Influence of space on the consistency and identifiability of soccer pitch-passing networks

In this work, we use network science to investigate the influence of spatial information on the consistency and identifiability of soccer teams. First, we construct the pitch-passing networks that topologically represent intra-team passing patterns between predefined pitch regions during match play. Next, we investigate how robust is the characteristic organization of pitch-passing networks of teams in the Spanish national league along a whole season, quantifying the team’s consistency. Building upon the concept of differential identifiability, i.e., the ability to distinguish a team from the rest based on the analysis of their pitch-passing networks, we propose a methodology to quantify the contribution of each region to a team’s overall identifiability. By exploiting the Euclidean spatial encoding of pitch-passing networks, our approach enables the identification of regions that are most characteristic of each team’s playing style. This spatial perspective shows how, on average, a higher consistency exists on the side lanes of a pitch, decreasing in the rest of the regions. This fact is reported in all teams of the competition. On the contrary, region identifiability is particularly low on average and it is only informative for certain teams, specially those at the top of the ranking. By incorporating spatial information into the analysis, our methodology enhances the interpretability of consistency and identifiability metrics, offering valuable insights for practitioners aiming to understand and compare team strategies more effectively. Furthermore, our results introduce the concept of region consistency and region identifiability, which can be adapted to other kind of spatial networks.