There is common agreement that an appropriate framework for training and adaptation needs to consider the complexity and non-linearity of athletic performance and its response to training. General concepts like dynamical systems theory (DST) deal with such characteristics under the idea of selforganizing systems but still lack of empirical verification in the field of training and adaptation. In this paper, first, a DST approach to training is detailed. Second, on the basis of empirical data of cycling training, it is evaluated whether training process related data support the proposed DST approach. Therefore, training and performance development of ten cyclists (recreational to competitive level) were monitored during a ten-week cycling training program. Additionally, changes occurring at the microscopic level in the neuromuscular system were analyzed by means of surface electromyography (EMG). The results underpin the non-linear and time-delayed relation of training and performance. Further, the performance development observed during the training program can be characterized by stable patterns in the performance dynamics, described as performance states and transitions between these states, promoting the concept of self-organizing states observable at the macroscopic level.
© Copyright 2011 International Journal of Computer Science in Sport. Sciendo. Alle Rechte vorbehalten.