Ideally, athlete selection would reflect the end result of explicit choices regarding what characteristics are best suited for long-term success. Unfortunately, in sport settings, it is not possible to know the accuracy of deciding to choose one athlete over another since their careers diverge after this selection is made. One athlete moves forward into a system where they compete against better competition, with more capable peers and access to superior resources, while the other is often left outside this system, either by the system explicitly denying access or implicitly by the athlete integrating this information into their motivation and competency beliefs. Moreover, early attraction effects almost certainly lead to improved survival effects, such as when a player who is ‘likeable’ (an attraction effect) is given more opportunities to develop than someone ‘less likeable’ and eventually develops into a superior player due to these skill development opportunities. Due to these design limitations, the ultimate value of qualities such as coachability  and/or anti-social personality characteristics  as predictors of future success is unknown. Over time, these traits will be over-selected for, demonstrating strong consistency and robustness thereby giving the impression they are important for long term development when in fact they are only important for ‘selection’. Dawkins’  concept of ‘selfish genes’ is based on the notion that the qualities selected for in a competitive environment are not necessarily the most beneficial for the long-term success of the species, but rather are the most desirable genes/traits at that given moment in time. Similarly, our understanding of whether profiles of current athletes represent the actual elements of performance necessary for success or simply those most needed for selection is extremely limited.
The interplay between survival and attraction advantages may help us to better understand the secular trends and challenges of talent identification and development. The specific constraints  associated with performance in varying sports, and resultant sport-specific athlete characteristics , likely reflect a kind of ‘evolutionary environment of adaptedness’Footnote 2 (EEA). That is, athletes’ characteristics (e.g., anthropometrics) and performances reflect the environments in which those elements provide advantages. However, these constraints change and evolve over time. Additionally, the elements related to performance demands in the elite, adult context are fundamentally different (and possibly diametrically opposed) to those at lower levels of development. Although this has yet to be confirmed empirically, given the changes noted in youth sport over the past few decades, we hypothesize that the attraction advantages resulting from and associated with the EEA at the highest levels of sport, now pervade youth sport. This overgeneralization and/or extension of the attraction advantages related to elite, adult sport to youth contexts is at least partially responsible for the biases and inaccuracies (Type I and II errors) associated with early talent identification. Alternatively, youth selectors who are more concerned with immediate short-term performance than with long-term potential, will simply be selecting characteristics that fit an EEA for developmental sport. Because the survival and attraction advantages of youth and elite adult sport reflect the same characteristics to some degree,Footnote 3 albeit on different scales, the systems’ two EEA (developmental and elite) can and do produce expert athletes. However, like judging a peacock’s tail before it is fully grown, conferring attraction advantages to youth athletes increases the cost of misestimating the survival advantages of an athlete.
To be sure, these effects will be difficult to test experimentally, at least using traditional approaches. The time frames associated with athlete development and the range of potential variables affecting development make designs like randomized controlled trials prohibitive or unfeasible. However, if we focus on the convergence of evidence in an area, produced through research synthesis methods like meta-analyses as well as systematic and scoping reviews, answers may be possible. For instance, if we consider the research evidence on the value of personality characteristics for athletic success, certain facets of personality (e.g., conscientiousness) are related to team and individual performance . However, it is not possible to determine the extent to which these qualities were important for long-term success or just important at one point in time for being retained in the athlete development system. Moreover, without appropriate control groups and longitudinal designs, it will be impossible to tell. However, it may be possible to compare across sports having more or less rigid development structures, to see if these same qualities are equally valued across contexts. The notion here is that the qualities necessary for eventual success should be relatively stable across similar domains.
Importantly, the distinction between attraction and survival effects emphasizes the importance of establishing a strong theoretical rationale for the relationships under investigation. For instance, how is the personality characteristic of conscientiousness linked with performance? Is it because those who are conscientious are more obvious in how they engage and attend during training and therefore are more likely to be noticed by their coach as being ‘coachable’ and/or is being conscientious directly related to greater success?
Using evolutionary models to shift how we conceptualize athlete selection would involve a change in how we think about cause and effect. Evolutionary psychology, for instance, makes the distinction between ‘proximate’ or ‘immediate’ causes (e.g., an athlete was selected because they were seen as ‘coachable’) and ‘ultimate’ causes (e.g., ‘coachability’ is more likely to be chosen in selection settings because the athlete-coach relationship often requires a close connection and coaches may be drawn to athletes who are perceived to be ‘coachable’). Both of these examples deal with the same characteristic (i.e., coachability) but position it in different ways. While both are important for understanding this process and how it can be effectively managed by practitioners, research in this area has largely neglected ultimate causes for proximal ones.