Last season 15 NBA teams had partnered with STATS LLC to put special motion-tracking cameras that record every move a player makes several times a second. It’s a flood of information but it can be useful — the information can tell you how well a player shoots after two dribbles vs. a catch-and-shoot, then it can overlay that with a shot chart to show if the player is strong on the right or left side of the court. It can measure a player’s speed, leaping ability, where a rebound from a missed elbow jumper tends to go, pretty much anything and everything.
And now that system is going to be in every NBA arena.
The league itself is stepping up to foot the bill, reports Zach Lowe at Grantland.
The NBA and an outside tech consultant have reached an agreement to install fancy data-tracking cameras in all 29 league arenas before the start of next season, according to several sources familiar with the matter…
The cameras cost about $100,000 per year, and the expense is one reason 15 teams hadn’t yet subscribed. Some of those teams were waiting in hopes the NBA would foot the bill, and the league has apparently decided to do so sooner than many of those teams expected. Installing the cameras in all 30 arenas will expand the data to include every game played, providing teams with a more complete and reliable data set. It also raises the possibility of the league using statistical nuggets from the cameras during television broadcasts. A few teams have used in-game data at halftime to show players specific examples of things like rebounds they didn’t contest aggressively, or evidence they weren’t running as hard as usual.
This information is great — I am a proponent of teams, GMs and coaches gathering as much data as possible from every source (cameras, advanced stats, traditional scouts, their own eyes) to help them make better decisions.
The challenge is twofold: 1) How to find useful information in the flood of data? 2) How to pass that information along to players in a digestible form?
Just like “big data” in business or the flood of stats in baseball (and increasingly football, soccer and other sports) NBA teams need to figure out how to mine this information for things they can use to both teach players and win games. Toronto has taken an interesting approach to this, but we are early in the “what do we do with this info?” phase. Some teams are far more adept at this than others.
The bigger challenge is making it work for players. While a handful of players such as Shane Battier can digest a lot of information and transform it into useful knowledge, most players do not. First, you don’t want a player thinking too much on the court, you need to get them information in a time and way that they can incorporate it and make it more natural. Second, a lot of players (like most people in the general population) are visual learners — it’s far more effective to show them video clips of what they did right or wrong, or what you want them to do, than it is to just tell them. You need to make it second nature so that if a player sees an elbow jumper being taken he moves instinctively to where the rebound is most likely to go.
It’s a process, but pretty soon teams are going to have a lot of data at their disposal.