Drafting for the Dwight Howard effect

Mark Zerof-US PRESSWIRE

Let's figure out how to play defense

I've been putting these thoughts together in my head for a while, but haven't been able to get them on digital paper until now.

First, let me qualify: I definitely want Pekovic to stay here. A lot of what I've said and what I'm about to say has/will sound like I don't like him much. That's not true.

What I will say is that, again, I think he's getting a $13.5-14mil offer from someone this summer and I don't think he's worth that much. Assuming Rubio gets a near-max contract (and believe me, he will. Cities like New York will pay him anything he wants just for name brand alone) that will effectively tie up $45mil+ in just three players long term. That's a lot of money, especially for a core three that hasn't even made the playoffs together yet.

In my mind...and I know full well that some of you disagree with this, and that's fine....but in my mind, you do not pay 3 guys $14-15mil apiece unless they give you a Duncan/Ginobili/Parker level of play. And as good as Love, Rubio and Pek are, it's extremely unlikely they will ever be as good together as Duncan, Ginobili, and Parker.

So that's where I'm coming from. Just to put it out there. Yeah.

ANYWAYS.

The Dwight Effect. It's a fantastic paper done by Kirk Goldsberry and Eric Weiss for the Sloan conference, in which they made a spectacular attempt at truly putting NBA defense in the realm of analytics. Their study looked at a few key aspects, then drew consistencies from those results where possible.

In regards to the Timberwolves....well, first, the bad news: 'Sota has been epically bad at defense the last few seasons, particularly in the paint.

I think this gets overshadowed by our struggles defending the three point line and our own bias towards believing that the strength of our team is our big men, but the bottom line is the bottom line: last season, the Wolves gave up 58% shooting to opponents within 10 feet of the hoop...second worst in the league. That ain't good. It creates a bad cycle on defense. Those shots are the easiest to make, and the more they make, the further in the defense feels it needs to be, which leaves more and more guys open from great beyond.

So the question is why are the Wolves so bad at this? Well, more bad news.

The Dwight Effect also looked at individual big men and their relative effectiveness at playing defense in close. The model Goldsberry settled on was what he called 'Proximity Defense', which the papers describes as:

Proximal FG%: the relative efficiencies of shooters in the proximity of the defender. Overall, when there is a qualifying interior defender within 5 feet of a shot attempt, the NBA shoots 45.6% from the field; however this value varies considerably depending on which defender that is.

So basically, how good is any given player at forcing a missed shot from someone he's right next to.

The best player at this turned out to be Larry Sanders, who gave up just 35%...a full 10% below the average Goldsberry and Weiss established for this scenario. Conversely, #50 on the list was David Lee, who gave up 53%.

The bad news for the Wolves is that Kevin Love came in at #48, giving up 52%. We could debate the rest of our lives why this is or if it's even accurate, but I personally think it does seem fair.

If you go through it all, there are...at least by eye test...oddities in the results. Andrea Bargnani came in #2 on the list with a ProxyFG% of 35.2%, clearly in defiance of his reputation as a terrible defender. I asked the Raptors guys and was told (and have since seen it put out in several other places) that that is in fact pretty accurate....Bargnani is really really good at forcing misses when he closes out. The problem simply is he almost never actually closes out.

Conversely, I think Love's habit of effectively giving up on defense early to box out drives his number up.

This finding led to an expanded dataset where Kirk and Eric looked at where on the floor this was happening. Using Larry Sanders and David Lee as the bookends again, they came up with this:

Kirk_interior_defense_medium

So as one would expect, people shoot better going straight at the hoop than coming at it sideways. The baseline defender and all that.

But clearly, the difference between Sanders and Lee is pretty drastic here. Lee is completely red (ProxyFG% of over 60%) inside the circle, while Sanders doesn't have a single cell of red on his whole chart. And considering that number of 60%...well, let's state the obvious: letting opponents make 6 out of every 10 shots from right under the hoop is not good.

And this leads to what I found to be the most interesting and useful part of the study, which they called 'Basket Proximity Condition', which basically boils down to 'how does the other team react to facing a very good (or very bad) interior defender?'

What they found was that the mere presence of a great interior defender....even if he's not directly involved in a particular play....drastically decreases the number of interior shots opponents even attempt. Just by being on the floor, a great paint defender forces opponents into taking more jumpers than normal.

The study found that the most effective player at this was Dwight Howard, which is how the study came to be named the Dwight Effect:

We found that the most deterrent interior defender in this sense was Dwight Howard. Overall, when a qualifying defender is within 5 feet of the basket, the NBA shoots 57.2% of its attempts close to the basket; however, when Dwight Howard was the interior defender this number dropped to 48.2% (Appendix 1A). This is what we call the "Dwight Effect" – the most effective way to defend close range shots is to prevent them from even happening. Although Howard does not lead the league in blocks, he does lead the league in "invisible blocks," which may prove to be markedly more significant. When Howard is protecting the basket, opponents shoot many fewer close range shots than average, and settle for many more mid-range shots, which are the least productive shots in the NBA.

S-n-P said several times that he believed the idea of defensive big men was dying out because of the changes in the way the game is played: far fewer traditional back-to-the-basket scorers, far more face-up, pick-and-roll guys. And now with the change in handchecking, the floor is wide open for anyone...many of the shots being taken near the hoop now are from perimeter players who use the new dynamic to get into the lane easy. We were a really good case study in this last season as we routinely got torched in the lanes by rookie Damien Lillard.

We also had a great example of this dynamic just the past few weeks, watching LeBron and Wade....two guys who are historically good at creating shots from in close...struggled to get those shots against Roy Hibbert and Tim Duncan. The Pacers, it could be argued, gave away Game 1 when they sat Roy Hibbert for the closing possession and watched LeBron...who had been forced into jumpers most of the game...go straight to the hoop for the winning layup.


The resulting sentiment was put into words by James right after, when he insinuated he made the decision to go to for the layup the instant he saw Hibbert sit. This was echoed by Paul George, who answered a question about why he overplayed the inbounds by saying he was setting up to contest a jumper because he forgot Hibbert wasn't there to protect the rim.

Contrast this with, say, Game 6, in which Hibbert played 42 minutes and the Pacers trounced the Heat by allowing just 22 points from the restricted are all game.

And this carries pretty consistently across the field with these types of players. Larry Sanders also ranked very high. So did Tyson Chandler, Marc Gasol....the reigning Defensive Player of the Year....was found to not only to be a tremendous shot deterrent (his 95.4 Defensive Efficiency rating ties Dikembe Mutombo for 4th all time among DPotY winners), but also that his presence on the floor generated an abnormally high number of steals for his team...a hallmark of the Grizzlies' defensive strategy.

This also isn't necessarily tied to simply being a good shotblocker. Serge Ibaka led the league in blocks (or did for the year this was studied, anyway) and is really quite bad. 3.6 blocks/game really comes out to a single block every 8 minutes. Ibaka is great at contesting shots in close, but not so great at intimidating opponents out of those shots in the first place (this is what I mean when I tell you guys when I say so-and-so is not good at defending 'in space', as in closing off the open areas of the floor) Kendrick Perkins, on the other hand, averaged just 1 block/game that season, but ranked in the top 5 in ProxyFG% and middled in Proximity Condition.

It's also not tied to just sheer size, like the names Howard, Hibbert and Gasol might imply. Also ranking very high are Joakim Noah, KG, Amir Johnson and even Nick Collison (which might explain his freakishly consistent +/- awesomeness)

So I wouldn't say that the idea of big men needing to be outstanding defenders is dying....it's just changing. The need to be great at stopping a post power like Shaq, Hakeem or David Robinson isn't necessary, no. But it is necessary to have a big man who great at those 'invisible blocks'. And this is why I keep repeating...often to the aggravation of you all...that the Wolves have to...have to....especially with Love and Pek, get a big man who can lock down the paint.

Again, we give up near 60% to opponents within 10 feet. 6 of every ten layups against us go in. We're a whole team of David Lees. We need someone to change that.

We need a guy who can be all like

Because what we have right now (in Pek) is more a guy who goes

Which, k'now, is great for the whole mafia strongman image, but on the basketball court...as Pekovic learned quite pointedly his first couple NBA seasons...it just results in a ton of foul calls.

SO. Where to find all this...

Well, there is the draft. Which is where everyone finds them, if you really think about it...

Ahaha....eeehhhh.....

Which anyway, the draft is always always a gamble, and even more so when it comes to defense, because it's pretty much impossible to statistically quantify playing defense at any level. Add onto that that many of the best defenders over the last 15 years or so have come from the high school ranks or Europe....KG, Jermaine O'Neal, Chandler, Howard, Marc Gasol....you're basically down to that blasphemous 'gut feeling'. Sometimes you get KG. Sometimes you get Kwame Brown.

Still, we can make a pretty educated guess based on what we do know.

Nerlens Noel was playing defense at a historic level before tearing his ACL, averaging not only 4.4 blocks, but 2 steals as well...the first time anyone in the college ranks has hit those defensive numbers since David Robinson's senior year in 1986. Noel showed tremendous instinct and timing, and has (assuming he returns to full health) the rare ability to get off the floor like a rocket. Very few guys can volley on the defensive end of the court. I mean, Noel contests shots at the same velocity Love tips rebounds to himself. Pretty much everyone I can think of that can do that has worked out in the NBA.

Plus, he has an epic high top.

Usp-ncaa-basketball_-kentucky-media-day-4_3_r536_c534_medium

I mean, c'mon man. He's probably like, 7'4" if you count the hair.

The other name in this draft that has genuine 'scare the bad guys into jump shots' potential is Rudy Gobert. The French center stands at 7'2" with a tremendous 7'8,5" wingspan (tied for 5th all time in DX's database), which gives him a towering 9'7" standing reach. That's a lot to try and get a shot off around.

It's always taken some serious extrapolation to look at Euro guys statistically, but John Hollinger took a pretty good shot at it a few years ago:

• Scoring rate decreases 25 percent
• Rebound rate increases by 18 percent (there are more missed shots in NBA play)
• Assist rate increases by 31 percent (Euro scorers are tightwads with assists)
• Shooting percentage drops by 12 percent
• Overall, PER drops by 30 percent.

It's not a straight 1-to-1 thing of course, but then again, neither is translating numbers from the NCAA either. His model worked nearly perfectly in retrospect looking at Bellineli, Scola and Juan Carlos Navarro, and he correctly called the futures of Gallinari, Pekovic, Casspi, Asik and Splitter. The only one he got wrong in his explanation/example was Nic Batum.

So throwing Gobert through the formula, you end up with something about:

PPG 8
RPG 7.6
APG 0.6
FG% 63
PER 14

His PER is high (for a projected rookie, anyways) because his Euro PER is REALLY high....over 20. I mean, the guy shot flippin' 72% from the field (and it's not all cleanup duty....70% from the FT line) Not bad at all for a 20 year old.

I don't know how this stuff translates beyond that...I'll leave it to vjl or Mad Dan to figure that out. But given the guy projects to shoot well over 50% on offense and has genuinely intimidating size and a good sample of effective work on defense (and French players have been quite successful in the NBA so far...Batum, Tony Parker, Boris Diaw....) I'd say Gobert is worth a long look if he's there in the 18-28 range.

Also of intrigue are Jeff Withey, who played outstanding defense for Kansas, and Lucas Noguiera, a Brazilian with likewise commanding stature and a high motor.

Beyond that, it's trades and free agency, and....well, let's just say that when teams find big guys who can really defend, they hang on to them. Howard is a max contact player. Hibbert is a max contract player. Noah, Gasol, Chandler, Garnett....all guys getting paid $12mil+. If you can snag one from a team looking to go in a different direction, then great. But don't count on it. By and large, teams need to get these guys in the draft and develop them themselves.

And this certainly isn't to say that we should be gunning all out for one of these guys Thursday. I wouldn't pick a Gobert of Withey over an Oladipo or KCP (Noel is a different story...) But if one of them can be picked up for like, Ridnour? Yeah, I say do that.

Defense has been a problem for this team ever since KG switched to a different shade of green, and this team has consistently and repeatedly overlooked opportunities to fix that. Time to fix that.

(To hype you up, I give you Dikembe Mutombo rejecting things)


X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join Canis Hoopus

You must be a member of Canis Hoopus to participate.

We have our own Community Guidelines at Canis Hoopus. You should read them.

Join Canis Hoopus

You must be a member of Canis Hoopus to participate.

We have our own Community Guidelines at Canis Hoopus. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9347_tracker