Cluster analysis on ball recovery and handling from top 5 European leagues reveals emerging tactical patterns in soccer

Abstract

Although playing styles in soccer have been widely studied—both qualitatively and quantitatively—the identification of emergent patterns resulting from the combination of these playing styles has not yet been addressed through a quantitative approach. Applying cluster analysis based on performance indicators related only to ball recovery and ball handling from 1826 matches in 98 teams across the top five European leagues, we propose three game models associated with different combinations of defensive and attacking strategies. High-block recovery with sustained possession (game model 1, 20 teams), medium-block recovery with a direct attacking approach (game model 2, 33 teams), and low-block recovery with a counterattacking strategy (game model 3, 45 teams). Significant differences in ball handling and recovery were observed between models (p ≤ 0.05). Teams employing game model 1 achieved more points per match (1.94 ± 0.4) and qualified more frequently for international tournaments, whereas teams employing game model 3 recorded fewer points per match (1.10 ± 0.28) and more relegations. The Bundesliga dominated by game models 2 (39%) and 3 (44%), exhibited the most balanced representation of game models, while Ligue 1 had the most diverse prevalence of models (35%, 5%, and 60%). This study provides a novel methodological perspective by focusing on emerging playing patterns (models), encompassing playing styles and supporting the adaptation of tactical preparation and player profiling selection according to the characteristics of the opponent.