Runze Feng

This study aimed to identify and classify team playing styles in women's football and to examine their relationship with performance outcomes. A dataset of 2322 team-match observations from the English Women's Super League and domestic cup matches (2017–2025 seasons) was analysed using a set of technical and tactical variables describing possession, ball progression, attacking organisation, and defensive behaviour. Principal component analysis (PCA) revealed two main underlying dimensions of playing style, a possession–progression axis and a directness–tempo axis. Based on these dimensions, cluster analysis identified three distinct playing styles: low-possession reactive, direct long-ball, and possession-based progressive. External validation analyses showed significant differences between clusters across key performance indicators, including expected goals (xG), xG difference, shots on target, shots conceded, and match outcomes. In particular, the possession-based progressive style was associated with higher xG values, a positive xG difference, greater attacking output, lower defensive exposure, and a substantially higher probability of winning compared with the other styles. These findings indicate that differences in playing styles are closely related to both the creation and prevention of goal-scoring opportunities. Overall, this study provides an integrated perspective on team performance in women's football by linking playing style classification with probabilistic performance metrics, offering practical implications for coaching, tactical analysis, and match preparation.

This study examined whether travel direction, travel distance, opponent strength and match-site thermal conditions were associated with the performance of away teams in the Chinese Super League. Data from 701 away matches from 2017 to 2019 were obtained from Amisco Pro. Seventeen technical and physical indicators were reduced by principal component analysis to four match-performance dimensions, accounting for 67.9% of the variance. Linear mixed-effects models were estimated for away-team goal difference and for each dimension, with travel category, geodesic distance, pre-match Elo difference, temperature, relative humidity and their interaction specified in a single model. Neither travel category (F(3, 150.9) = 1.19, p = .317) nor geodesic distance (β = 0.066 goals per SD, 95% CI [−0.101, 0.233]) was clearly associated with goal difference. A 1-SD increase in home-minus-away pre-match Elo difference was associated with a 0.771-goal reduction in away-team goal difference (95% CI [−0.912, −0.630]). Higher temperature (β = −0.455 SD, 95% CI [−0.556, −0.353]) and relative humidity (β = −0.142 SD, 95% CI [−0.217, −0.067]) were associated with lower running volume and high-speed activity, with a negative temperature-by-humidity interaction (β = −0.135 SD, 95% CI [−0.191, −0.079]). Thermal conditions were not clearly associated with goal difference. Northward and southward journeys involved substantial origin-to-destination thermal contrasts, but these provided no clear information beyond match-site conditions.

This study examines the impact of age (U17/U21) and match type (friendly/official) on external load indicators in elite youth football players during key developmental transitions. Twenty-nine male players (U21: n = 13; U17: n = 16) from a Chinese professional club were tracked over a full season (U17: 37 matches; U21: 29 matches) using 10 Hz GPS. Data were analyzed using Mann-Whitney U tests, Cliff's Delta (δ), and Hodges-Lehmann median differences. In official matches, all external load indicators for U21 players were significantly higher than U17s, with the largest effect size in running distance at 15–20 km/h (δ = 0.593) versus total distance (δ = 0.202). In friendly matches, U21s remained superior in distance >25 km/h, accelerations, distance 15–20 km/h, and maximum speed, though the max speed gap narrowed by 21.94% compared to official matches. Except for maximum speed and distance >25 km/h, U21 external loads were significantly higher in official matches than friendly ones, particularly regarding deceleration. Conversely, U17 running distances at 15–25 km/h were significantly higher in friendly matches (p < 0.05). Therefore, this study found that U21 and U17 players performed better in official matches than in friendly matches; deceleration control ability is a key indicator of U21 players’ adaptation to high-intensity competitions; the unique negative correlation pattern of ‘medium-high intensity load-match type’ in the U17 group may be related to the decline in running performance caused by psychological pressure, conservative tactics or delayed decision-making in official matches.