Analyzing the influence of player tracking statistics on winning basketball teams

Basketball player tracking and hustle statistics became available since 2013-2014 season from National Basketball Association (NBA), USA. These statistics provided us with more detailed information about the played games. In this paper, we analyze statistically significant differences in these recent statistical categories between winning and losing teams. The main goal is to identify the most significant differences and thus obtain new insight about what it usually may take to be a winner. The analysis is done on three different scales: marking a winner in each game as a winning team, marking teams with 50 or more wins at the end of the season as a winning team, and marking teams with 50 or more wins in a season, but considering only their winning games, as a winning team. The results of the analysis reveal a few categories that are significantly different between the winning and the losing teams, such as: the number of uncontested shots made, the number of assists and secondary assists, and the number of defensive rebound chances. Based on these results, we propose the effective passing ratio, a novel statistical category, which also demonstrates large differences between winning and losing teams.