INTRODUCTION: Over the last decades, the prognostic relevance of performance assessments in talent identification and development (TID) programs in soccer is critically discussed in talent research. While a major part of studies deals with variable-centered approaches investigating the predictive value of single tests and/or their combination, also person-oriented analyses focusing on the player holistically, e.g., by examining player profiles, seem promising [1;2]. Therefore, the aims of the present study were (a) to explore different player profiles based on multidimensional performance assessments, and (b) to examine differences among those profiles regarding players` future success in youth elite soccer. METHODS: The study sample consisted of N=6523 male U12 players participating in nationwide conducted multidimensional assessments (12 outcome variables) in the German TID program [3]. To address (a), a deep learning factor analysis identified four underlying latent factors behind the assessments: (1) subjective coach evaluations of players` tactical, technical, and psychosocial skills; (2) age-related and anthropometric measurements, (3) technical skills; and (4) speed abilities. Those were used to discover player profiles via a k-means-cluster analysis. Regarding (b), it was assessed, whether players transitioned into a U15 at a German youth academy three years after the assessment. 570 players (8.7%) met the criterion. Chi-square tests examined differences between identified profiles and success rates. Odds ratios for being selected (for each profile) served as effect size. RESULTS: (a) The cluster analysis revealed six different player profiles: "Anthropometrically advanced high performers" (n=788), "Subjectively low rated, anthropometrically advantaged" (n=604), "Subjectively high rated and technically skilled" (n=1189), "Average performers" (n=1344), "Subjectively low rated players with low technical skills" (n=1157), and "Anthropometrically and technically below average" (n=1440). (b) Significantly different proportions of successful players among profiles were detected (p<.001). "Anthropometrically advanced high performers" obtained the highest chances for future success (22.6%, OR=2.6, p<.05), while the lowest chances occurred for the profile "Anthropometrically and technically below average" (1.7%, OR=0.2, p<.05). CONCLUSION: The results indicate differences in future success among the identified performance profiles of players. Similar to former variable-centered approaches, the person-oriented analyses confirmed the predictive value of the multidimensional performance assessments. Whether and to what extent benefits of the person-oriented approach (e.g., the potential to discover compensation effects) may provide valuable information for research and applied practice of TID processes needs to be examined in future studies.
© Copyright 2023 28th Annual Congress of the European College of Sport Science, 4-7 July 2023, Paris, France. Veröffentlicht von European College of Sport Science. Alle Rechte vorbehalten.