Pitch networks reveal organizational and spatial patterns of Guardiola's F.C. Barcelona

Abstract

We investigated the particular organization of Guardiola's F.C. Barcelona during season 2009/2010, using datasets from the Spanish National League La Liga. Specifically, we constructed the corresponding pitch networks, obtained from all passes successfully performed by a team during a football match. Pitch networks are composed of nodes consisting of particular subdivisions of the field, which are connected through links whose weight ωi,j corresponds to the number of passes made from region i to region j. We performed a multi-scale analysis focused on evaluating the properties of pitch networks at different scales, from a partition of the pitch into 2 × 2 to 10 × 10 areas. For each scale, we calculated a diversity of network parameters of F.C. Barcelona and its opponents during the whole season. Next, we compared the properties of F.C. Barcelona pitch networks with the networks of its rivals. Our results show how, depending on the spatial scale, there are statistically significant differences between F.C. Barcelona and the rest of the teams of the Spanish league. These differences are particularly significant at the clustering coefficient, the network average shortest-path, and the number of nodes occupied by a team for partitions with a high number of subdivisions.