ABSTRACT

Besides technical skills and physical fitness, tactics is another major success factor in sports like team handball. Tactics in team handball and other sports is determined by the continuous interaction between the players of one's own and opposing teams. One approach to model the sequence of actions in sports is the analysis of so-called temporal patterns (T-patterns). The main idea of the pattern detection is the temporal relationship between the events. An alternative approach to model action sequences in sports is to analyze the temporal behavior of exact position data and classify similar patterns by means of artificial neural networks (ANNS). In this approach, so-called self-organizing maps (SOMs) allow automatic clustering of complex data to detect data patterns. The chapter describes the challenges, the methods and examples of results of these applications in team handball. It uses ANNs to analyze offensive and defensive behavior in team handball based on exact position data of specific game events.