The Next Big Thing in Basketball Analytics

Advanced metrics and NBA stats have done great things for the game, but the most important development would be the ability to quantify a player’s defense! Imagine that there is an opportunity to quantify the in-depth defensive abilities of NBA players just like we can measure offensive abilities. This would be an excellent improvement for the NBA community and the global basketball player market.

Based on STATS LLC‘s data generated from 3D, high-tech cameras, Sandy Weil presented a worthwhile stuff at the MIT Sloan Sports Analytics Conference

This season, 5 NBA arenas (San Antonio, Houston, Dallas, Oklahoma City, and Golden State) have those cameras which capture 25 images per second and record all activity on the court plus the geographical location (x,y coordinates). Not only the players, referees but even the movement of the ball throughout the court can also be tracked.

I’m fascinated by the fact that defender proximity to a player is going to be available. In other words, the defense will be more quantifiable with what this optical tracking technology brings to the table. Let’s hope that STATS LLC puts those cameras to 25 other arenas and starts to collect league-wide data. Then, this could be the pretty next big thing in basketball analytics.

Here’s other notables from the SSAC sessions:

  • General Managers and team owners are not interested in systems that measure player/team performance. Can analytics help teams win more games? They are looking forward to answering this question!
  • Companies in the oddsmaking industry are getting more involved in data-driven models. They continuously crunch numbers and adjust them to account for how teams are playing lately.
  • In the meantime, recency desires heavier weights than the averages for the season does. Assuming that it is normally distributed, calculating the standard deviation and mean helps us see whether the team is underachieving or overachieving for the compared time period.
  • The way the results of analytics work are being presented is so important. It really needs to be easily understandable.