By Jeremy Stierwalt
Service Line Director, Analytics
Optimal Solutions an NTT Data Company
With the NBA All-Star game over, my thoughts turn to the remainder of the season… and how these players will round out the year.
In my last few posts, I’ve had some fun exploring the growing role that big data and advanced analytics is playing in professional sports as we continue to develop the discipline. With the NBA season starting to heat up, it’s only natural that my attention has turned to basketball. Basketball, with it’s rapid-fire pace of play, dynamo stars and unique formations, lends itself to analytical scrutiny in a special way. It’s a game of strategy as much as action; many plays must be executed in order to achieve an end, and advanced big data analytics are perfect tools for harnessing this kind of tactical (and strategic) planning. And while the high scoring player who usually gets the credit and the applause, data can help shed light on all the steps (and supporting players) that got him there.
Two statisticians from Harvard, Dan Cervone and Alex D’Amour, have been working on a basketball analytics project to better measure the individual player performances, team dynamics and strategic factors that tell the ‘behind-the-scenes’ story of the game. They started with the theory that every ‘state’ of play in the game has an effect on the outcome, and attempted to develop a model that could measure for such factors such as position on the court, individual scoring abilities, locations of players, ball possession and on-ball tendencies.
Looking at all the small things, in theory, would illuminate bigger picture of the sport in ways before untold. It could also help give more credit to the ancillary players and their ‘assisting’ actions that ultimately help win games. A big name point guard like Tony Parker, for example might share some of his glory with a shooting guard or forward. I’ve tried to summarize this project, but for all the details see the link below!
MUST READ: DATABALL – http://grantland.com/features/expected-value-possession-nba-analytics/
The project is a work in process, so it’s impossible to determine the impact it could have on the game, however it’s something to ponder. Changing the way we measure the value add of each play and athlete could help coaches optimize their team’s overall performance, not to mention the effect it might have on individual salaries, their overall health, and team performance.
The possibilities are intriguing as big data solutions continue to evolve. Advanced Analytics, unlike basketball, seems to have a shortage of talent, however. It’s up to us in the field to continue to advance the cause and recruit for our ranks. Nate Silver might be the Tony Parker of basketball analytics, but the geeks in the trenches are needed to keep evolving the industry and the solutions that advance them.