Data and the NBA: A Slam Dunk Approach to Basketball

“Welcome to the most unpredictable playoffs ever.”

Reggie Miller, former Indiana Pacer and three-point shooting legend, closed this year’s play-off promo with these words. While the top seeds from each conference did end up in the finals, this post-season also saw untimely starter injury, sensational comebacks and pivotal plays by unlikely stars. These playoffs were unpredictable, drama-filled and had the NBA world on their toes.

Certain aspects of basketball will never be predictable or measurable, and therein lies the thrill of it all. However, the expanding ability to collect and analyze data in the NBA is creating avenues to make the game bigger and better for everyone. Number crunching for sports is a little more exciting than you might think.

No one can deny the growing buzz around basketball analytics.

Easy access to websites like Basketball-Reference, a holistic database of game data, puts analysis in the hands of everyone from professional analysts to bloggers. Statistical models created with this data can be used to put a probability on a team’s win or subsantiate arguments about the MVP race. Given the enormity of data collected, the potential applications are endless and the number of basketball enthusiasts trying their hands at analysis is exploding. 

Here is a breakdown of the various 5-man combinations that the Houston Rockets played during the 2014-2015 season. It is straightforward to navigate the different stats on Basketball Reference to see which lineups are optimal for different in-game situations.


Let’s not forget that data collection was once a grueling task.

SportVU motion cameras were only installed in basketball arenas two seasons ago. These cameras support a technology called Player Tracking and have the ability to capture 25 frames of data per second. With these installations, the NBA is promising improvement in referee calls and the ability to capture each player’s every move. As the data is further explored, analysts will have the ability to quantify the effect and value of different plays and tactics across players and teams, letting coaches determine where a player truly excels. This would inevitably aid coaches in creating smarter strategy.

Not everyone is on board with data and analytics.

Serious interest for a movement towards analytics-focused basketball strategy has been voiced from higher-ups within the NBA. The eight-year Houston Rockets General Manager, Daryl Morey, founded the MIT Sloan Sports Analytics Conference to gather the most forward thinking and outspoken minds in basketball analytics. Many of the conference’s supporters are coaches and NBA players.

On the other hand, Hall-of-Famer Charles Barkley is anything but quiet about his disdain for sports analytics. And while I don’t doubt that other players, current and retired, privately agree with Barkley, I do think that having better education of how data collection and analytics can positively affect the game should be promoted. While Barkley’s career-ending knee injury probably couldn’t have been predicted, as many accidents can’t be, what if data could predict a good handful of injuries? Any player or coach would agree that injury prevention always trumps injury rehabilitation.

Some teams are already leading the way with great results.

Steve Kerr, Coach of the 2015 NBA Champions, used various wearable technologies such as Athos and Catapult to monitor and measure each of his players during practice. While looking at the data to measure player performance and ability was important, Kerr also used this real-time data to ensure he was benching fatigued players before they were at risk of injury. These technologies capture hundreds of data points per second and also offer the ability to quantify rehabilitation for injured players. Most notably, the Warriors used trends in the data to make real change, such as changing flight times after games so the team could get optimal rest and reduce stress. This season, the Warriors lost less minutes to injury than any other team in the NBA.

Don’t forget to cater to the fans.

Data can help those mapping out strategy on the floor and those hitting the buzzer-beaters. However, fans are the supporting base of the NBA and single-handedly move the league forward. While big data is increasingly being used across industries to improve customer experiences, the NBA has plenty of untapped opportunities to further delight fans. Social data can be used, for example, to measure fan sentiment about promotional campaigns to ensure marketing success. One example is the Indiana Pacers who utilized customer data collected from social data and arena wifi to improve one-to-one relationships with fans, create customer loyalty programs and improve their sponsorships.

Everything in basketball can’t be predicted. Everything in basketball shouldn’t be predicted. Data is never going to explain or predict the sheer excitement experienced during Tracy McGrady’s feat of 13 points in 35 seconds or Tyreke Evans’ half-court game-winning shot.  Instead, we should look at the opportunities data presents as a way to see more of what we love and to continue the NBA’s legacy as a place where amazing happens.