Detecting regularities in soccer dynamics: A T-pattern approach

Abstract

The dynamics of play during professional soccer matches are complex phenomena that traditional approaches to the quantification of team sports are not fully able to identify. The aim of this study was to detect such dynamics through an analysis of temporal patterns. In particular, the objective was to reveal the hidden yet stable structures which underlie the interactive situations that determine the attack actions of play in soccer. The methodological approach is based on observational design, supported by digital recordings and computer analysis. Data were analyzed with Theme 6 beta software, which detects the temporal and sequential structure of data sets, revealing repeated patterns that may regularly or irregularly occur within a period of observation. Theme detected many temporal patterns (T-patterns) in the soccer matches analyzed. Striking differences were found when won and lost matches were compared. The number of pattern occurrences and the number of different T-patterns detected was greater for lost matches and lower for the won matches, whereas the number of events coded was similar. Theme software and T-pattern enhance research opportunities by moving further than frequency-based analysis of performance, making this method an effective research and support tool for sports analysis. Our results indicate a need for further investigation upon the possible links between temporal structure detection and human observations in soccer performance. This approach could assist both soccer teams’ staff and coaches in obtaining a greater understanding of game dynamics, providing information that current methods may overlook or not detect at all.