The paper discusses the problem of analysing the psychophysiological indicators of athletes using modern information technologies for decision-making support. There is proposed a solution using the example of tennis, used to predict athletes’ victories, cluster tennis players by level of play, and deter-mine the playing style using a heat map of shots. An original mathematical model is used in the tournament recommendation module to calculate predicted tournament ranking points. The modal was probated using a dataset describing 1500 matches held in the Samara region in 2021. Clustering of tennis players allows evaluating changes in the level of play of different regions. The cluster-ing method uses three parameters to determine groups of players: age, rating and number of matches played per year. Based on the heat map, which shows the places where the player most often encounters the ball, the algorithm draws conclusions about the tennis player’s playing styles. Combination of the pro-posed solutions supports automated coach decision-making to improve the effectiveness of athletes’ training. #COMESYSO1120.