Validity of a Microsensor-Based Algorithm for Detecting Scrum Events in Rugby Union

Chambers R.M., Gabbett T.J., Cole M.H.

Purpose: This study aimed to investigate whether data derived from wearable microsensors could be used to develop an algorithm that automatically detects scrum events in rugby union training and match-play.

Methods: Data were collected from 30 elite rugby players wearing a Catapult S5 Optimeye microtechnology device during competitive matches (n=46) and training sessions (n=51). A total of 97 files were used to train an algorithm to automatically detect scrum events using random forest machine learning. An additional 310 files from training (n=167) and match-play (n=143) sessions were used to validate the algorithm’s performance.

Results: Across all positions (front row, second row, and back row), the algorithm demonstrated good sensitivity (91%) and specificity (91%) for detecting scrum events in both training and match-play when the confidence level of the random forest was set to 50%. Generally, the algorithm had better accuracy for match-play events (93.6%) compared to training events (87.6%). The algorithm performed better for scrums involving five players or more, suggesting it may not be suitable for scrums involving only three players, such as in Rugby Sevens.

Conclusions: The developed scrum algorithm accurately detected scrum events for different positions in rugby union. However, practitioners are advised to adjust the confidence level based on the position to limit false positives. Further research is needed to develop additional algorithms for detecting other aspects of rugby union demands, such as contact and collision events.

View this research