Bowling workload is linked to injury risk in cricket fast bowlers. This study investigated the validity of microtechnology in the automated detection of bowling counts and events, including run-up distance and velocity, in cricket fast bowlers.
Twelve highly skilled fast bowlers (mean ± SD age 23.5 ± 3.7 y) performed a series of bowling, throwing, and fielding activities in an outdoor environment during training and competition while wearing a microtechnology unit (MinimaxX). Sensitivity and specificity of a bowling-detection algorithm were determined by comparing the outputs from the device with manually recorded bowling counts. Run-up distance and run-up velocity were measured and compared with microtechnology outputs.
No significant differences were observed between direct measures of bowling and nonbowling events and true positive and true negative events recorded by the MinimaxX unit (P = .34, r = .99). The bowling-detection algorithm was shown to be sensitive in both training (99.0%) and competition (99.5%). Specificity was 98.1% during training and 74.0% during competition. Run-up distance was accurately recorded by the unit, with a percentage bias of 0.8% (r = .90). The final 10-m (–8.9%, r = .88) and 5-m (–7.3%, r = .90) run-up velocities were less accurate.
The bowling-detection algorithm from the MinimaxX device is sensitive to detect bowling counts in both cricket training and competition. Although specificity is high during training, the number of false positive events increased during competition. Additional bowling workload measures require further development.
McNamara and Gabbett are with the School of Exercise Science, Australian Catholic University, Brisbane, QLD, Australia. Chapman is with New South Wales Cricket, Paddington, NSW, Australia. Naughton is with the School of Exercise Science, Australian Catholic University, Melbourne, VIC, Australia. Farhart is with City Edge Physio, Ultimo, NSW, Australia. Address author correspondence to Dean McNamara at firstname.lastname@example.org.