In this paper team sports games, as are: water polo, soccer, basketball and handball, were analysed using bottom-up approach. Based on (kinesiolgical) activity analysis, and, referring to the fields of game theory, robotics (AI) and complex system theory, granule is identified as basic ingredient of team sports games. Obtained lump of such non-atomic elements (e.g. moving, passing, dribbling, orientation, positioning, player role, playing rules, playing field, tactic etc.), i.e., granules, is found to be structured in such a way that micro, intermediate, and, macro levels can be identified. While each level in the structure is internally consistent set of simple elements, to explain their collective properties, or, moving upward along structure levels, the well-known emergence property was necessary to be employed. Thus, the necessity of introduction of complex system theory in modelling of team sports games is concluded. Aim of this analysis was formalization of team sports games to enable computer simulation. Simulation is suggested to be most appropriate scientific tool of today for modelling of team sports games. Some existing software packages and/or platforms (Soar, SDML, Swarm, RoboCup simulator) are shortly discussed and evaluated for the possible use in modelling and simulation of team sports games. Since different complex systems, in general, can be grouped according to some common features, intuition and insight gained in one can be transferred to another, it is suggested that studying of team sports games can yield to the development of complex system theory itself, and, also, beside kinesiology/sportscience to some other scientific fields as are sociology, organizational theory, or biology.
© Copyright 2002 International Journal of Computer Science in Sport. Sciendo. Alle Rechte vorbehalten.