The relationship between collision metrics from micro-sensor technology and video-coded events in rugby union

This study aimed to determine the relationship between collision metrics from a commercially available micro-sensor technology unit (MST) and the count of collisions coded by expert video analysts in professional rugby union. Forty-four professional rugby union players wore MST units during match play. We analyzed 245 combined data files from 11 competitive matches, resulting in the inclusion of a total of 9202 individual collision events. Collision metrics (the count of collisions and the Collision Load™) were analyzed via the manufacturer's software. Each match was also video recorded and evaluated by two expert video analysts. Pearson's correlation coefficients were used to determine the relationship between the count of collisions coded by the expert video analysts, and both MST collision metrics. One-way ANOVA was used to determine whether differences in the Collision Load™ for individual collision events existed between different playing positions. Very large-nearly perfect correlations were observed between the count of collisions coded by the expert video analysts and both MST collision metrics (the count of collisions: r = 0.91, 90% CI = 0.89-0.93; the Collision Load™: r = 0.89; 90% CI = 0.87-0.91). Differences in the Collision Load™ for individual collision events were identified between different playing positions. Collision metrics registered by the MST software relate very strongly with the count of collisions coded by expert video analysts. The typical Collision Load^TM per individual collision event varies depending on player position. The application of automated collision detection for rugby union appears feasible.