Classifying Player Profiles in Elite Women’s Football: A K-Means Clustering Analysis of Physical and Technical Data from the 2023 FIFA Women’s World Cup

SHEN, Q, DING, Junyuan, SUMMERS., H.D., KUBAYI, A., STRAFFORD, Ben and STONE, Joseph (2026). Classifying Player Profiles in Elite Women’s Football: A K-Means Clustering Analysis of Physical and Technical Data from the 2023 FIFA Women’s World Cup. Football Studies: 100026. [Article]

Abstract
As elite women’s football becomes increasingly physical and technically demanding, practitioners require objective, scalable methods to profile player behaviours and match-specific demands to inform coaching, recruitment, and training design. However, integrated clustering of physical tracking outputs and technical event indicators remains underused in this context. An exploratory, data-driven approach was used to classify player behaviours based on integrated physical tracking and offensive technical event data from the 2023 FIFA Women’s World Cup. The full dataset comprised 1,885 player-match performances; after applying the study inclusion criteria, 1,599 player-match records from 539 players were retained for analysis. The K-means clustering analysis applied to the combined physical and technical dataset identified eighteen player clusters, reflecting distinct offensive behaviours operationalised as joint patterns of technical execution and physical output. The Principal Component Analysis (PCA) projection was used to provide a low-dimensional visual summary of cluster separability. Most clusters appeared well separated (77.2%), whereas a subset were positioned in close proximity with partial overlap of observations in the projected space. Independently of the PCA visualisation, longitudinal cluster membership indicated that 184 players showed stable single-cluster assignment across matches (specialised profiles), while the remaining players were assigned to two or more clusters across matches (hybrid profiles). These results suggest the relevance of integrating technical and physical data and the value of unsupervised learning approaches in capturing the diversity of player behaviours within women’s football. The findings offer insights for coaching, recruitment, and training design by identifying nuanced player profiles, such as high-intensity forwards, aerial target strikers, and impact substitutes.
More Information
Metrics

Altmetric Badge

Dimensions Badge

Share
Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email

Actions (login required)

View Item View Item