Data-Driven Tactical Archetypes in European Professional Football: A K-Means Clustering Approach

Positional labels in football no longer adequately reflect the fluid and role-based nature of modern play. This study examines whether player roles can be identified directly from match performance data, without relying on predefined positional categories. Using a dataset of 1988 outfield players from the top five European leagues (2024/2025 season), K-means clustering was applied to eight per-appearance performance metrics. The optimal number of clusters was determined using the Elbow method and Silhouette scores. The analysis produced five statistically distinct player archetypes: Conservative/Disciplined players, Defensive Enforcers, Primary Attackers/Finishers, Active Wingers, and Creative Playmakers. The results demonstrate that data-driven clustering captures functional differences in player behaviour that are not reflected in traditional positional classifications. In particular, the model distinguishes clearly between defensive disruption roles, high-output attacking profiles, and intermediary creative functions. These findings support a shift from position-based to behaviour-based player evaluation. The study contributes to football analytics by providing an interpretable framework for tactical archetyping based on unsupervised learning. Practical implications relate to scouting and recruitment, where player identification can be aligned with functional performance profiles. Limitations arise from the absence of defensive tracking data, and future research should incorporate richer datasets and alternative clustering approaches.