Quantifying important differences in athlete movement during collision-based team sports: Accelerometers outperform Global Positioning Systems

In this study accelerometers were more effective at quantifying important differences in movement between player positions and match halves compared to global positioning systems (GPS) during matches in a professional collision-based team sport, rugby union. The purpose of the present investigation was to determine the effectiveness of the respective technologies for detecting differences in measures of maximum mean movement between positions and halves during professional rugby union match-play. METHODS: Movement data were collected via integrated GPS and accelerometer units (OptimEye S5, Catapult Sports, Melbourne, Australia) for 30 professional rugby union players (16 forwards, 14 backs) across an eight-match season (256 match-half files). GPS sampling frequency was 10 Hz, while the accelerometer sampled at 100 Hz with an output range of ± 16 g. Accelerometer-derived PlayerLoad™ and GPS-derived measures of mean speed and metabolic power were analysed using a rolling average method to identify the maximum mean values across a 600-s epoch. General linear mixed modelling was used to predict measures, with fixed effects for player positions (backs, forwards) and match half (1st, 2nd), and with random effects for within-player variabilities, between-player differences, and between-match differences. Mean differences (backs - forwards) and changes (1st - 2nd) were assessed via standardization and magnitude based inference. RESULTS: Accelerometer derived PlayerLoad™ displayed moderate to large clear positional differences by half (standardized difference, ±90% confidence limits: 1st half -1.50, ±1.03; 2nd half -0.75; ±0.71) compared to generally unclear differences as quantified by GPS. All measures displayed clear half differences; however, PlayerLoad™ quantified the greatest magnitude of decline between-halves (forwards 1.66, ±0.71; backs 0.90, ±0.48). CONCLUSIONS: Positional and half differences in player maximum mean movement showed that GPS technology alone underestimated the activity of athletes during professional rugby union match-play. Accelerometers provide additional meaningful information that may aid practitioners in physically preparing and monitoring athletes in collision-based team sports.