e-space
Manchester Metropolitan University's Research Repository

    A multivariate and cluster analysis of diverse playing styles across European football leagues

    Plakias, Spyridon, Moustakidis, Eserafim, Mitrotasios, Michalis, Kokkotis, Christos, Tsatalas, Themistoklis, Papalexi, Marina ORCID logoORCID: https://orcid.org/0000-0003-1125-7015, Giakas, Giannis and Tsaopoulos, Dimitrios (2023) A multivariate and cluster analysis of diverse playing styles across European football leagues. Journal of Physical Education and Sport, 23 (7). pp. 1631-1641. ISSN 2247-8051

    [img]
    Preview
    Published Version
    Available under License Creative Commons Attribution Non-commercial No Derivatives.

    Download (283kB) | Preview
    Official URL: https://www.efsupit.ro/

    Abstract

    Performance analysis is a valuable tool for team coaches and has been the subject of extensive study in international research. A significant portion of the scientific literature in the field of football has been devoted to studying playing styles in recent years. The identification of playing styles is now regarded as crucial for conducting an efficient performance analysis. This study aimed to explore the variances in playing styles among eleven distinct European domestic football leagues. A comprehensive sample of 2996 matches, accounting for 5992 observations, was scrutinized. Nineteen latent variables, representing thirty-eight different game styles previously identified in sports science literature, served as the basis for this investigation. Multivariate analysis of variance (MANOVA) revealed significant differences across countries in ten out of nineteen variables. The variables with the highest effect sizes (partial η2) were transition game, effective game, and defending aggressively, implying that these factors contributed to the most substantial differences among countries. To visualize these disparities, the t-distributed stochastic neighbor embedding (t-SNE) method was employed. Subsequently, k-means clustering was applied to the t-SNE results, grouping the eleven participating countries into five distinct clusters. A unique playing style was discerned in the Scottish league (Cluster 4), setting it apart from all other leagues. Other clusters included Austria, Belgium, and Switzerland (Cluster 1); Spain, Turkey, and Croatia (Cluster 2); Greece and Italy (Cluster 3); and Germany and England (Cluster 5). The findings offer valuable insights for coaches, managers, scouts, and sporting directors, potentially guiding the development of effective game styles and enhancing recruitment strategies for both players and coaches.

    Impact and Reach

    Statistics

    Activity Overview
    6 month trend
    594Downloads
    6 month trend
    439Hits

    Additional statistics for this dataset are available via IRStats2.

    Altmetric

    Repository staff only

    Edit record Edit record