Network models have plenty applications in sports analytics. In addition to many types of social networks in teams, organizations, and markets, two sport-specific types are of interest for collective behavior in team sports: interaction networks in which nodes represent players or spaces, and state-transition networks in which nodes represent the state of the game. After providing some background on the use of data in sports analytics, I will focus on association football (soccer) and discuss ongoing work on the detection of tactical player positions and team formations from tracking data.
↧