Issues in Athlete Identification and Selection: Are We Compromising Talent?October 10, 2018
Despite some important ethical and developmental concerns, early identification and selection is the modus operandi of high performance sport. Most sport systems internationally have limited resources for high performance athlete development and, as a result, have to make predictions about who has the greatest likelihood of future success. Notions of talent also play critical roles in most athlete development models, and despite any strong evidence either for or against the role of genetic factors in predicting long-term performance, the concept isn’t going anywhere. The purpose of this article is to highlight a number of issues related to the identification and selection of athletes and discuss what they mean for coaches and sport administrators.
0.1 Early talent selection assumes talent is a fixed capacity that can be identified early
In most sports, selection decisions happen on a fairly regular basis across an athlete’s development (e.g. ice hockey programs choosing athletes to play on representative-level teams or provincial teams choosing junior athletes to participate in the Canada Games). However, high performance sport is increasingly focusing on the identification of athletes at earlier and earlier phases of development; perhaps the most extreme example we are aware of is the signing of an 18-month old child to a 10 year “symbolic” contract with a Dutch professional soccer club. This is done in an attempt to provide athletes with the optimal developmental environment to achieve future success. However, there are considerable consequences associated with this approach (e.g. overemphasis on winning at the cost of fundamental motor skill development, greater likelihood of burnout and dropout from sport). Furthermore, selecting athletes early in development assumes that the factors associated with early success (e.g. superior physical skills) are stable indicators of what performance will look like in the future, an assumption that does not have good research support (Baker & Wattie, 2018). Key performance capacities like the ability to read patterns of play (e.g., in sports like soccer and football), anticipate the forthcoming actions of opponents (e.g., in sports like tennis and squash) and make good decisions about the best options in specific situations (e.g., in decision-making sports like volleyball and baseball) only emerge after considerable time spent practicing. As a result, they are currently impossible to identify in early development because there are no good early indicators.
0.2 Talent identification is not done on a level playing field
The reality is that the identification and development of athletes is not a meritocracy. A number of constraints influence who gets labelled as “talented”, and whether or not athletes have opportunities to progress through a high performance developmental pathway. In some cases policies can influence selection: the relative age effect sees older youth in their age groups more likely to be selected to competitive teams likely because they’re simply physically larger, more psychologically mature, and developmentally advanced (Wattie, Schorer, & Baker, 2015). Aspects of youths’ immediate developmental environment also influence their experiences in sport. For example, there is some research suggesting obvious factors like socioeconomic resources as well as less obvious factors such as broad, geographic variables (e.g., population) limit opportunities for athlete development (Woolcock & Burke, 2013). These factors, and others, can compromise the accuracy of selection decisions, as well as influence the size of the talent pool from which to select and develop athletes.
0.3 Multivariate approaches to talent selection may be problematic
A significant advancement in sport science has been the increased use of “big data” among high performance coaches and administrators to understand the complexities of play at the highest levels of competition. On the one hand, the trickle down of these big data strategies to talent identification and athlete selection seems reasonable – if big data can improve our understanding at the elite level, surely it can have some utility at lower levels to identify athletes with the greatest potential. However, these approaches tend to be multivariate (i.e., considering the relationships between a combination of different skills and outcomes), which may not be appropriate for identifying a coach or a team’s specific need (e.g., does a team need the best well-rounded player or a player with strengths in one key area?). In many sports, the performer with the best potential for success is not the one with the greatest combination of all-round skill. Furthermore, because multivariate approaches require larger than normal datasets, they typically have to rely on data from players who may not have played in recent years. This is problematic because it assumes that the variables that predict player selection do not change over time (see #4 below for more on this). However, in many sports, particularly team sports, athlete selection is based on who is on the team at this specific point in time, how the team’s needs will change in the future, and what players are available to maintain or expand the team’s repertoire of capabilities.
0.4 Athlete selection requires predicting the future of sport
A related point to the one made above involves the need for high performance coaches to predict how their sport will change over time. In essence, when coaches make decisions about athlete selection, they should be making predictions about what performance in their sport will look like in the X number of years left between an athlete’s current age and their age of peak performance, and whether the athlete has the skills and capabilities to reach that level of performance. Predicting what could win today in a developing athlete ignores the reality that their sport will evolve during their window of development (e.g., due to rule changes, advances in equipment and technology, etc.); the longer the window of development, the greater the potential for change. One prominent example is Usain bolt, who in younger years was considered to be too tall to be a world class 100m-sprinter because coaches believed that only smaller sprinters could reach the step frequency needed to be quick enough.
0.5 Short-term priorities undermine talent selection and development
One of the challenges to developing elite athletes is the conflict that exists between short-term and long-term goals. Identifying talent begins at young ages in many of our sport systems, and the process of developing talent into expertise (i.e., performance at the highest levels of a sport) usually occurs over many years, through many different training environments. During this process coaches and selectors often prioritize short-term goals: winning this year’s games, tournaments and championships. This can result in “performance identification” – the selection of athletes that serve immediate performance goals rather than athletes that have talent and tremendous long-term potential. Furthermore, when short term priorities dominate (e.g., when linked to financial incentives), the risk that training and recovery practices fail to serve the long-term interest of individual athletes increases. We can see this in youth leagues where teams are stacked with a large proportion of top players instead of spreading them around the league. This results in the team having a high likelihood of success that year, but a sub-optimal environment for development since challenge may be low. It also compromises the learning opportunities for athletes on other teams. In circumstances where such short-term priorities undermine talent selection, organizations may need shifts in culture and incentives that align with long-term athlete development priorities.
0.6 Increased competition for talent between sports undermines athlete development
Most athlete development models advocate a broad and diverse foundation of movement experiences during early phases of development (e.g., https://playmoresports.activeforlife.com/) while, paradoxically, the early experiences of high performance athletes have become more and more specialized. One factor driving this effect relates to the protectionist and isolationist approaches many sports have to talent identification and development. In many high performance systems, sports compete against each other for the highest quality youth samples from which to identify and develop athletes. For instance, we have heard of several examples of elite coaches who do not want their athletes to participate in other sports in the off season out of concerns that they get an injury that would affect performance in their main sport. As a result, they design a 12-month training program to keep their athlete focused in one sport. This inter- and intra-organization competition constrains opportunities for athlete development across the system. Isolationist approaches lead to an emphasis on what is best for the sport (e.g., 12-month training) instead of what might be best for the athlete (e.g., a diversified involvement where athletes play different sports during an off season). While we often see this happen in sports like ice-hockey, it may be particularly relevant in less popular sports that have a need to maintain minimum numbers to allow systems to run efficiently.
Due to the limited resources available in most high performance systems around the world, identification and selection will remain parts of an athlete’s journey from grassroots to greatness. That said, frank and honest discussions with stakeholders regarding the realities of working in high performance systems are necessary. While researchers may squabble philosophically about whether talent exists, those working within the athlete development system understand the rationale for selection quite simply. To them, it reflects a decision about the most effective use of available (and often very limited) resources. Better alignment between sports would facilitate greater opportunities for athletes to experience success through practices such as “talent transfer” between sports, and potentially allow sports to maximize the pool of talented athletes and the use of limited resources.
In summary, several issues compromise effective talent identification and development in sport. Moreover, these issues are not always mutually exclusive, further complicating the already challenging practice of developing high performance athletes. The challenge for researchers and practitioners alike is to test and implement creative strategies to mitigate the factors that negatively impact the efficacy of athlete development initiatives.
About the Author(s)
Joe Baker (@bakerjyorku) is professor of sport science at York University. He has been examining the factors affecting long-term development and performance of high performance athletes for over two decades. He currently works with several NSOs and PSOs in Canada (e.g., Wheelchair Basketball Canada, Golf Canada, Canadian Paralympic Committee, Canadian Sport Institute Ontario) to improve models of athlete development and the delivery of evidence-based approaches to skill acquisition.
Nick Wattie (@wattien) is an Assistant Professor in the Faculty of Health Sciences at the University of Ontario Institute of Technology. His researchers various factors related to talent identification and development in sport, expertise development, skill acquisition, and positive youth development through sport. He has worked and consulted with a number of sport organizations, including Wheelchair Basketball Canada, the Canadian Sport Institute Ontario, the Canadian Paralympic Committee and Ontario Soccer.
Jörg Schorer (@jrschorer) is a professor of sport and movement science at the University of Oldenburg in Germany. His research focuses on expertise development over the life-span. He has worked with the German Handball Federation, the German Baseball and Softball Association, the German Table Tennis Federation and the German Field Hockey Association.
Baker, J., Schorer, J. & Wattie, N. (2018). Compromising talent: Issues in identifying and selecting talent in sport, Quest, 70, 48-63.
Baker, J & Wattie, N. (2018). Innate talent in sport: Separating myth from reality. Current Issues in Sport Science, 3, 006.
Baker, J., Wattie, N. & Gilson, L. (2016). A decision-matrix for determining risk in talent selection. HP-SIRCuit Winter Issue.
Wattie, N. & Baker, J. (2014). Improving how we think about talent: Step 1, stop talking about talent. HP-SIRCuit, Fall Issue.
Wattie, N., Schorer, J., & Baker, J. (2015). The relative age effect in sport: A developmental systems model. Sports Medicine, 45, 83-94.
Woolcock, G., & Burke, M. (2013) Measuring spatial variations in sports talent development: The approach, methods and measures of “talent tracker”. Australian Geographer, 44, 23–29.
The information presented in SIRC blogs and SIRCuit articles is accurate and reliable as of the date of publication. Developments that occur after the date of publication may impact the current accuracy of the information presented in a previously published blog or article.