Performance analysis is a rapidly evolving field in football and a subject of extensive international scientific research. Recognizing playing styles is now considered essential for effective performance analysis. This study aimed to create a map of 174 teams from 11 European leagues that could, through visualization, provide practical insights applicable to football teams' daily practice. The t-distributed Stochastic Neighbor Embedding (t-SNE) method was used to reduce the dimensions of 19 tactical situations derived from previous research. The resulting two coordinates were employed to generate a scatter plot, and simultaneous k-means cluster analysis (k= 11) was conducted. Greece (86%) and Scotland (83%) had the highest percentages of teams within the same cluster as their country's average, while Germany (11%) and Croatia (10%) had the lowest percentages. In terms of cluster dispersion, England ranked first with 9 clusters, followed by Spain and Germany with 7 clusters, while Greece and Scotland had the least with 2 clusters. The visualization and clustering of teams led to the following conclusions. a) There are variations in playing styles not only between teams from different countries but also within the same country, particularly when there is a disparity in quality. b) Coaches' philosophies and implemented strategies significantly influence the adoption of playing styles by teams. These findings provide valuable information for coaches, analysts, and team scouts, assisting them in their respective roles. By understanding the diverse playing styles present in European football, practitioners can tailor their approaches to optimize team performance and gain a competitive edge.
© Copyright 2023 Journal of Physical Education and Sport. University of Pitesti. Alle Rechte vorbehalten.