3 NBA Insights From The San Francisco Sports Analytics Summit

From Catapult.

Brad Tilley (@bradjtilley) is an Analytics Consultant in the health insurance industry. He lives in Bountiful, Utah with his wife and two daughters. This is his second freelance article for Salt City Hoops. He recently attended the Innovation Enterprise Sports Analytics Summit in San Francisco (#SportsSF), and gave us a report of what he learned. Thanks, Brad!

I attended this Sports Analytics Summit as part of my pursuit of my dream job to work in basketball analytics, but I learned many things that I think any NBA and Utah Jazz fan would be interested in. I was fortunate to hear from and speak to many professionals from the sports industry; specifically, the Golden State Warriors, Phoenix Suns and Los Angeles Lakers (yes, the Lakers) sent their analytics departments to attend. A handful of takeaways stood out to me that I thought would be of interest to Salt City Hoops readers.

SportVU data is not the only revolutionary basketball data set available.

At professional conferences like this, there is usually one presentation that gets the attendees buzzing. For this summit, that was the NBA data presentation from Vantage Sports’ Philip Maymin. Vantage Sports¹ takes basketball data a step further than most by employing dozens of analysts (or grunts), to watch every second of every NBA game and track every minute detail of every play. Everything from how a point guard defended against a screen, to who the help defender was, to whether the defender had his hand up to defend a shot.

On the other hand, the SportVU automated technology consists of several cameras installed in each NBA arena, which track the locations of the ball and the 10 players on the court. A major drawback to SportVU data is its raw size (imagine trying to figure out how to interpret (x,y) coordinate data that tracks the ball and every player 25 times per second). The graphic below gives some insight into the size of the types of NBA data available for a typical game.

Vantage data can provide users with up to 75 new statistics such as “Deflected Pass Rate”, “Solid Screen %”, and “Cut Efficiency”. This data, according to Maymin, is much easier to collect, and does not require complex computing power and algorithms which SportVU data analysis requires. My opinion is that both technologies have their merits. Vantage helps to bring in some of the subjective analysis that SportVU will never be able to capture. But the opportunities available to have data that can track how a player moves and how the ball moves during the course of a game opens up major possibilities for basketball analysis.

Sports science and medicine is the next big thing in basketball analytics.

“Wearable technology” was probably the most-used buzzword during the conference. This technology, such as Catapult Sports’ popular GPS-enabled vest, can track the distance a player travels in practice, heart rates, breathing patterns, and overall workload. This data can aid in evaluating a player’s practice workload and effort, as well as helping a team to monitor a player when they are returning from an injury. Interestingly, a team can only have players wear these devices for practice and not games, because of restrictions in the Collective Bargaining Agreement with the Players’ Union.

Another area of sports medicine that was highlighted by multiple presenters, including Nike, was the importance of athlete recovery after training or a game. A Sports Scientist from Nike showed this equation during his presentation: Performance = Fitness – Fatigue. His main point was that typically teams are so engrossed in conditioning their athletes (fitness) that they miss the whole other part of that equation (fatigue). The sports scientists at Nike work with individual athletes to track their training performance and recovery. One hot topic, and a large part of recovery, is “sleep performance”. Anyone with the right type of Fitbit can track their sleep patterns, but Nike takes it a step further. An example of the sleep data that they can capture for athletes is shown below.

Sleep data example.

Besides sleep, other types of recovery include ice baths, massages, acupuncture, inflatable compression sleeves, and cryogenic chamber rest. Dave Hamilton, the Director of Performance Science for the U.S. Field Hockey Team cited two very interesting studies. The first found that regardless of recovery type, athletes actually recover from workouts better when they are together with their team as opposed to recovering alone. Another study showed that athletes recover better after a win compared to a loss.

The Jazz data that I could find that was relevant to these sports science topics comes from SportVU data that has been organized on the NBA stats website. The table below shows the breakdown of total distance the players travelled over the season, as well as speed and distance per game. This is a snippet of data compared to the data that the Jazz could theoretically capture if the players practiced with wearable technology that could track their speed, distance, and overall workload during the workout.

Tim DiFrancesco, the Head Strength and Conditioning Coach of the Lakers, highlighted the recent change in how teams manage a player’s workout. In the past, a player “arrived for the workout, warmed-up, and trained”. Now a player “arrives, applies wearable technology, submits physiological and psychological data, warms-up, trains, recovers and reviews performance data”. All of these examples provided by industry experts illustrate how sports performance is becoming much more of a science. Following these sports science and medicine trends give teams a leg up in developing elite-level athletes and teams.

“Don’t believe everything you read.”

NBA teams keep their analytical capabilities very close to the chest. That is starting to change, and some of my experiences at this conference were evidence of that. If you follow the NBA, you know that the stigma around the Lakers’ analytics department is that it is either non-existent or a non-factor. However, I met two members of the Lakers analytics department and there were others at the conference as well. I asked the Lakers stats expert what he thought about certain media coverage around the Lakers, and specifically the ESPN Great Analytics Ranking article. This article classified the Lakers as “Nonbelievers” and ranked them #113 out of #122 sports teams in terms of the strength of their analytics department. He cursed in disagreement to the article, then told me to “not believe everything I read” and that the Lakers have this stigma because they have not publicized their analytical work like other teams. Shortly after the conference I asked the same question to the Lakers Director of Analytics. “Now that we’ve been catching heat about it they’ve let us come out of the closet”, he told me as he stepped into his black, private car headed to the airport.

The Jazz have not been quite as tight-lipped as the Lakers in regards to their analytics department. In the last two years, the Jazz have announced the hirings of two Video Analysts, a Coordinator of Analytics, and recently a Manager of Basketball Strategy/Technology, which shows their commitment to using data to improve the performance of the team. We know that the Spurs have been analytical innovators for years, as well as the reigning champion Warriors (the Assistant GM of the Warriors presented on their analytical and sports medicine work). The fact that the Jazz are on the same path as these elite teams should instill hope and confidence in Jazz fans.