Analysis of Key Performance Indicators and Major Winning Rules in Chinese Basketball Professional League

Abstract

Abstract Background Effective analysis of the game performance of basketball will help to explore the hidden, unknown and useful information in the game, which is able to help the coaches to clarify the technical and tactical deficiencies at this stage and optimize the training system and game tactical programs. The Chinese Basketball Association (CBA) represents the highest competition level of basketball in China and is highly correlated with the competition level of the Chinese national basketball team. The Chinese basketball national team, after missing out on the 2020 Tokyo Olympics, is once again missing out on the 2024 Paris Olympics, and there are currently few analytical studies that have been conducted on the performance of the CBA league games. Therefore, what are the key performance indicators (KPIs) for the different-paced games in the CBA league? Do the winning characteristics in different-paced games in the CBA league remain the same? What are the differences between the CBA league and high-level leagues? These questions require further in-depth study. This study aims to identify the KPIs for games with different rhythms in the CBA league and clarify the disparities with high-level leagues. This endeavor would contribute to the improvement of the CBA league and the overall competitiveness of Chinese basketball, ultimately fostering the vibrant development of basketball in China. Methods Using the 814 games from the 2021-2022 CBA League season (comprising 760 regular-season games and 54 postseason games) as a sample, machine learning algorithms such as K-Means clustering and C5.0 decision trees were employed. The Ball Possession Rate (BP) was used as the clustering indicator, and decision tree models were constructed for games with different tempos. Results (1) The constructed C5.0 decision tree models of the CBAshowed high prediction accuracy, with 93% for the mixed group, 91.43% for the fast-paced group, and 94.38% for the slow-paced group. (2) KPIs for the mixed group were 2P% (0.29), DRB% (0.20), and 3P% (0.18). The major winning rules are as follows: a) When 2P% > 50%, the winning probability will be > 65%. b) In this condition, if 3P% > 41%, the winning probability will be > 88%. (3) KPIs for the fast-paced group were 2P% (0.31), DRB% (0.14), and STL% (0.13). The major winning rules are as follows: a) When 2P% > 51%, the winning probability will be > 61%. b) In this condition, if 3P% > 41%, the winning probability will > 94%. (4) KPIs for the slow-paced group were DRB% (0.26), 3P% (0.24), and 2P% (0.22). The major winning rules are as follows: a) When DRB% > 73%, the winning probability will be > 75%. b) In this condition, if 3P% > 38%, the winning probability will > 93%. Conclusion Most winning teams in the CBA League are characterized by a balanced offense and defense, as well as a high field goal percentage. The fast-paced game group tends to favor proactive offensive strategies, while the slow-paced game group is more inclined to seek counterattack opportunities from defense. Through a comparison with previous studies on high-level leagues, it has been observed that 2-point field goals and defensive rebounds are common KPIs in the CBA League. In the context of the CBA League, the importance of 2-point field goals is higher than that in the NBA. Furthermore, turnovers, assists, free throw shooting and other technical indicators do not significantly impact winning outcomes in the CBA League, marking a distinct difference from high-level leagues.