Tactical solution with network science in soccer: Example of team Turkish super league from 2019-2020

Purpose of the research: The aim of this study is to make a tactical analysis of a team in the Turkish Super League using a network science approach. Method: In the study, the pass data of 34 matches in a season of a team in the Turkish Super League were used. The data of the matches are taken from the data source where the match data of the club is recorded, with the permission of the club's officials. Within the scope of the study, tactical analysis was carried out by using the modularity measure, one of the network science clustering analysis measures, through pas networks. In the tactical analysis, the variables of the match result, the goals scored and the goals conceded are analyzed with the modularity measure. Cluster analysis and modularity coefficients were made through the Nodexl program, and the widely used clause-newman-moore algorithm was used. Correlation analyzes of modularity values and goal statistics were performed using SPSS 26 package program. Results: The correlation coefficient between modularity and the variable of goals scored was calculated as -0.11. This negative relationship confirms our hypothesis, but the relationship was not confirmed because statistically the null hypothesis (absence hypothesis) could not be rejected. Likewise, although the correlation coefficient between total goals and modularity was calculated as 0.20 (weak relationship), this relationship could not be confirmed statistically. The correlation coefficient between the goals conceded by the team, which is the subject of our research, and modularity was calculated as 0.40, and there is a moderately similar relationship between these two variables, p=0.008<0.01. The modularity value gives us an insight into the tactical understanding of the team and the path to the goal. Our second hypothesis is that there is a similar and moderate relationship between goals scored and modularity value. Conclusion: These results show that the importance of the passing net in football is increasing day by day. The results of our study show that there is a relationship between the goals conceded by the team, which we have done the pass net analysis, and the net structure during the season. In this study, the matches of a team in a season are considered. However, it is thought that the analyzes that will be made with the participation of the whole league or more teams will reveal more important developments in this field and contribute to the literature. Keywords: Football, Cluster analysis, Modularity, Passing network, Super League