Search

The aim was to determine the validity of an open-source algorithm for measuring jump height and frequency in ballet using a wearable accelerometer. Nine professional ballet dancers completed a routine ballet class whilst wearing an accelerometer positioned at the waist. Two investigators independently conducted time-motion analysis to identify time-points at which jumps occurred. Accelerometer data were cross-referenced with time-motion data to determine classification accuracy. To determine the validity of the measurement of jump height, five participants completed nine jetés , nine sautés and three double tour en l'air from a force plate. The jump height predicted by the accelerometer algorithm was compared to the force plate jump height to determine agreement. Across 1440 jumps observed in time-motion analysis, 1371 true positives, 34 false positives and 69 false negatives were identified by the algorithm, resulting in a sensitivity of 0.98, a precision of 0.95 and a miss rate of 0.05. For all jump types, mean absolute error was 2.6 cm and the repeated measures correlation coefficient was 0.97. Bias was 1.2 cm and 95% limits of agreement were -4.9 to 7.2 cm. The algorithm may be used to manage jump load, implement periodization strategies, or plan return-to-jump pathways for rehabilitating athletes.

This investigation examined the effect of prior workload on high-intensity football match performance. Player load variables were recorded using a global positioning system and converted into composite variables: rolling season accumulated load (AL), exponentially weighted moving average acute, chronic and acute:chronic workload ratio (A:C). Match-play high-intensity performance-per-minute: accelerations (ACC), sprints, high-speed running (HSR) and high metabolic load (HMLd) distances; and situational and contextual variables were recorded for all games. Partial least squares modelling, and backward stepwise selection determined the most parsimonious model for each performance variable. Quadratic relationships of small to moderate effect sizes were identified for sprint AL and sprint performance, HSR AL and HSR performance, acute HMLd and HMLd performance, acute sprint load and ACC performance and A:C sprint load and ACC performance. Match performance was typically greatest between the mean and +1SD. High chronic HMLd, and combined acceleration and deceleration (ACC+DEC) load exerted small beneficial effects on HMLd and HSR performance, whereas high acute load exerted trivial to moderate negative effects. High sprint A:C exerted a small beneficial effect on sprint performance and playing position exerted small effects on HSR and HMLd performance. Prior workload has trivial to moderate effects on high-intensity match performance in professional players.