Structural and dynamical differences between short and long pass sequences: A temporal network analysis

We propose a temporal network framework to analyze football passing sequences, moving beyond static aggregate graphs to reveal dynamic structural properties. Analyzing 302 matches from the South Korean national team (1980–2010), Euro 2024, and World Cup 2022, we classify sequences into Short (SPS, s = 3 , 4 ) and Long (LPS, s = 5 − 8 ) Pass Sequences using a variance-maximization threshold. We identify a functional dichotomy: SPS acts as the primary driver of effective attacks, characterized by high forward velocity, directional synchronization ( r ≈ 1 ), and triangular motifs. Conversely, LPS exhibits a transition into a diffusive, low-velocity state used primarily for stability. Spatially, a universal core-periphery structure emerges where LPS dominates the midfield, while SPS characterizes critical zones. Comparative analysis reveals a key evolution: while the historical Korean team used LPS reactively when blocked, elite modern teams utilize it proactively to manipulate defenses. These findings establish sequence size as a critical order parameter governing the trade-off between attacking efficiency and structural complexity.