Prognostic Relevance of Motor Predictors in Early Adolescence for Reaching Professional Soccer Level in Adulthood

Abstract

Several talent development programs in soccer have implemented diagnostics measuring motor performance factors. However, there is debate in the field of talent research about the prognostic relevance of such diagnostics in early adolescence for adult performance and about possible biases in diagnostics caused by maturation related characteristics. PURPOSE: Focusing on players’ speed abilities (SA) and technical skills (TS) with simultaneous consideration for anthropometric and relative age characteristics, this study analyzed the long-term prognostic relevance of a nationwide assessment conducted within the German soccer talent development program. METHODS: The prospective cohort study investigated N=14.178 talented players from the Under-12 age group born between 1993 and 1995. These players participated in a motor diagnostics (sprinting, agility, dribbling, ball control and shooting) carried out in 2004, 2005, or 2006, respectively. The measurement model comprised the five motor tests as well as three additional covariates (players’ height, weight and relative age). As criterion for the adult performance level (APL) in the 2014/15 season, players were categorized as professional (N=89), semiprofessional (N=913), or amateur league players (N=13.176). The prognostic relevance for each motor predictor was determined using one-way ANOVAs. Additionally, in a logistic regression model, APL was predicted by two latent variable factors (SA, TS) and the maturation related characteristics as covariates. RESULTS: Each motor predictor discriminated between the APL with small effect sizes (each p<.001; η2≤.02). In the logistic regression model, TS (OR=6.72; p<.001) and SA (OR=4.58; p<.05) predicted the APL after controlling for height, weight and relative age. This multivariate threshold model explained R2=24.8% of the APL variance. CONCLUSIONS: The study proved prognostic validity of motor predictors over a long-term period (≈10 years). However, the effect sizes indicated that the predictors explained only small proportions of the APL variance. Therefore, repeated measurements and further predictors (e.g. personality or cognitive performance factors) in a dynamic and multidimensional approach should supplement the diagnostics. This study was funded by the German Soccer Association.