Some approaches dealing with position data of the players and the ball are introduced, and, among other things, it is shown how complex team interaction can be reduced to the interaction of simple patterns by means of artificial neural networks. This approach of net-based pattern analysis is dealt with in more detail in this chapter, where Tilp models and analyzes sequences of actions in handball. After introducing a number of up-to-date approaches, he describes an attempt of recognizing and characterizing offensive playing processes based on self-organizing neural networks. One promising finding of his analyses is that: In this approach, the deviation of only 1.2 m demonstrated a promising accordance of single action sequences with the related patterns. One important advantage of this approach compared to classical approaches is that context information is taken into account. Even when the final shot Position of two action sequences is the same, the preceding actions might have been significantly different and lead to different defensive countermeasures.
© Copyright 2020 Modelling and simulation in sport and exercise. Veröffentlicht von Routledge. Alle Rechte vorbehalten.