Deriving Movement Categories in Rugby Sevens

The primary aim of this study was to generate sport-specific movement category velocity thresholds for elite rugby sevens male and female players. Match activity data were collected via Global Positioning Systems (GPS) (10 Hz) from 19 male and 11 female players during 88 competitive international fixtures during the 2022/2023 and 2023/2024 seasons. A two-stage unsupervised clustering method was applied. The elbow method, a technique used to determine the optimal number of clusters in a dataset, was first applied to identify the number of movement categories. Spectral clustering was then used to define the velocity thresholds corresponding to each category. For both male and female rugby sevens, four movement categories were identified with varying velocity thresholds. The male movement category velocity thresholds were low (0.0-2.05 m.s^-1), moderate (2.06-4.26 m.s^-1), high (4.27-7.20 m.s^-1) and very high (> 7.20 m.s^-1). Although the female movement category velocity thresholds were low (0.0-1.87 m.s^-1), moderate (1.88-3.74 m.s^-1), high (3.75-5.97 m.s^-1) and very high (> 5.97 m.s^-1). A comparison of the total distance covered in the respective gender-specific zones revealed that females covered a significantly less distance in the low-velocity movement category (p = 0.02) and a significantly more distance in the very-high-velocity movement category (p < 0.001). This work informs customised movement categories that allow for better physical load assessments in male and female rugby sevens and the provision of sport-specific and gender-specific conditioning programmes.