Comparative analysis of the technical–tactical characteristics and position-specific features of the top four teams and their opponents in the 2023 FIFA Women’s World Cup
Description
Objectives: The aim of this study was to compare the technical–tactical and performance indicators of the top four teams and their opponents at the 2023 FIFA Women’s World Cup, with particular focus on positional differences. Methods: The analysis included 38 key variables related to passing, attacking, defending, and physical performance. Data processing was conducted using SPSS 29.0, applying a General Linear Model (GLM) and post hoc EMMeans analyses. The main effects and interactions of team (Top 4 vs. opponents) and position were examined. Results: The team had a significant effect on the number of attempts passes (p<0.001; η²p=0.031), the number of completed passes (p<0.001; η²p=0.043), the completions of passes (p<0.001; η²p=0.042), completed of line breaks (p<0.004; η²p=0.01), the completion of line breaks (p<0.002; η²p=0.012), indirect pressure (p<0.001; η²p =0.017), pushing on (p<0.003; η²p=0.011), and pushing on into (p<0.004; η²p=0.01). The defenders and midfielders of the Top 4 teams showed higher passing activity, more completed and attempted line break, and better high-intensity running performance in Zone 4 than their opponents playing in these positions. Conclusions: The findings confirm that the top four teams’ success was based on a structured, possession-oriented playing style characterized by high passing and attacking activity. Performance analysis provides valuable insights not only for scientific under-standing but also for practical football applications, including training design, tactical planning, and youth development. Identifying the indicators that distinguish elite teams helps to better understand the elements leading to success and to apply them effectively in practice.
Files
Steps to reproduce
Study Design Matches from the 2023 FIFA Women’s World Cup were analyzed using an obser-vational, cross-sectional design based on official performance indicators. The analysis included the teams that reached the semifinals—England, Spain, Sweden, and Aus-tralia—as well as all teams that played at least one match against these teams during the tournament. The primary independent variable was team type (Top 4 vs. opponents), while playing positions were categorized into four groups: goalkeeper, defender, mid-fielder, and forward. Variables Data were collected from the official post-match statistical reports available on the FIFA website. These data are publicly accessible; therefore, ethical approval was not required for this study. A total of 38 technical and physical performance indicators were analyzed, resulting in 823 observational units, where one unit represents the performance of one player in a given match. All count-based variables were standardized per 90 minutes to avoid bias due to differences in playing time. Prior to statistical analysis, tests of normality were conducted. The Kolmogorov–Smirnov and Shapiro–Wilk tests indicated deviations from normal distribution for several variables. However, the applied general linear model (GLM) is sufficiently robust to violations of normality, particularly in large samples. Therefore, further analyses were conducted, and significance levels were adjusted using Bonferroni correction. Statistical Analyses Statistical analyses were performed using IBM SPSS software. To identify differences between teams, a univariate general linear model (GLM) was applied with two fixed factors—team (Top 4 vs. opponents) and playing position (goalkeeper, defender, mid-fielder, forward)—including their interaction. Main effects and interactions were tested using F-tests, and effect sizes were reported as partial eta squared (η²p). The following benchmark values were used for interpretation: η²p ≈ 0.01 = small, 0.06 = medium, and 0.14 = large effect [10]. Pairwise comparisons (estimated marginal means, EMMeans) with Bonferroni ad-justment were conducted to examine positional differences. For significant team × position interactions, EMMeans analyses were used to identify the positions at which significant differences between team types occurred. Cohen’s d effect sizes were also calculated to estimate the magnitude of these differences, using the following formula: d ≈ Δ / √MSE For all results, the magnitude of the difference (Δ), 95% confidence intervals, and Bonferroni-adjusted p-values are reported.
Institutions
- Pecsi Tudomanyegyetem Egeszsegtudomany Kar