of this research relates to how coaches and other stakeholders actively createthese environments. Although there has been some examination and discussion ofTDEs (see Ivarsson et al., 2015; Li et al., 2014; particularly, Martindale, Collins, &Daubney, 2005), greater work is needed, particularly in tracking changes in thisenvironment over developmental time.

(3) Greater engagement from all stakeholders in the talent identification, selection, anddevelopment systems is essential. While sport stakeholders are obviously engaged inthe athlete development process, it seems at times as if stakeholders and scientistsare not working together as effectively as possible. Developing an efficient andeffectual athlete development system requires that everyone work together with ashared goal and clear objectives. To do this would require engagement from allcategories of stakeholders—from athletes and parents, to coaches and administra-tors—as well as basic and applied scientists.

(4) Frank and honest discussions with stakeholders regarding the realities of workingin these systems is necessary. Although researchers may squabble about theprecise definition of talent, those working within the athlete development systemunderstand it quite simply. To them, it reflects a decision about the most effectiveuse of available (and often very limited) resources. It would be important forscientists working in high performance sport to understand the realities of appliedwork where differences between medaling and not medaling are determined bysmall percentages in absolute performance—differences that may not be readilydetermined using traditional statistical analyses. Importantly, it is critical that keystakeholders within the sport system create and then defend the standards ofmeasurement and improvement. For instance, a recent examination of the efficacyof talent selection in the NFL draft found a success rate of approximately 9% (i.e.,9% of eventual games played in the NFL were predicted by draft round; Koz et al.,

58 J. BAKER ET AL.

2012). While this seems low, the ultimate assessment of this value needs to comefrom those making key decisions within the sport, not from outside.Unfortunately, research (e.g., Pankhurst, Collins, & MacNamara, 2013) high-lighted the lack of consistency between groups of stakeholders (parents, coaches,and the sport’s national governing body) regarding issues related to talent identi-fication and development.

Concluding thoughts and implications for practice

Despite movements to emphasize the role of the coach and the importance of high quality“deliberate practice,” the notion of talent is firmly entrenched in athlete development systems. Atits essence, athlete selection decisions reflect a coach, administrator, or parent’s evaluation of thefuture prospects for their athlete. In the sections above, we highlighted several concernswith howmany sport systems conceptualize and measure talent, as well as emphasized key consequencesof current approaches. Despite limited research on the validity of the idea of sporting talent, andthe consequences of approaches centered on this notion, there are important implications forthoseworking in talent selection and athlete development. First, practitioners should not focus sointently on identifying and selecting talent. Evidence suggests that if it does exist, we do not knowwhat it looks like, and are poor predictors of athlete potential. Second, practitioners areencouraged to consider approaches to athlete development that allow athletes to move acrosslevels of the systemmore easily. At themoment, these systems are very restrictive; once an athleteis removed from the system, it is very difficult for them to re-enter. Talent transfer initiatives are agood step, but these usually happen relatively late in development. Finally, it is critical toemphasize the role of the coach as an applied sport scientist. Our experience is that coachesusually have considerable datasets that they accumulated over years of involvement.While somemay be hesitant at first to share the results, when they do, the system, as a whole, will improve.Ultimately, scientists and stakeholders need to do more to evaluate talent and its relevance forpromoting athlete development. This will involve much greater engagement and interactionamong those with vested interests in athlete development.

Notes1. Interestingly, there is some research (e.g., Baker & Logan, 2007; McCarthy, Collins, & Court, 2016)

to suggest that although relatively younger players are less likely to be selected into developmentsystems, those who are selected are more likely to emerge as more skilled performers in adulthood.

ORCID

Joseph Baker http://orcid.org/0000-0002-5686-1737Jörg Schorer http://orcid.org/0000-0002-4888-7048Nick Wattie http://orcid.org/0000-0003-2623-000X

References

Abbott, A., Button, C., Pepping, G. J., & Collins, D. (2005). Unnatural selection: Talentidentification and development in sport. Nonlinear Dynamics, Psychology, and Life Sciences,9, 61–88.

QUEST 59

Ahmetov, I. I., Kulemin, N. A., Popov, D. V., Naumov, V. A., Akimov, E. B., Bravy, Y., . . . Govorun,V. (2015). Genome-wide association study identifies three novel genetic markers associated withelite endurance performance. Biology of Sport, 32, 3–9. doi:10.5604/20831862.1124568

Bailey, R., & Toms, M. (2011). Coaching and the ethics of youth talent identification: Rethinkingluck and justice. In A. R. Hardman, & C. Jones (Eds.), The ethics of sport coaching (pp. 149–164).London, England: Routledge.

Baker, J., Cobley, S., & Fraser-Thomas, J. (2009). What do we know about early sport specialization?Not much! High Ability Studies, 20, 77–89. doi:10.1080/13598130902860507

Baker, J., Cobley, S., Montelpare, W. J., Wattie, N., Faught, B., & the Ontario Hockey ResearchGroup. (2010). Exploring mechanisms of the relative age effect in Canadian Minor Hockey.International Journal of Sport Psychology, 41, 148–159.

Baker, J., Cobley, S., Schorer, J., & Wattie, N. (2017). The Routledge handbook of talent identificationand development in sport. London, England: Routledge.

Baker, J., & Horton, S. (2004). A review of primary and secondary influences on sport expertise.High Ability Studies, 15, 211–228. doi:10.1080/1359813042000314781

Baker, J., & Logan, A. J. (2007). Developmental contexts and sporting success: Birthdate andbirthplace effects in NHL draftees 2000–2005. British Journal of Sports Medicine, 41, 515–517.doi:10.1136/bjsm.2006.033977

Baker, J., Wattie, N., & Schorer, J. (2015). Defining expertise: A taxonomy for researchers in skillacquisition and expertise. In J. Baker, & D. Farrow (Eds.), The Routledge handbook of sportexpertise (pp. 145–155). London, England: Routledge.

Baker, J., & Young, B. (2014). 20 years later: Deliberate practice and the development of expertise insport. International Review of Sport and Exercise Psychology, 7, 135–157. doi:10.1080/1750984X.2014.896024

Barreiros, A., Côté, J., & Fonseca, A. M. (2014). From early to adult sport success: Analysingathletes’ progression in national squads. European Journal of Sport Science, 14(sup1), S178–S182.doi:10.1080/17461391.2012.671368

Barreiros, A. N., & Fonseca, A. M. (2012). A retrospective analysis of Portuguese elite athletes’involvement in international competitions. International Journal of Sport Science and Coaching,7, 593–600. doi:10.1260/1747-9541.7.3.593

Beamish, R. (1990). The persistence of inequality: An analysis of participation patterns amongCanada’s high performance athletes. International Review for Sociology of Sport, 25, 143–152.doi:10.1177/101269029002500204

Beamish, R. (1992). Towards a socio-cultural profile of Canada’s high performance athletes.International Review for Sociology of Sport, 27, 279–288. doi:10.1177/101269029202700401

Bridge, M. W., & Toms, M. R. (2013). The specialising or sampling debate: A retrospective analysisof adolescent sports participation in the UK. Journal of Sports Sciences, 31, 87–96. doi:10.1080/02640414.2012.721560

Brouwers, J., De Bosscher, V., & Sotiriadou, P. (2012). An examination of the importance ofperformances in youth and junior competition as an indicator of later success in tennis. SportManagement Review, 15, 461–475. doi:10.1016/j.smr.2012.05.002

Bullock, N., Gulbin, J. P., Martin, D. T., Ross, A., Holland, T., & Marino, F. (2009). Talentidentification and deliberate programming in skeleton: Ice novice to Winter Olympian in 14months. Journal of Sports Sciences, 27, 397–404. doi:10.1080/02640410802549751

Campbell, K., & Parcels, J. (2013). Selling the dream: How hockey parents and their kids are payingthe price for our National obsession. Toronto, ON: Penguin.

Charney Lawyers. (2016). CHL class action lawsuit: A lawsuit to pay the players. Retrieved fromwww.chlclassaction.com

Cobley, S., Baker, J., Wattie, N., & McKenna, J. (2009). Annual age-grouping and athlete develop-ment: A meta-analytical review of relative age effects in sport. Sports Medicine, 39, 235–256.doi:10.2165/00007256-200939030-00005

Cobley, S., Schorer, J., & Baker, J. (2012). Identification & development of sport talent: A briefintroduction to a growing field of research and practice. In J. Baker, S. Cobley, & J. Schorer (Eds.),Talent identification and development in sport: International perspectives. London, UK: Routledge.

60 J. BAKER ET AL.

Davids, K., Araújo, D., Vilar, L., Renshaw, I., & Pinder, R. (2013). An ecological dynamics approachto skill acquisition: Implications for development of talent in sport. Talent Development &Excellence, 5, 21–34.

Dries, N. (2013). The psychology of talent management: A review and research agenda. HumanResource Management Review, 23, 272–285. doi:10.1016/j.hrmr.2013.05.001

Dweck, C. S. (1999). Self-theories: Their role in motivation, personality, and development.Philadelphia, PA: Psychology.

Dweck, C. S. (2003). Ability conceptions, motivation and development. Development andMotivation, 2, 13–27.

Dweck, C. S., Chiu, C., & Hong, Y. (1995). Implicit theories and their role in judgments andreactions: A world from two perspectives. Psychological Inquiry, 6, 267–285. doi:10.1207/s15327965pli0604_1

Elferink-Gemser, M., Visscher, C., Lemmink, K., & Mulder, T. (2004). Relation between multi-dimensional performance characteristics and level of performance in talented youth field hockeyplayers. Journal of Sports Sciences, 22, 1053–1063. doi:10.1080/02640410410001729991

Elferink-Gemser, M., Visscher, C., Lemmink, K., & Mulder, T. (2007). Multidimensionalperformance characteristics and standard of performance in talented youth field hockeyplayers: A longitudinal study. Journal of Sports Sciences, 25, 481–489. doi:10.1080/02640410600719945

Gagné, F. (2013). The DMGT: Changes within, beneath, and beyond. Talent Development andExcellence, 5, 5–19.

Heritage Canada. (2013). Sport participation 2010 research paper. Retrieved from http://publications.gc.ca/site/eng/434212/publication.html

Howe, M. J. A., Davidson, J. W., & Sloboda, J. A. (1998). Innate talents: Reality or myth? Brain andBehavioral Sciences, 21, 399–442. doi:10.1017/S0140525X9800123X

Hruby, P. (2016, October 19). The NCAA lets college Olympians collect cash for gold, becauseamateurism is a self-serving lie. VICE Sports. Retrieved from https://sports.vice.com/en_us/article/the-ncaa-lets-college-olympians-collect-cash-for-gold-because-amateurism-is-a-self-serving-lie

Hyman, M. (2012). The most expensive game in town: The rising cost of youth sports and the toll ontoday’s families. Boston, MA: Beacon.

Ivarsson, A., Stenling, A., Fallby, J., Johnson, U., Borg, E., & Johansson, G. (2015). The predictiveability of the talent development environment on youth elite football players’ well being. Aperson-centred approach. Psychology of Sport and Exercise, 16, 15–23. doi:10.1016/j.psychsport.2014.09.006

Jourden, F. J., Bandura, A., & Banfield, J. T. (1991). The impact of concepts of ability on self-regulatory factors and motor skill acquisition. Journal of Sport & Exercise Psychology, 8, 213–226.doi:10.1123/jsep.13.3.213

Kerr, G., & Stirling, A. (2013). Putting the child back in chidren’s sport: Nurturing youth talent in adevelopmentally appropriate manner. In S. Harvey & R. L. Light (Eds.), Ethics in youth sport:Policy and pedagogical applications (pp. 25–39). London, England: Routledge.

Koz, D., Fraser-Thomas, J., & Baker, J. (2012). Accuracy of professional sports drafts in predictingcareer potential. Scandinavian Journal of Medicine and Science in Sports, 22, e64–e69.doi:10.1111/j.1600-0838.2011.01408.x

Li, C., Wang, C. J., & Pyun, D. Y. (2014). Talent development environmental factors in sport: Areview and taxonomic classification. Quest, 66, 433–447. doi:10.1080/00336297.2014.944715

Loland, S. (2015). Against genetic tests for athletic talent: The primacy of the phenotype. SportsMedicine, 45, 1229–1233. doi:10.1007/s40279-015-0352-5

Martindale, R. J. J., Collins, D., & Daubney, J. (2005). Talent development: A guide for practice andresearch within sport. Quest, 57, 353–375. doi:10.1080/00336297.2005.10491862

McCarthy, N., Collins, D., & Court, D. (2016). Start hard, finish better: Further evidence for thereversal of the RAE advantage. Journal of Sports Sciences, 34, 1461–1465. doi:10.1080/02640414.2015.1119297

QUEST 61

Mokone, G. G., Schwellnus, M. P., Noakes, T. D., & Collins, M. (2006). The COL5A1 gene andAchilles tendon pathology. Scandinavian Journal of Medicine and Science in Sport, 16, 19–26.doi:10.1111/j.1600-0838.2005.00439.x

Musch, J., & Grondin, S. (2001). Unequal competition as an impediment to personal development:A review of the relative age effect in sport. Developmental Review, 21, 147–167. doi:10.1006/drev.2000.0516

Ommundsen, Y. (2001). Self-handicapping strategies in physical education classes: The influence ofimplicit theories of the nature of ability and achievement goal orientations. Psychology of Sportand Exercise, 2, 139–156. doi:10.1016/S1469-0292(00)00019-4

Ontario Hockey League. (2016). OHL priority selection history. Retrieved from http://ontariohockeyleague.com/priority-selection-history

Pankhurst, A., Collins, D., & MacNamara, A. (2013). Talent development: Linking the stakeholdersto the process. Journal of Sports Sciences, 31, 370–380. doi:10.1080/02640414.2012.733821

Papadimitriou, I. D., Lucia, A., Pitsiladis, Y. P., Pushkarev, V. P., Dyatlov, D. A., Orekhov, E. F., . . .Eynon, N. (2016). ACTN3 R577X and ACE I/D gene variants influence performance in elitesprinters: A multi-cohort study. BMC Genomics, 17, 285. doi:10.1186/s12864-016-2462-3

Pearson, D. T., Naughton, G. A., & Torode, M. (2006). Predictability of physiological testing and therole of maturation in talent identification for adolescent team sports. Journal of Science andMedicine in Sport, 9, 277–287. doi:10.1016/j.jsams.2006.05.020

Posthumus, M., September, A. V., Keegan, M., O’Cuinneagain, D., Van der Merwe, W., Schwellnus,M. P., & Collins, M. (2009). Genetic risk factors for anterior cruciate ligament ruptures: COL1A1gene variant. British Journal of Sports Medicine, 43, 352–356. doi:10.1136/bjsm.2008.056150

Rea, T., & Lavallee, D. (2017). The structured resychling of talent: Talent transfer. In J. Baker, S.Cobley, J. Schorer, & N. Wattie (Eds.), The Routledge handbook of talent identification anddevelopment in sport (pp. 443–454). London, England: Routledge.

Robinson, K., Wattie, N., Schorer, J., & Baker, J. (2016). Talent identification in sport: A systematicreview of 25 years of research. Manuscript under review.

Salo, A. I., Bezodis, I. N., Batterham, A. M., & Kerwin, D. G. (2011). Elite sprinting: Are athletesindividually step-frequency or step-length reliant? Medicine and Science in Sports and Exercise,43, 1055–1062. doi:10.1249/MSS.0b013e318201f6f8

Schorer, J., Büsch, D., Fischer, L., Pabst, J., Rienhoff, R., Sichelschmidt, P., & Strauß, B. (2012).Back to the future—A case report of the ongoing evaluation process of the German handballtalent selection and development system. In J. Baker, S. Cobley, & J. Schorer (Eds.), Talentidentification and development in sport: International perspectives (pp. 119–129). London,England: Routledge.

Schorer, J., Wattie, N., & Baker, J. (2013). A new dimension to relative age effects: Constant yeareffects in German youth handball. Plos ONE, 8, e60336. doi:10.1371/journal.pone.0060336

Simonton, D. K. (1999). Talent and its development: An emergenic and epigenetic model.Psychological Review, 106, 435–457. doi:10.1037/0033-295X.106.3.435

Steingröver, C., Wattie, N., Baker, J., Helsen, W., & Schorer, J. (2017). Geographical variations inthe interaction of relative age effects in youth and adult elite soccer. Frontiers in Psychology:Movement Science and Sport Psychology, 8, 278. doi:10.3389/fpsyg.2017.00278

Steingröver, C., Wattie, N., Helsen, W., Baker, J., & Schorer, J. (2017). The interaction betweenconstituent year and within-one-year effects in elite German youth basketball. ScandinavianJournal of Medicine and Science in Sports, 27, 627–633. doi:10.1111/sms.12672

Tetlock, P. E. (2005). Expert political judgment: How good is it? How can we know? Princeton, NJ:Princeton University Press.

Thomas, C. (2016, August 20). Built for speed: What makes Usain Bolt so fast? The Telegraph.Retrieved from http://www.telegraph.co.uk/usain-bolt-worlds-fastest-man/0/built-for-speed-what-makes-usain-bolt-so-fast/

Till, K., Cobley, S., Wattie, N., O’Hara, J., Cooke, C., & Chapman, C. (2010). The prevalence,influential factors and mechanisms of relative age effects in UK Rugby League. ScandinavianJournal of Medicine & Science in Sports, 20, 320–329. doi:10.1111/j.1600-0838.2009.00884.x

62 J. BAKER ET AL.

Vaeyens, R., Lenoir, M., Williams, A. M., & Philippaerts, R. M. (2008). Talent identification anddevelopment programmes in sport: Current models and future directions. Sports Medicine, 38,703–714. doi:10.2165/00007256-200838090-00001

Wattie, N., & Baker, J. (2017). Why conceptualizations of talent matter. In J. Baker, S. Cobley, J.Schorer, & N. Wattie (Eds.), The Routledge handbook of talent identification and development insport (pp. 69–79). London, England: Routledge.

Wattie, N., Schorer, J., & Baker, J. (2015). The relative age effect in sport: A developmental systemsmodel. Sports Medicine, 45, 83–89. doi:10.1007/s40279-014-0248-9

Webborn, N., Williams, A., McNamee, M., Bouchard, C., Pitsiladis, Y., Ahmetov, I., . . . Collins, M.(2015). Direct-to-consumer genetic testing for predicting sports performance and talent identi-fication: Consensus statement. British Journal of Sports Medicine, 49, 1486–1491. doi:10.1136/bjsports-2015-095343

Wulf, G., & Lewthwaite, R. (2009). Conceptions of ability affect motor learning. Journal of MotorBehavior, 41, 461–467. doi:10.3200/35-08-083

QUEST 63