Abstract
In the present study the course of basketball game is observed as a separate and comprehensive system consisting of the succession of characteristic game situations being defi ned as states of the game. Precise identification and follow-up of various game states enables the explanation of game flow. In accordance with that, a formal mathematical model of the system "basketball game" has been founded from the aspect of kinematic description. The model enables the recognition of two basic system states which were in the paper defined like the set offense/defense and the transition offense/defense. The basic aim of both teams engaged in a match is to maintain balance in their own system of game states. Large number of states in set or positional and transition game have been listed. The system for assessing basketball game states will enable, through its empirical procedures, the computation of transition probability among states. Such an analysis of states and substates in transition and set offense and defense should facilitate understanding of the structure of the game and scientific research and evaluation of performance. This new methodological approach, based on the formal mathematical models, can be a prerequisite for research studies on discrete stochastic processes using the Markov chains. The elaborated paradigm of the system analysis of basketball game states can be applied, subject to certain modifications, to the other team sports games with the ball.