The company that churns out a million points of data for each NBA game is now looking to go even deeper, mining those numbers for insight into players' fitness and training with an eye toward someday preventing injuries before they happen.
STATS is the provider of the SportVU system, a set of high-speed cameras now installed at 15 NBA arenas that track the movements of all 10 players, the ball, and the referees throughout the game, recording information 25 times per second. As a result, teams can see how far or fast a player runs during a game or uncover specifics. For example, in Game 1 of the playoffs against the Houston Rockets, Oklahoma City's Nick Collison had six touches from the elbow and the Thunder scored 1.5 points per possession in those instances.
Now STATS is working with six NBA teams on how to take that same data and apply it to getting not just the most effort from their players, but also the most efficient effort. The key lies in two measurements: training load and intensity. Paul Robbins, STATS's director of elite performance, calculates training load by multiplying a player's weight by their average velocity by the distance traveled, while intensity is a measure of how much energy a player expends.
With these figures in hand, teams can then personalize nearly every aspect of a player's routine, recommending what to eat and how much, the type of exercise they should take part in as well as the duration, how much sleep to get, all directed toward a goal of giving the best effort on the court.
"If a coach or trainer knows a guy is going to play 35 minutes, I want to prepare him to make those 35 minutes on game day the optimal 35 minutes it can be," said Brian Kopp, head of STATS's Sports Solutions Group. It's interesting to note that there isn't any new data being created for these performance analysis studies. The system works from the existing information captured by the SportVU system and sorts it in a new way.
The use of heart monitors or accelerometers (a device that measures acceleration) isn't particularly new; players have long been rigged up during practice so trainers can keep an eye on their movements and vital signs. Although those are still important measurements, Robbins said, as teams cut back on practice time later in the season, the in-game data become more important.
"What the strength coaches really want is to understand how the game load affected that player and what they can do to recover and get ready for the next game," said Robbins, an expert in cardiovascular programming and metabolic assessment who joined the STATS team last year. "If they have a back-to-back game, or if they're traveling somewhere else, get them the data as soon as we can so they know how to work with that player the next day."
The Phoenix Suns shared some of their recovery techniques in an azcentral.com story earlier this year. If a player needs to bounce back quickly after a high-intensity game, the Suns may strap them in knee-high compression boots that flush lactic acid to speed up circulation. Tyler Wallace, the Suns' performance recovery consultant, said the addition of objective data from a game adds to the training staff's mix of information and leads to workouts that are specific to each player.
"Looking at the physical and mechanical differences in each player allows us to individualize the [recovery] strategies," Wallace said.
Players have begun to buy in to this new approach as well. Robbins recalls sitting with one player for more than an hour, reviewing charts and graphs from past games in which that player said he felt more tired than usual. Robbins made some specific conditioning recommendations to help that player cope with the physical demands of back-to-back games.
It's not hard to look at this precise tracking of physical data and imagine what might be possible. With the right amount of information in hand, could teams eventually predict and prevent injuries from overuse? What if the Lakers knew that Kobe Bryant was pushing the limits of his training load after playing heavy minutes in the six games leading up to his torn Achilles on April 12?
Robbins said we're still three to four years away from compiling enough data from NBA players to begin to address injury prevention. And while he steered clear of any speculation about Bryant specifically, he did note that certain injuries appeared to be correlated with high training loads.
"Right now, there are certain players, there are certain injuries, that I believe we saw this year that if we went back and looked at it, we'd see that those loads were extremely high prior to that injury," Robbins said. "Now I can't say it's because of those loads caused that injury at this time, but once we have enough data we start cross-referencing that with injuries, we could have a better picture of that."