Affiliation:
1. UFMG Soccer Science Center, Sports Department, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil
2. Chair of Performance Analysis and Sports Informatics, Technical University of Munich, Munich, Germany
Abstract
This study aimed to compare position-based variables between winning, losing, and drawing teams and to identify which among these variables would better predict the goal difference in the 2021–2022 German Bundesliga. The dataset comprised 275 official games, monitored through a semi-automatic camera system. For each game, players’ and ball’s positions were captured and exported in 10 Hz. Position-based variables comprised teams’ length, width, length per width ratio (LPWratio), stretching index, and near-the-ball space control in both offensive and defensive phases, as well as in the first and the second halves. Data were compared between the game outcomes using a Generalized Linear Model. Results indicated that winning teams in the Bundesliga tend to approach the first half of the game by expanding the field during the offensive phase (evidenced by a higher offensive width and stretching index), pressing high during the defensive phase (manifested through a greater defensive length), and exerting control over the space in the second half by concentrating players near the ball (indicated by higher near-the-ball space control). Differences between halves indicate that teams must be able to adapt their tactics during the game to succeed in the competition.