Defense is boring. Defense is ugly. Most people don’t pay attention to defense. Actually, most stats don’t pay attention to defense, either.
The defensive component of Expected Value does a decent job though, incorporating all forced turnovers (not just steals), defensive errors, field goal percentage against and defensive rebounds. We can use that as a basis to ballpark the best defenders in the league this year — apologies to the players on the last place teams, who failed to accrue enough possessions in the sample. (See the stat-tracking page for further details.)
Below are the leaders in Defensive EV from the 2011 season, by position. Included for reference are the number of defensive possessions they were involved in for the Defensive EV sample, and courtesy of 82games, their opponent PER, opponent eFG% and opponent Free throw attempts (FTA) at their primary position. Also from 82games is the On/Off difference in their team’s defensive rating. (Negative scores are good in that category.)
Stats in green from 82games (through April 6, 2011)
Rajon Rondo looks the best outside of EV, but remember none of the other metrics in that table would penalize him for missing steal opportunities. He’s 16th in my regular season sample in blow by’s per 100 possessions, trailing only Ty Lawson, Tony Parker and Derek Fisher in that category among point guards. Brandon Jenning’s sample is small, although the other numbers suggest that he’s a good defensive point guard. Jason Kidd echoes similar numbers from the 2010 playoffs, even at his advanced age. For those wondering, future MVP Derrick Rose (12.7 opponent PER) has a 0.6 Defensive EV.
Stats in green from 82games (through April 6, 2011)
Ronnie Brewer is a certifiable pest. Opponents are shooting 20% (8-40) against Brewer. Mike Miller has also been surprisingly effective on the ball for Miami, holding opponents to just 31.0% shooting(18-57). Dwyane Wade looks solid across the board, with his usual all-around defensive contributions and large defensive role (21.2 Defensive Usage, 4th in the sample among SG’s).
Stats in green from 82games (through April 6, 2011)
I’m including Andre Iguodala, despite a small sample size, because of his deserved defensive hype this year. In only a few 76er games that I’ve tracked, Iggy has guarded incredibly well; he rarely fouls and is rarely scored on. LeBron James has been a defensive beast this year, boasting a ridiculous 26.7 defensive usage for Miami’s 5th-ranked defense while holding opponents to 30.6% shooting (30-97). Luol Deng has been excellent for Chicago too, and while Grant Hill looks good, it’s hard to make a legitimate case for him over the guys ahead of him.
Stats in green from 82games (through April 6, 2011)
The largest impact defenders in the league are usually the bigs — they typically carry the largest burden — and Kevin Garnett‘s numbers here reflect his reputation. Conversely, what is DeJuan Blair doing there? Opponents shoot a high percentage against him (48.2%), so there’s a good reason he was replaced in San Antonio’s starting lineup. At least he’s a big steals/blocks/rebounds player who rarely ever fouls (3.5 free throws against per 100 possessions)…
And what to make of Gerald Wallace? All but one of Wallace’s games in my sample is from Portland, where he’s primarily playing power forward (he played more small forward in Charlotte). I’m not sold either way, as I see a really good rebounder, someone who can both block shots, take charges (~1 every 100 possessions) and generate steals (2.0 per game). Portland’s defensive rating with him has improved only slightly, so I’m not entirely sold on him being that good. Still, he made the 2010 all-defensive team for a reason…
Stats in green from 82games (through April 6, 2011)
Marcin Gortat has had a good run of games, with opponents shooting just 32.8% (29-89) against him. He’s clearly helped Phoenix’s defense, moving them from bottom of the league to average in his time there. He’s also drawing offensive fouls at nearly 1 per 100 possessions (6th among qualifying players) and 17th in blocks per 36 minutes. Orlando never saw a huge dropoff from Dwight Howard to Gortat, which suggests he could anchor an elite defense, although he’s never had to play huge minutes. Gorat (2.3 free throws/100) and Howard (2.9 free throws/100) foul the least of any centers in the league.
The other names are big defenders by reputation: Andrew Bynum has been a monster across the board, along with Andrew Bogut. DeAndre Jordan has show incredible promise, and the league has a hard time shooting against him, shooting 30.0% (44-137) while DJ is 4th in blocks per 36. Chicago’s Omer Asik also deserves a mention here, posting wonderful numbers in reserve in all of these categories.
Overall, here are the top players by Defensive EV:
Green stats from 82games; D RAPM is Defensive Regularized Adjusted plus-minus from stats-for-the-nba.appspot.com/
EDIT: I’ve added a D RAPM to the above table from here.
FWIW, my all-defensive team would be:
1st
PG Kidd
SG Brewer
SF Iguodala
PF Garnett
C Howard
2nd
PG Rondo
SG Wade
SF James
PF Wallace
C Bynum
Hard to see anyone else in the 2-3-5 spots…would be pretty bad if Bynum didn’t make one of the teams.
What was Dirk Nowitzki’s Kobe Bryant’s and Carmelo Anthony’s defensive EV
ElGee, you have the image of the chart saved in .TIFF format, which most browsers can’t see. Can you convert it to PNG or JPG, please?
+1 on this. I’ve got to click on each image (and have it take me to a new screen) before I see it.
Beyond that, fascinating stuff. Curious why you don’t address Howard’s good but not great showing. Saving that for a DPOY post?
OK fixed. Thanks for the catch guys.
re: Howard. No DPOY post. My impression is that people think his defense is slightly below where it was a year ago. That said, Bynum is really the only center I think with a viable case against him, as DH still looks pretty darn good across the board. Most stats are teammate dependent to a degree, and considering who Orlando has, Howard’s counterpart numbers are more impressive to me…
That said, it’s a fair point Matt. His D *hasn’t* been as dominant (he killed DEV in the 2010 playoffs.) Are you suggesting someone like Garnett for DPOY?
If I just looked at that chart, I’d be hard pressed to think why it shouldn’t be Garnett.
Also compelling in his favor are his team’s DRtgs in with him (97.84) versus without (104.47).
Howard’s with (101.95) vs without (104.98) is not quite so compelling.
For the record, the more I think about it, the more I’d probably vote for Garnett for DPOY. The on/off stuff can be noisy and it’s unadjusted for opponent strength…but what is compelling is that KG has nearly a decade of ridiculous defensive +/- data. (He’s one of the all-time greats at defensive play-calling/communication.)
Furthermore, Dwight’s counterpart data is always against a center. Garnett will guard some centers/athletic centers like Stoudmire, and all the top 4′s in the league. Which is a tougher batch of players?
Centers: http://tinyurl.com/3vztsy4
Power Forwards (roughly): http://tinyurl.com/3lns7ck
Maybe I should write a post on this…
KG as DPOY is exactly the question I’d have. I’d like to hear your take on it.
Also though, I’d also love to see you wax loquacious on the nature of basketball’s “middle linebackers”, perhaps with Garnett as your focal point.
Haha get a room, you two.
I suppose it is kind of creepy to use the word “love” twice followed by “loquacious”.
Howard’s defense is starting to reach Kobe status when it comes to perception vs the story that +/- tells. It makes them interesting test cases. They both put up huge numbers in the traditional box score stats that most pay attention to when making their judgments…Howard’s combo of blocks/steals/rebounds are certainly the best these days. Likewise, both have a lot of team success to bolster (or perhaps generate?) said public opinion. And both have generally played a lot of minutes, making it difficult to argue that the team (or in Howard’s case the team defense) might not suffer so much without them.
The thing is, in Howard’s case, the evidence is slowly starting to build up for me. The biggest case that I’ve heard for him as the DPoY, and the one that I think will eventually win him the day, is that his teammates are known to be poor defenders so for the team to be 3rd in defensive rating it must be due to his defensive impact. And that certainly is an intuitive position. But we now have years of on/off +/- data suggesting that the team doesn’t fall off so much defensively when he’s not on the court. He used to never miss a game, suggesting that this was purely a situational effect…but he’s had to miss a handful in the last couple of years, and as far as I know the team defense hasn’t missed much without him. Then, the story was that it was Marcin Gortat’s unheralded defense behind him that kept the team steady without Howard. But Gortat’s gone, and neither the defensive +/- nor the team defense when Howard’s out have reflected any change.
I don’t know. I’ve accepted that Howard is going to win his third straight DPoY by another landslide this year. And I consider him a great defender. But I do wonder if his defensive impact isn’t a bit exaggerated.
That’s a bit exaggerated about Kobe and +/- — Howard’s 3 year RAPM on defense here (http://stats-for-the-nba.appspot.com/4-year-ranking) is 3.6. Kobe’s is 0.4.
I think the Van Gundy’s are fantastic defensive coaches and that should be considered. Of course, Marcin Gortat was Howard’s backup, which allowed them to continue to play a similar scheme with a big roaming shot-blocker. But I see Howard as having big lift on D, but not in the sense that Orlando would fall apart on that end without him.
The last game without Howard they were torched (Chicago) so I’m not sure what you’re referencing regarding his injuries. His on/off this year is 102 –> 105 DRtg…that doesn’t seem significant as a ballpark indicator for you?
Yeah, I would say that Howard’s definitely been a good defender this year. A -3 on/off differential is nothing to sneeze at.
Just nowhere near what KG has been this year.
Essentially, my point is similar to what aversiveness wrote. I’m not saying that Howard is a no-impact defender, just like I would never say that Kobe’s not a great player. My point is, both are often credited as being the BEST at what they do despite a yearly progression of stats (that I tend to believe useful) repeatedly saying that they aren’t the best.
So yeah, I buy that Howard makes them better on defense. But what I’m not buying as much is that his defensive impact jives with the notion that he is hands-down without peer the best defensive player in the NBA (as his soon to be third consecutive DPoY emphatically states).
Fantastic work!
My only observation would be that Kidd spends the majority of his time defending shooting guards at this point. Terry, Bubois (Sp?) and Barea do most of the PG defense per 48.
Surprised to see Blair ahead of Duncan.
Thanks Jason. Blair is weird — I’d suggest sample but people are scoring on him easily. So his defensive value is coming from the other areas I mention, and those seem more consistent over time. Quite an interesting case…
Kidd does defend a lot of 2′s. His counterpart data at 82games is: 16.0 PER/.502 eFG%
Also I have KG as DPoY. The data from the last four years suggests that his presence as well as his ability changes the dynamic of how well his whole team defends.
I suspect Wallace didn’t effect Portland’s defensive rating as much as you might expect because with playing PF aldridge was moved to C bumping camby out of the staring lineup. Any thoughts?
Interesting thought. One thing I’ll give to Wallace is how versatile he is as a defender. Then again, if he plays next to a traditional center, is he capable of having a huge impact alongside him? I’m not sure…
I have been looking for something like this. I’m one of those rare basketball fans that really enjoy the defensive side of the ball (at least that’s what I consider myself as everybody the media/peers seem much less interested in defense). But I am interested to know where Tony Allen of the Grizzlies fits in these statistics.. He definitely passes the eye test for defense when I watch him play, and his steal numbers are great also.
I could never be a gambler because my math skills are so bad I can’t ever understand any of this stuff.
tony allen is one of, if not the best perimeter defender in the nba iyam. he’s just a sick combo of incredible lateral quickness and general athleticism, length, wirey strength and most of all effort. never dies on screens, always contests the jumper. iirc last year he led the league in blks+stls per 48 in crunch time (82games.com)
Jennings the best defensive PG? Something is flawed in your formula.
Hard to dispute anything in the report. Very well done! But there has to be something said for the ability to guard multiple positions. I have seen Shawn Marion guard Kobe, Rajon Rondo, Lebron and Durant in a ten day span and he did a good job on each one! Marion can guard every position except center, but I realize that its hard to quantify that sort of thing with numbers.