Using Diagnostic Analysis to Discover Offensive Patterns in a Football Game

Abstract

Football is a popular team sport, for which analyzing a team strategies can reveal useful information for understanding and improving a team’s performance. For this purpose, a promising approach is to analyze data collected about a match using data mining algorithms. However, designing such approach is not trivial as a football match involves both the time dimension and the spatial dimension. In this paper, a diagnostic analysis based approach is proposed, which consists of preparing data from a match by considering the spatial dimension and then extracting sequential rules from the data. The proposed approach is illustrated in a case study to analyze the match between Germany and Italy at the 2012 European Championship. Results of this study show that threatening offensive patterns from the Germany team are identified, illustrating complex interactions between players for performance analysis.