A mathematical model based on discriminant function analysis for screening youth male football players with physical fitness profiles resembling professional standards

To propose a discriminant-function-based mathematical model to classify youth male football players according to the similarity of their physical fitness profiles to those observed in professional players. In this cross-sectional study, 171 Brazilian male outfield players (90 youth; 81 professionals) completed anthropometric assessment and performance tests (vertical jumps, 20-m sprint, 20-m change of direction, and Yo-Yo IRL1). Group differences were tested, and Fisher’s linear discriminant function analysis was used to derive a classification model (α < 0.05). Professionals were older, presented higher body mass, and showed better squat jump, countermovement jump, drop jump, and COD 20-m performance than youth players (all p < 0.05), while 20-m sprint and Yo-Yo IRL1 did not differ significantly between groups ( p > 0.05). The final discriminant model retained body mass, countermovement jump, and COD 20-m, showing a canonical correlation of 0.419, with group membership accounting for 17.6% of the variance in the discriminant function. The original classification accuracy was 67.8%, while the leave-one-out cross-validated accuracy was 66.7%. A parsimonious model based on routinely collected physical measures provides a complementary screening tool for identifying physical profiles resembling professional standards. Considering the classification accuracy, particularly among youth players, the model should be used alongside multidimensional evaluations.