May 11, 2018
Philip Maymin is an Associate Professor of Analytics and Finance at the University of Bridgeport Trefz School of Business. He is the managing editor of Algorithmic Finance and the co-editor-in-chief of the Journal of Sports Analytics. He has been an analytics consultant with several NBA teams and is the Chief Analytics Officer for Vantage Sports.
Philip brings the perspective of “Moneyball” to basketball. The Celtics, for example, have done a great job of putting players in positions that play to their strengths. They do this by analyzing the data of their players better than most other teams. High frequency data is in sports now, not just in trading. There are cameras in every professional basketball arena that produce play by play data showing summary statistics that coaching staff and in some instances, the public, can see.
Teams are producing models for how each player moves in combination with the other players on the court. The two most basic questions you ask of a player is “What does he do really well?” And “What does he do often?” Sometimes what players do often, doesn’t correlate to what they do well. Philip also discusses outlier players like Marcus Smart. He doesn’t have amazing scoring stats, blocking stats or anything else particularly extraordinary, yet when he is on the court, his team has a dramatically higher chance of winning.
How could we start going down the path of using data to create and put a team together? The myth is that GM’s are brilliant and have foresight on who to draft. If that were the case wouldn’t they be making dramatically more money? Also, how are they making decisions? GM’s mostly rely on their gut. Philip believes that EVERYTHING should be put in a system. If you take scouts opinions, analytics, GM’s opinions, etc. and put it all in a soup then you are just trading off your gut. You have essentially de-systematized the system.