How Strong Is Strong Enough? Assessing When Physical Performance Tests Cease to Be Predictive of Sprint Performance...

Vial, S, Scanlan, M, Beranek, P, Kadlec, D, Barley, OR, and Cochrane Wilkie, J. How strong is strong enough? Assessing when physical performance tests cease to be predictive of sprint performance in trained football players. J Strength Cond Res 39(10): e1195-e1201, 2025-Quantifying underlying capacities of sprint performance are useful for monitoring training and guiding interventions. Traditional statistical techniques, such as correlation and regression analyses, have been widely used to model relationships between physical capacity and sprint performance. However, these methods often assume linearity, potentially oversimplifying complex, nonlinear interactions, and generalizing group-level data that may misrepresent individual trends. For instance, strength metrics such as maximal half-squat strength have been correlated with sprint performance, suggesting that speed should increase proportionally with strength. Yet evidence indicates that, in practice, this relationship plateaus, with further gains yielding diminishing returns. Although general guidelines exist, no established methods exist for quantifying the individual contribution of physical performance metrics to sprint performance. Using a novel approach, this study identified the "saturation point" at which physical performance tests cease to predict sprint performance. We used random forest regression and partial dependence plots to analyze data from 60 male football players who completed strength (isometric mid-thigh pull [IMTP] and Nordic eccentric hamstring strength) and jump tests (countermovement jump and standing broad jump [SBJ]). The random forest model demonstrated high predictive power ( R2 = 0.85-0.87) for 20 and 40 m sprint times, identifying IMTP and SBJ as key predictors. Partial dependence plots revealed a saturation effect for IMTP at 2.0 body mass and SBJ at 0.29 m·s -1 , where further increases yielded diminishing returns on sprint performance. These findings suggest that while strength and power improvements benefit sprint performance, further gains beyond specific thresholds may contribute little to additional improvements in sprint performance.