The Celtics have prided themselves on being a great defensive team since Kevin Garnett put on the green for the first time. The defense is something that they’ve been able to hang their hats on–even with a decreasingly efficient offense. They’ve been top 10 in defensive efficiency every year since Kevin Garnett has been a Celtic, and that’s in a large part due to his all around stellar defense on the interior.
This weekend, if you didn’t know, the 7th annual Sloan Sports Athletic Conference took place at MIT. I didn’t have the privelege of going to the conference myself and interacting with everyone there, but I did get a chance to do some reading on some of the research papers. You should take a read yourself when you get the chance.
There was one presentation that was really interesting to me. It was called the Dwight Effect by Kirk Goldsberry of Grantland.com and Eric Weiss of SportsVu.
What they set out to do was determine a way to measure interior defense through spatial analytics. The NBA is a league that lacks a proper statistical measuring tool for defense and defensive ability. The public generally looks at steals, blocks, and defensive rebounds to judge defensive ability. Advanced stat geeks and NBA aficionados will look to defensive efficiency, but even that isn’t the most accurate manifesto of defensive ability.
We’ve looked at plus/minus numbers to determine who’s good defensively and what team gets worse without an interior presence, but there are a lot of factors that go into that as well. For example, last season the Bulls were better defensively when Joakim Noah was off the floor. That wasn’t because Noah wasn’t a good defender, but because the defensive combination of Omer Asik and Taj Gibson was better than Joakim Noah and Carlos Boozer collectively. There were plenty of other lineups that had the same characteristics as well.
By recognizing space in defending and finding a proper way to measure it through proximal field goal percentage. They measured where interior defenders were when players were going to the rim and they also measured how close players were to a player when he took a shot.
They broke the measurements up in three spatial distances. From three point range, from midrange, and from close range. Their final results showed that Larry Sanders–or LARRY SANDERS!–is actually the best interior defender in the NBA this season.
Kevin Garnett actually made the top 30 in most categories that Goldsberry and Weiss explored. Its very well worth a read. I’ll have some more information from the conference as I read more research papers. But it looks like we’re heading into an era of defensive analytics and I couldn’t be more excited.