Understanding Acceleration-Based Load Metrics: From Concepts to Implementation

Abstract

Accelerometer-based wearables offer a cost-effective solution for managing match and training loads in invasion team sports. However, a multitude of acceleration-derived metrics, each employing different algorithms, has led to inconsistent and incomparable outcomes across studies and devices. This article reviews the mathematical procedures underlying whole-body mechanical load metrics, clarifies their conceptual differences, and proposes refinements to enhance standardization. Synthetic data were employed to investigate conceptual differences, while experimental accelerometric data (463 time series) from a set of elite handball training sessions (involving 16 players) were used to implement the corrected equations and analyze statistical relationships. Analysis of synthetic data revealed that derivative-based metrics, such as Jerk Modulus (typically referred to as Player Load) and corrected Accel'Rate (cAccel'Rate), tend to amplify noise compared to acceleration-based metrics, such as universal Dynamic Stress Load (uDSL) and Body Load. Experimental results indicated that when metrics were summed, their values were nearly identical. In time-series comparisons, Jerk Modulus and cAccel'Rate were predictably found to be nearly identical, while Body Load was the most distinct. Acceleration-based metrics are preferable to derivative-based ones. Sports scientists should lead the design and validation of such metrics, ensuring methodological rigor, transparency, and innovation while preventing commercial interests from introducing rebranded variables with undisclosed scaling factors and unclear calculations.

Keywords: Accel’Rate; Dynamic Stress Load; Jerk; accelerometry; body load; workload.

Figure 1

Player Load estimation using published equation [11,21] from artificial acceleration data sampled at 273 Hz (blue) and 12 Hz (orange).

Figure 2

Jerk Modulus (corrected Player Load) computed from artificial acceleration data sampled at 273 Hz (blue) and 12 Hz (orange).

Figure 3

Jerk Modulus (corrected Player Load) computed from artificial acceleration data sampled at 273 Hz, with two points of disturbance (a value of 18 milliG).

Figure 4

Accel’Rate computed from artificial acceleration data sampled at 273 Hz (blue) and 12 Hz (orange).

Figure 5

cAccel’Rate computed from artificial acceleration data sampled at 273 Hz (blue) and 12 Hz (orange).

Figure 6

cAccel’Rate computed from artificial acceleration data sampled at 273 Hz, with two points of disturbance (a value of 18 milliG).

Figure 7

Body load [12] computed from artificial acceleration data sampled at 273 Hz (blue) and 12 Hz (orange), with two points of disturbance (a value of 18 milliG).

Figure 8

Comparison between the possible definition of impact in DSL by Gaudino et al. [15] (red) and the definition proposed in this study (green), computed from artificial acceleration data. The blue dashed line marks the 2 G threshold defined as the minimum acceleration value to be considered.

Figure 9

uDSL, computed using Equation (3), from artificial acceleration data with two points of disturbance (a value of 18 milliG), sampled at 273 (blue) and 12 (orange) Hz.

Figure 10

Effects of introducing noise-like data points into artificial acceleration data across each metric. On the left, the axis is normalized; on the right, the y-axis scale is logarithmic.

Figure 11

Time-series comparison between the normalized acceleration modulus (blue) and normalized mechanical load metrics. On the left, Jerk Modulus (green) and cAccel’Rate (orange). On the right, uDSL (orange) and Body Load (green).

Figure 12

Time sequence of all metrics synchronized. From top to bottom: Jerk Modulus, cAccel’Rate, Body Load, uDSL, and original acceleration data across three axes (Z: vertical, Y: medial-lateral, X: anteroposterior).

Figure 13

Scatterplot between every pair of metrics, with the regression line (black), the 95% confidence region (orange lines), and the 95% prediction bands (dashed lines). From top to bottom, left to right: (top left) uDSL (G³⋅s, x-axis) vs. Body Load (arb. u., y-axis); (top right) uDSL (G³⋅s, x-axis) vs. cAccel’Rate (G, y-axis); (centre left) uDSL (G³⋅s, x-axis) vs. Jerk Modulus (G, y-axis); (centre right) Body Load (arb. u., x-axis) vs. cAccel’Rate (G, y-axis); (bottom left) Body Load (arb. u., x-axis) vs. Jerk Modulus (G; y-axis); (bottom right) cAccel’Rate (G, x-axis) vs. Jerk Modulus (G, y-axis).