therealbig3 wrote:The RAPM which says Dirk supposedly got way better on defense, to me, says something that mysticbb has touched on before. His defensive RAPM looks good not because he himself is an amazing defender, but because he can carry an offense to the point that he allows the Mavs to play strong defensive lineups around him. For example, as I've mentioned before, with Kidd/Terry/Marion/Dirk/Chandler (only one self-creator outside of Dirk in Terry), the Mavs posted a 125.57 ORating (+20.01) with a 90.70 DRating (-19.92) in the 2011 playoffs...so they were a +34.87 team with that lineup on the court in the playoffs.
So it may seem like his defensive RAPM is getting better, while his offense has remained constant, so what gives? If he improved his offense, why is his offense constant while the defense is better? I think it's because of the different lineups that are used with him now. They're not using Nash/Finley with him anymore, they're using past their prime version of Kidd and Marion (or in 09, J. Howard), one decent ball handler and creator in Terry, and a defensive minded center, such as Haywood or Dampier or Chandler. That is really not strong offensive support at all, so I can understand why his offensive RAPM hasn't really improved, even though he has personally become a better offensive player. I can also see why his defensive RAPM has vastly improved, even though he hasn't become a better defensive player.
So imo, the overall impact that RAPM says he's having is there, it's just that the splits between offense and defense aren't being properly assigned, or are being affected by the types of lineups that Dallas has mainly used around Nowitzki.
And his production in 09 was really quite strong actually. Compare his per 36 RS and PS numbers in 09 to 06:
06 RS: 25.1 ppg, 8.5 rpg, 2.6 apg, 1.8 TOpg, 58.9% TS
06 PS: 22.7 ppg, 9.8 rpg, 2.5 apg, 1.8 TOpg, 59.6% TS
09 RS: 24.7 ppg, 8.0 rpg, 2.3 apg, 1.9 TOpg, 56.4% TS
09 PS: 24.5 ppg, 9.2 rpg, 2.8 apg, 2.1 TOpg, 63.5% TS
His RAPM was a little lower in 09, but his +/- for the PS that year was +14.37. Now, I know they got torched by the Nuggets in the playoffs that year, so I wouldn't expect Dirk's defensive numbers to look great, since he plays most of the same minutes as Melo and Billups. But look at the offense of the two most common lineups that the Mavs played in the playoffs that year:
Kidd/Terry/Howard/Dirk/Dampier = 113.14 ORating
Kidd/Barea/Howard/Dirk/Dampier = 130.77 ORating
This is more quality thought.
The idea that a player's defensive +/- might look good because his offense allows better defensive guys on the court is something that's been around since people started talking about +/-, but it's never gotten well developed so far as I know because of lack of evidence. By that I mean, a player might seem to be showing that in one year, but it doesn't maintain in any ongoing fashion even when that player's role seems to stay the same when we used advanced metrics. So so far as I was concerned, it was noise, but I'm always open to new scenarios.
Let me also say that although we're at this point emphasizing the defensive shift, the overall shift was more dramatic than that and included offense as well. As mentioned, Dirk doubled his deviation in one year. That's a huge leap forward to the point where I think we're a bit beyond saying "Dirk's gotten so good that they can shift lineups around him for defense".
I'll revisit what I said before, when I said, "Either Dirk got better, the team got a better fit around Dirk, or something weirder happened." To me if a player isn't getting better, but his RAPM is showing an abrupt change, I'm not sure what else to call that but "fit" in the broad sense that includes team strategy. If someone has a different way to explain it, then I'd like to hear that. Obviously "fit" is a bit of a charged word for good reason. It tends to imply that there's some baseline level that's "real", and superior fit adds a bonus to that.
As a rule, I don't subscribe to that. To my mind, the baseline is more on the high end of things regarding fit, because I often times see scenarios where a player is simply being misused. However, I do consider it a fair point that when we compare players with different degrees of "fit" on their teams, it doesn't make sense to act like their both at the baseline if we're talking about how "good" they are as opposed to the more concrete measurement of "value".
I don't contest Dirk's value to these teams, however when we are trying to judge the goodness of these players I get nervous when I point to a +/- statistics that I can't put into more concrete terms. And Dirk's numbers going through the roof so late in his career makes me think that the Mavs have hit the fit jackpot around Dirk in a way that I don't know if it's ever been done before. Can you think of another player who in his 13th year on a team, 11th year in prime, and after what seemed like a pretty graceful arch, all of a sudden saw his value explode like this?
And this is where I start feeling the need to tip my cap to the Mavs. A team who has been building around this star for most of that 11 years tinkering with different approaches, and all the while being the inventors of modern +/- analysis, if anyone was going to see shocking breakthroughs in the +/- of their star despite no glaring improvement from him, wouldn't it be the Mavs?
):
