A Bayesian perspective on geographic influences in coach decision-making for athlete identification and selection

Decision-making science examines not only what choices people make but how and why they make them. This study investigated whether systematic geographical differences exist in athlete selection within a development program. With 4805 athletes from a state-wide talent search program across 16 sports (19 disciplines) in Australia, selection patterns were analyzed between ‘Regional’ and ‘Metropolitan’ (South East Queensland) athletes. Athletes completed anthropometric, physical, and physiological assessments. A Bayesian hierarchical model revealed Regional athletes face systematically stricter selection thresholds, requiring 1.009 standard deviations higher than Metropolitan athletes to be selected (95% CI [0.656, 1.355], P = 100%). This effect size indicates that Regional athletes must perform at approximately the 84th percentile to achieve the same offering probability as an Metropolitan athlete performing at the 50th percentile. This varied substantially by sport, from 0.899 SD (Archery) to − 0.386 SD (Athletics 400 m). Notably, 16 disciplines showed no clear disparities for Regional athletes (effects ranging from − 0.216 to 0.211 SD). These findings suggest geographical location creates systematic disadvantages in talent selection, likely driven by economic realities, as supporting Regional athletes costs several times more through travel, accommodation, and facilities. Consequently, Regional athletes may require superior performance across selection criteria to justify these additional costs.