Great Lakes Analytics in Sports Conference
Associate Professor | University of San Francisco
Spectral Analysis of NBA Data
This presentation addresses the question of how to analyze group effects in NBA lineups. We show that noncommutative harmonic analysis is well suited to uncovering and quantifying group effects of all orders, from individuals, to pairs, triples, and groups of four and five players. We illustrate these ideas using NBA play-by-play data. Attendees will be able to explain the group-effects problem in analyzing NBA team performance, identify noncommutative harmonic analysis as a potential solution to the group effects problem, and explain how noncommutative harmonic analysis quantifies group effects and brings insight to NBA team evaluation.
Stephen Devlin has a PhD in mathematics from the University of Maryland and is associate professor at the University of San Francisco. He is a faculty affiliate of the masters in data science program at USF, and has served as the director of the undergraduate data science program.
Listen to a podcast interview with Steve and GLASC director Scott Tappa
Watch a video of Stephen's 2018 GLASC presentation
Back to Great Lakes Analytics in Sports Conference