"One of the greatest problems in understanding, planning, and evaluation of adaptation in high load training processes in sport is the number and the interaction of variables that influence the desired parameters ... " This first sentence of a contribution2, which dealt with an interdisciplinary project concerning sport medicine as well as computer science, describes a well-known problem in applied modelling and simulation: Although the struc-ture of a system on the first glance seems to be not too complex, the model finally becomes complicate and non-transparent due to the lots of components and parameters, and their interactions and influences. Doing it bottom up, it is difficult to decide which of the components are really necessary, and which interactions are really essen-tial for the system behaviour. Improper system understanding and inadequate reductions, however, can cause mistakes, which overlay or falsify simulations. Expensive evaluation processes followed by revisions or even redesign are the consequences. One solution for solving this problem - or at least to reduce the above-mentioned fatal consequences - is the development of meta-models in combination with on line-simulation.
© Copyright 2000 Alle Rechte vorbehalten.