Mathematical modeling and optimization of athletic performance: tapering and periodization

Abstract We conduct a mathematical optimization of the training impulse profile to maximize performance for two seminal athletic performance models: the Banister et al. (1975. A systems model of training for athletic performance. Aust. J. Sci. Med. Sport 7: 57–61) fitness–fatigue impulse response model and the Busso (2003. Variable dose-response relationship between exercise training and performance. Med. Sci. Sports Exercise 35: 1188–1195) variable dose–response model. We discuss discrepancies in both the quantitative and qualitative aspects of the optimized training impulse profiles compared to what is physically plausible and relative to common training practices recommended in the empirical sports science literature, such as periodization and tapering. We analytically prove that, under the Banister formulation, the optimal training strategy always consists of sustained maximal training, followed by the complete cessation of training prior to the prioritized performance date, and therefore does not yield periodization, or tapering beyond this trivial form. This highlights a gap between theoretical prediction and empirically supported practice, underscoring the need for models that incorporate additional physiological effects. We then propose a set of illustrative and interpretable nonlinear modifications in the magnitude and time response to training in the fitness–fatigue impulse response model such that the optimized training impulse profile demonstrates these qualities.