Exploring Key Technical Indicators for Victory in Men's and Women's 3x3 Basketball Competitions Based on Machine Learning Models: A Case Study of the 2019-2024 World Cup and Olympics

In order to investigate the gender differences in key winning indicators of 3x3 basketball games, all men's and women's 3x3 competitions (World Cup and Olympic Games) from 2019 - 2023 were selected for this research. A decision tree model was constructed for men's and women's 3x3 competitions with 16 technical-tactical indicators as independent variables and game results as dependent variables. The results showed that (1) in men's and women's 3x3 basketball games, the number of 2-point shots made, the number of 1-point shots made, and the number of free throws made were the key technical-tactical indicators affecting the outcome of the matches. (2) In men's 3x3 basketball, defensive rebounds have a very important impact on the outcome of the game. A team has a 93% probability of winning when it makes at least four 2-point shots and at least 9 one-point shots, which represents a max sample size of 16%. (3) In women's 3x3 basketball, the number of 2-point field goals made, the number of free throws made, the number of missed and made 1-point shots, and blocks is a very important indicator of the outcome of the game. When a team made at least 9 1-point shots, at least 2 2-point shots, and at least 2 free throws shots, the probability of winning was relatively high at 92%, which represented 18% of the maximum sample size. By analyzing the results of the decision tree model, the relationship between the independent and dependent variables can be reflected in more detail, providing coaches with a comprehensive report of the game and a scientific basis for improving the team's competitive performance.