The use of accelerometers to quantify external load in rugby league match play.

Abstract

Introduction: This study quantified the distribution of accelerometer data in rugby league competition.

Methods: A squad of 35 male professional National Rugby League (NRL) players (age: 26.0 ± 3.4 years; height 187.8 ± 6.4 cm; weight 98.7 ± 9.7 kg) wore inertial measurements units (Vector S7, firmware; 8.1, Catapult Sports, Victoria, Australia) during the 2023 NRL season. Three-dimensional 100 Hz acceleration data were exported and manually processed for each individual game file. An "acceleration index" was calculated by combining the three accelerometer signals, subtracting the influence of gravity, and removing periods of rest (5 s of less than 0.3 G). The "acceleration index" demonstrated a bimodal distribution of each individual player game file. A Gaussian mixture model was fitted to the acceleration index resulting in a mean, standard deviation and proportion of low and high-intensity activity for each individual player game file. Linear mixed models were used to quantify the magnitude of differences in each of these variables between positional groups.

Results: Overall, across all positions, 33% of game play is spent completing low-intensity activity and 67% of high-intensity activity. There were minimal differences between positional groups in the mean, standard deviation and proportion of low and high-intensity activity, with total volume reflective of time spent on the field.

Discussion: The lack of differences between positional groups suggests that accelerometers quantify both the running and contact that occurs within rugby league competition. As such, accelerometers may provide a measure of total high and low-intensity work in a contact based running sports, providing alternative and/or additional information to running metrics derived from global positioning systems.