In order to analyze the interaction between training and performance in Olympic archery an antagonistic model (PerPot) and neural networks were used to model individual performances of seven archers on the basis of training and competition documentations. The modeling resulted in moderate model fits for the PerPot-model, suggesting the use of the model for the rough planning of overall training load. The rather poor model-fits for the neural networks may arise from methodological problems of the approach as well as limitations in the quantification of training and performance in archery.
© Copyright 2008 International Journal of Computer Science in Sport. Sciendo. Alle Rechte vorbehalten.