MODELLING AND PROGNOSIS OF COMPETITIVE PERFORMANCES IN ELITE SWIMMING

Abstract

The study demonstrates that the performance of an elite female swimmer in the finals of the 200 m backstroke at the Olympic Games 2000 in Sydney can be predicted by means of the nonlinear mathematical method of a neural back-propagation network. The analysis included the performance output data of 19 competitions prior to the Olympics within a time period of 95 successive weeks and the training input data of the last four weeks prior to each competition. The training data were divided into two phases: (1) a two-week taper cycle, and (2) an earlier two-week high load cycle. The trained neural network was not only able to model the 19 competitive performances, but also to predict the performance in the semi final of the Olympic Games in Sydney on the basis of the two sets of training data during the preparation before that specific competition.