KZ Okpala and the application of physical tools

Often the most difficult challenge for prospects with strong physical tools is being able to connect them to the basketball court.  KZ Okpala, in at least one regard, has already done this.

Over the past few days, NBA draft twitter has put a specific focus on KZ Okpala. Most of it’s highly negative, in part because Okpala’s efficiency puts him in some unfortunately bad statistical company. While that’s likely something that would be mitigated by appropriate context, since Stanford’s spacing was exceptionally poor, even by college standards, and the statistical paradigms involved in forming those arguments are notably flawed to start with, it’s important to go back and actually put justification behind why Okpala is more than just his exceptional physical tools.

So let’s talk about one of the most direct applications of his physical tools, which is his ability to remain vertical while moving forward or backward at relatively quick speeds. Offensively, this represented itself in the form of his ability to get by defenders in driving situations while still staying upright enough to evade the defense by going to the other side of the rim, as seen here:

Okpala frequently used his first step to produce some impressive drives. Often they were drives into three defenders in a collapsed floor, but still. The more important setting in which this physical tool’s application was visible was his perimeter defense, in which he frequently played some of the most effective contain defense in this class.

And it is important to note the specific kind of defense he plays, because it changes the context of his steal and block numbers. When you’re comparing to other players that I’d describe as contain defenders, his numbers look outright normal. For example, compare his career 1.8 steal percentage and 1.7 block percentage to Michael Kidd-Gilchrist, who ended up at 1.9 percent and 2.8 percent, respectively. Torrey Craig, who played a significantly lower level of competition at USC-Upstate, ended up at 2.0 percent and 2.9 percent. Solomon Hill, though he had the peak of his career derailed by a hamstring tear, was at one point a very strong contain defender, and his career numbers at Arizona were 1.9 percent and 1.3 percent.

And part of that is that the pieces in question aren’t particularly indicative of good defense.

Nylon Calculus’s Andrew Johnson once showed that blocks are statistically insignificant at the NBA level at predicting defensive regularized adjusted plus-minus under a proper set of controls such that you aren’t subjecting your model to omitted variable bias. And part of that set of controls is field goal percentage allowed at the rim, something that’s fairly correlated to blocks, but there have been multiple players in recent years in the NBA that have demonstrated that it’s highly possible to have high blocks while being unsuccessful at protecting the rim, like Dwight Howard. The converse is also true. Derrick Favors actually led the league this year in percentage allowed at the rim with a minimum of 100 field goals contested, despite having the third-best block percentage on his own team.

Similarly, I suspect that steals become insignificant when you control for deflections but have not directly run that regression yet. The standard error on steals for this regression (Thanks to Drew Steele for running this for me while my laptop, where I’ve always coded from, is dead) is similar in magnitude to that of blocks, something we know to be the product of omitted variable bias. The items we know to be stable in value and not suffering from multicollinearity all have smaller standard errors, and those that we know have a fair bit of multicollinearity, like the shot types, all end up higher.

But more important than the question of the actual value of steals, since contain defenders tend to also have low deflection rates as well, is what the distribution of them looks like. I suspect that in that regard, steals look more like turnovers in that they likely have a bimodal distribution for what they say about player value. The details of turnovers can be found here, but functionally there’s a wide gap in the lost value to a dead ball turnover relative to a live ball turnover. Similarly, I suspect that players that gamble for steals produce less value with their steals than those that get steals while simultaneously playing sound defense. This seems to track with the number of visibly excellent defenders who carry low steal rates, and visibly poor defenders who carry some of the highest in the league.

Which is an excessively fancy way to say that blocks and steals are not particularly strong indicators for defense nor are they widely used as such at the next level.

But ultimately, what’s more key to understanding Okpala’s steal and block rates is looking at the actual film of how he defends, since that film most directly demonstrates that he is, in fact, the kind of contain defender for whom steal and block rates bias down relative to actual ability. In that film, his ability to keep his head and back upright while moving parallel to his defender’s line to the basket shines.

For example, in this play, he’s guarding Khalil Iverson in a sideline isolation situation.

Iverson goes baseline, and Okpala, despite being in his defensive stance, is able to deny him the spot and force the ball out of his hands because he’s able to stay relatively upright while retreating. This is physically difficult to execute for most players, but Okpala does it easily. Not only that but staying upright allows him to use his wingspan to deny any dump-offs to a cutter or context any runner.

But also, we can see him going in the opposite direction to contest a pull-up. Here, Kris Wilkes is isolated on him at the top of the key.

Wilkes pulls up because he perceives Okpala is giving him extra space. Okpala, however, demonstrates his ability to get quickly into and out of his stance and is into tight contest range by the time Wilkes is actually releasing.

But those are both plays on guys who are generally in the same area of height as he is. All the other contain defenders we’ve talked about are highly switchable and are able to contain different players of all archetypes. The good news is, though, that Okpala has that same ability.

For example, here, he’s in a sideline isolation defending David Crisp.

Crisp, at an optimistic 6-feet, is one of the quickest players in the PAC-12. Okpala denies him the opportunity to get to anything here. He retreats in a way that never allows Crisp to turn the corner, but also never gives him the space to get into the pull-up he’s willing to take.

And while I’ve been talking about these plays as a function of his ability to both change direction front to back and stay upright, he hardly lacks for lateral change of direction either, as you can see in this play where he completely denies Kris Wilkes a path into the paint.

And while those plays in this subsequence have focused on his ability to physically contain, now I’m going to move to a different kind of player for whom containment is more technical than physical, in Ethan Happ.

Happ is never going to out-athlete Okpala. Or much of anyone. What he tends to do instead is get defenders into compromised positions because they can only move so quickly and attack those. Because of Okpala’s ability to keep his back upright, however, Happ gains no breathing room. He doesn’t get a clear path to the paint, when he goes one way, Okpala is in position to contest, and when he fakes back the other, Okpala can contest there too. While this is a technical skill, it’s only made possible for Okpala because of his physical abilities. Because he’s upright, he can consistently react to the play as it happens and react quickly.

Further, because he’s already upright, he has an easier time with verticality, and that doesn’t just apply in the perimeter. He manages this effectively in the post too. Here, he’s posted by Khalil Iverson.

Iverson beats him on this play without a doubt. He gets him off his feet, gets into his body, and hits a relatively easy finish. And this wasn’t the only time a sequence like this played out — Okpala had at least two other times in the Wisconsin game alone where he bit on a pump fake, resulting in a weak contest, but stayed vertical to do no further damage. But what’s relevant here is that Okpala stays vertical such that even when he gets beaten there’s not a foul. Staying on your feet you can teach, especially for a prospect as relatively raw as Okpala. There’s a reason why staying on your feet on defense is referred to as staying disciplined, and that’s because it’s more about focus than it is about skill. That you can teach. The body control to stay vertical on that play, however, is not so easy to teach. It’s a physical skill that’s directly applicable to basketball.

Now let’s look at some commonalities across these plays, because we’ve been talking about differences. I’d contend there are three relatively important ones:

First, they all happen in Stanford’s man defense. Stanford used that man defense a fair bit, but also occasionally played a zone which was a true 2-3 rather than the more popular Syracuse matchup zone. That lead to a lot of Okpala guarding wood, to co-opt a term from football, and while that might have been by scheme, a large part of containing at the NBA level also includes making anticipatory rotations out of a zone-like setting such that it’s disappointing to see him, from the weak side of the zone, not rotate as readily onto the driver

Second, they all happen out of isolation plays. This was intentionally chosen on my end because Stanford’s pick-and-roll coverage was heavily push/ICE focused, as seen here.

This had varying levels of success at best, but also you can see why Okpala wasn’t able to showcase his ability to move front to back when he’s taking a scheme that prioritizes denying the screen at the cost of putting him off his line between the man and the basket over containing a single man like a switch coverage would.

Third, despite almost every play included containing good defense, and several of them generating turnovers, none of them produce a steal or a block for Okpala.

As a result, from the film, you can see that Okpala is a developing containment defender. While he is not yet flawless in this regard, he has already shown a clear ability to connect his exceptional physical tools to the basketball court, and defend in a specific way that the NBA is progressively making more and more use of. While he is still a raw prospect, and contrary to public conception still younger than much of his big wing competition, he’s not just a large ball of physical tools, and that should give him a leg up in the draft process.