Observation of team behaviour in games shows rather inhomogeneous distributions of the levels of activity and effectiveness along the game. It is to expect, of course, that even for a high level team a phase of great effort is followed by a phase of reduced activities, while the opponent
team is increasing its pressure. The result could be thought to be something like a rhythmic change, where the maximum activities of the one team correspond to the minimum ones of the other team. However, as we got from handball data that Martin Lames from the University of Ulm/Germany
recorded last year (private correspondence), the correspondences can be much more difficult and varying: Lames measured the scoring effectiveness depending on the aggregated numbers of ball possessions that formed a non-linear transformation of the time axis. The results were diagrams where
high and low performance intervals of both of the teams can alternate, follow and overlap each other or even can appear concurrently. Based on the interpretation that the subsequent effectiveness of the one team is a kind of performance, which in a delayed way is caused by the past and
present pressure or load from the opponent team – i.e. its activities and scoring – the game can be understood as a symmetric process of load-performance-interaction. Such processes can be analysed by means of PerPot, which originally was developed in order to analyse physiological
adaptation processes. The new aspect of the presented approach is that in case of games the performance output of the one team forms the load input of the opponent one, where the delay of the reaction characterizes the temporary state or the temporary fatigue of the respective team. The
first results of this approach are that the types of local correspondences of the effectiveness patterns can be simulated quite well. Even the qualitative reproduction of longer sequences of patterns is satisfying. The simulation is controlled only by the respective delay values. Currently,
some work has been done in order to improve those results and the understanding of the team interaction dynamics. The aim could be to predict up-coming weak phases earlier and control and optimize the activities of a team.