Abstract
Predicting success is a common goal for ecologists and sports scientists, yet these disciplines rarely interact. Sports scientists often use tests of closed-skill or game performances, but these are critiqued for their inherent uncertainties in predicting success. In contrast, ecologists embrace variance, measuring traits under controlled conditions to make probabilistic predictions of success. Integrating ecological perspectives could enhance team selection efficiency in youth sports. Here, we demonstrate this concept using territorial contests in crayfish. As in sports, individual traits in crayfish can be measured rapidly but do not perfectly predict contest outcome. First, we simulated populations of 100 male and 100 female crayfish that competed in 20 rounds of contests and estimated how many individuals must be selected to ensure the top 10% of performers are included. Selections were based on individual traits (body length, claw size and strength) and/or contest outcomes. When few contests had occurred, the top 10% of individuals were most efficiently selected on individual traits but increasingly more on contests as rounds progressed. Empirical data supported these theoretical simulations. We staged 10 rounds of contests among 27 male and 32 female Cherax destructor. After two rounds, ∼21 individuals were needed to capture the top 3; by round 10, ∼5 were required. Taken together, our study provides an initial but compelling demonstration of how ecological models can help improve talent identification strategies in sport. Such an adaptive selection framework efficiently narrows down selection of high-performing individuals under uncertainty and has the potential to be applied to reintroduction and translocation strategies in conservation.