Last week on the BS Report’s comprehensive preview of the NBA playoffs, Bill Simmons and Henry Abbott had the following exchange:
Simmons: Stats have kind of maxed out. I think they’ve found all the stats they need to find. Now its getting to psychology and the way your body is and–
Abbott: Can I get some money on that stats have maxed out thing? I’ll take the other side of that.
Abbott proceeded to tell Simmons about Optical Tracking Data, a new technology premiered at the MIT Sloan Sports Conference and blogged about here and on TrueHoop. Simmons hastily retreated, seemingly convinced that in fact what we consider “advanced” stats now is going to change a whole bunch in the next five years.
I attended the Optical Tracking presentation from Sandy Weil at Sloan, and came away impressed but also a bit hazy on exactly how the process works. In truth, Weil, Brian Kopp, and the gang at Chicago based Stats, Inc are still improving usability for their six unnamed NBA clients.
In a must read feature for Wired Magazine, Erik Malinowski tell us much more about how the future of Stats will change the way Simmons, Abbott and the rest of us understand the game. There’s looking at a spreadsheet of stats, there’s watching the game, then there’s capturing every movement on the court and logging it as data—25 times per second.
Here’s Malinowski on what the Golden State Warriors are finding:
For all that comp-sci wizardry, the Stats crew has helped articulate and define gads of previously unseen metrics. Through the Warriors’ first 14 SportVU-enabled home games at Oracle Arena, data showed that guards Monta Ellis and Stephen Curry were accounting for nearly 60 percent of the team’s entire ball possession.
Curry, the team’s point guard, had gotten 937 touches of the ball over that time, compared to 948 for Ellis, the team’s leading scorer. More interesting for the Warriors was seeing that the team had a 51.5 shooting percentage off passes from Ellis compared to 44.6 percent from Curry.
Stats can also take the data and turn it into top-down graphical representations of how each segment of the game played out, for a perspective of the game and its flow never before seen by players and coaches.
Stats brought this kind of insight to every other client as well. Should Houston guard Kevin Martin shoot the ball after holding it more than four seconds? Probably not.
Is there more to San Antonio center Matt Bonner’s remarkable three-point shooting this season than blind luck? Better believe it.
And let’s just say veteran point guard Jason Kidd still protects the ball as well as he ever did when he was a rising star at the University of California at Berkeley.
There have been plenty of other new data points, ranging from points per touch to catch-and-shoot field-goal percentage to secondary assists per game, even how physically far apart the players are from their defenders during the game. But what are coaches today doing with it, and is the sport prepared for such a statistical overhaul? These are questions that remain to be answered.
“Everyone sees that there’s going to be a lot of value here. It’s just a matter of what that value is going to be,” says Schlenk. He adds that while Smart and the Warriors’ coaching staff have seen some of the SportVU reports, “they haven’t really done a lot with it.”
In any modern venue, it’s normally a simple installation that could be done by a single technician in an afternoon. But because of the unique interior of Oracle Arena, Stats had to go back and remap where the cameras would be installed. It settled on two central overhead cameras — one above each side of the center-court line — and a pair overlooking each side of the court from above the seating bowl.
To nearby patrons, as well as members of the press sitting just 10 feet below, they look like something akin to high-end security cameras. But to Warriors executives, coaches and fans, they represent something much more intangible: hope.
I urge everyone to go read the piece in its entirety.
For now, keep this coaching truism in mind: we can’t improve what we don’t measure. Then flip it around a bit, and consider the implication: we can improve anything we measure.
Every team serious about winning an NBA title is seeking that edge, the undervalued segment of the market on which to capitalize. What could be a more undervalued piece of information than the one yet undiscovered, unmeasured, unquantified and unimagined.