Simulation of interactive action sequences using the example of high-performance soccer.

点击翻译按钮获取中文摘要

In the course of increasing electronic collectability and AI-supported analyzability of behavioral data, the expectation that the analyzed dynamic processes can be optimized is growing. In complex interactive-communicative processes, such technical controls are more problematic, since here cognitive and emotional levels are addressed in addition to functional ones. The problem with AI-supported analysis in cognitive processes is that while observable behavior can be categorized and evaluated, communicating and implementing adequate or optimal behavioral patterns is often difficult due to the complexity of the behavioral dynamics. This leads to the fact that complex recommendations for action are reduced by the agents to highly simplified patterns, which subsequently lead to stereotypical behavior when implemented uncritically. A socially highly relevant process form that contains a complex combination of cognitive, communicative and interactive dynamics is that of the (sports) game. Apart from some approaches to cooperative games, the majority of games are win-oriented, i.e., the game goal is to achieve a better result compared to the opponents. Sports games, which are a popular "playground" for computer-based data collection and behavioral optimization, especially belong to this category. A very prominent example of this is soccer, which most recently provided illustrative material for the possibilities and limitations of such computer- and AI-based analysis and optimization in the European Championship. The acquisition and interpretation technology for e.g. position data of players and ball, movement processes and action patterns up to the tactical-strategic evaluation of success is meanwhile very well developed. Less well developed is the ability to translate this into game action.The transfer of the potentially infinitely complex analysis results into practice and thus to the coach-player and player-player communications leads to a reduction to the basic recommendations and thus to an increasing stereotypy due to a lack of anticipation. An example of this is the "TikiTaka" ball possession strategy. The last few years have shown, parallel to the technical improvement of match data acquisition and match analysis, a significant increase of such reduced strategic concepts. And this year's European Championship in particular also showed stereotypical tactical interactions between the strategic chains over long stretches.The aim of the present application is, building on the positive results of the completed project, to establish a better link between the results of computer-aided game analysis and the possibilities of a more lively game design, in which success is achieved not through tactical "pushing" but through creative break-ups of tactical schemes. To this end, the approaches to creative training begun in the previous project are to be continued and specifically implemented, tested and evaluated at higher performance levels.