Análise multivariada para o talento esportivo de jogadores de futebol categoria sub-19 do Cuiabá Esporte Clube

Abstract

The process of detecting sport talent through methods for recruiting and selecting young athletes in sport is a task of great importance in view of their training for future participation in competitions. The aim of this study was to verify whether the variables obtained through the iSports talent detection protocol can be regrouped with new proposed tests and generate new variables that were not originally correlated, using the multivariate statistical model of Factor Analysis via Main Component Analysis. Thirty-eight athletes from the U-19 category of Cuiabá Esporte Clube participated in the sample. Two batteries were applied, one corresponding to the tests of the iSports protocol and the second corresponding to the additional laboratory tests, adding a total of 14 variables. When analyzing the data through analysis of main components were indicated 4 first components that together has the sum of the variation of 71.42% of the data. Through Factor Analysis, the variables were reorganized into 4 factors. In this way, we can group in a new set of variables the tests proposed by the iSports software and the additional tests proposed, leading us to reflect on the association between the existing valences behind the different tests and making it possible, for example, to model a new protocol that is capable of attending the multifactorial aspects of football, in the creation of constructs that include different tests and make the evaluation process more optimized.