Accelerometry Insights Into Gym-Based Plyometric Exercises; Evaluating Metrics, Reliability, and Sensor Placement.

McGillivray, I, Murphy, A, Reid, M, Perri, T, and Duffield, R. Accelerometry insights into gym-based plyometric exercises; evaluating metrics, reliability, and sensor placement. J Strength Cond Res 40(8): e784-e791, 2026-Quantification of training load is important for athletic performance, yet traditional field-based accelerometry often fails to capture gym-based training demands, particularly in the vertical plane. This study investigates the use of triaxial accelerometry for assessing training load in gym-based plyometric exercises, examining the influence of sensor placement and the reliability of accelerometry-derived metrics. Ten elite youth tennis players (16.0 ± 1.6 years) completed a standardized plyometric session on 2 occasions, consisting of 4 sets of 7 exercises, performed in a superset sequence. Inertial measurement units (IMUs) were worn at 4 anatomical locations (thoracic spine, lumbar spine, left ankle, right ankle), collecting 4 key metrics: peak upward acceleration (PUpAcc), 1D upward PlayerLoad (1DPL-up), 2D PlayerLoad (2DPL), and 3D PlayerLoad (3DPL). Results showed skater hops consistently produced the lowest acceleration magnitudes, whereas altitude landings recorded the highest. Ankle-mounted sensors recorded significantly higher acceleration values than thoracic and lumbar placements ( p < 0.05), confirming the attenuation of signals up the body. PlayerLoad metrics demonstrated strong agreement across sensor locations ( r = 0.50-0.96) in 119/126 comparisons, whereas PUpAcc showed lower associations ( r = 0.01-0.57). Moderate to good reliability values (intraclass correlation coefficient = 0.50-0.84) existed across all metrics, with greater consistency in PlayerLoad metrics compared with PUpAcc. Overall, sensor placement influences accelerometry-derived measures of plyometric activities, with PlayerLoad metrics demonstrating intersensor agreement and reliability across placements. Although thoracic-mounted IMUs capture whole-body motion, they may underestimate lower-limb loading. These preliminary findings support using PlayerLoad metrics for monitoring gym-based plyometrics while highlighting the need for further research into PUpAcc as a measure of impact-related acceleration.