The objective of this study was to identify parameters that best discriminate between selected and non-selected players for the Brazilian under-19 men's volleyball team and propose mathematical models to identify high-performance players. To this end, 18 selected (16.89±0.96 years) and 138 non-selected (16.91±0.74 years) players for the under-19 team were assessed for the training profile, anthropometric profile, and physical performance level. The discriminant function analysis was used to build the models, with a significance of α<0.05. The spike jump reach showed a greater correlation with the discriminant scores obtained in the two models (r=0.701; r=0.782). The 10 variables included in Model 1 helped identify 88.9% of the players selected in their group of origin; Model 2 - obtained by the spike jump reach and duration of playing experience - identified 83.3% of the players selected. Therefore, coaches should be aware that differences between the selected and non-selected players are multi-factorial, with the spike jump reach being the most relevant assessment factor. Furthermore, good players for the selection can be identified using the two models: Model 1 promises greater success with ten assessments, whereas Model 2 allows the identification of suitable players for the under-19 men's volleyball team with only two simple assessments.
Keywords: Vertical jump; change of direction; collective sports; physical fitness; talent identification.