Social Network Analysis of College and Professional Basketball

Abstract

Social network analysis (SNA) allows us to quantify cooperative team sports such as basketball, and compare different passing patterns’ effects on a successful game. We analyzed games at both the college and professional level to identify common network metrics that predict a win. For both levels of competition, individual players who scored the most points per game were unlikely to also occupy the central position in their team’s network; however, we found opposite effects of out-degree centralization on college vs. professional basketball. Other factors that significantly increase chances of success at the professional level include a team’s total number of passes and level of reciprocity. These results may be used as a tool for coaching staff to improve their teams’ strategies in concrete, measurable ways, and help track the level of cooperation among their players more easily.