Analyzing Team Performance Using Graph Theory Metrics

Resumo

In the realm of sports analytics, identifying the critical factors that distinguish successful football teams from those that underperform is of paramount importance. This paper conducts a comprehensive investigation into the similarities and differences between championship-winning teams and relegated teams in La Liga by harnessing the power of graph theory. We develop detailed player similarity networks, where individual nodes represent players and weighted edges denote the degree of performance similarity derived from an extensive set of statistical data. By employing a range of centrality and connectivity metrics, we uncover structural patterns that are intrinsically linked to team success. Our findings reveal that championship teams typically exhibit higher values in metrics such as degree, closeness centrality, clustering coefficient, and PageRank—indicators of robust team cohesion and distributed influence—while relegated teams often rely on a few key players with high betweenness centrality, a pattern that exposes vulnerabilities in their network structure. These insights contribute valuable perspectives to team performance analysis, recruitment strategies, and tactical planning in modern football.