Signals Against Iron Curtains: A DPOY Deep Dive

Change what you think you know about defense.

Photo by Tim Mossholder on Unsplash

Over at the Whiteboard, I talked about how guards deserve credit for their defense and should be in the DPOY conversation more. I wanted to expand on that a bit here, with some data.

I talked about this with MVP defense stuff, but yes, defensive metrics are tricky. What most people don’t know is how they’re tricky. They just want to say, “Oh, you can’t trust them,” usually because they don’t like what they find.

Defensive rating is the easiest to tear down because it’s a team metric. Keon Ellis doesn’t have a defensive rating; the Kings have a defensive rating when he’s on the floor. There’s a signal there, but it’s going to be noisy.

Bad defenders can have better on-court defensive ratings because they’re protected in a little cocoon of scheme and rim protection; good defenders can have bad on-court defensive ratings because they do their job, but the rest of their team can’t.

For me, it’s useful in evaluating the “what is.” You can argue a player’s impact on that number, but you can’t argue they’re making it worse, and you can prove that a team can have a good (or bad) defensive rating with them on the floor.

Synergy Data is suspect because it’s manually logged. Workers go through and log the plays and that involves human error. There are also guidelines. There are plays you’ll find where a player starts by guarding a ballhandler, switches off, goes to the corner to guard his other assignment, and gets credit for it. The aggregate of Synergy data I absolutely trust for being directional. It points you in the right direction.

Defensive EPM via DunksAndThrees.com, Defensive Real Adjusted Plus Minus, and BPM all factor in as well. I’ve reached the point of ignoring Defensive Box Plus-Minus for the most part, because I feel it over-indexes rebounding. Nikola Jokic is neither as bad of a defender as his detractors suggest nor as good as the metrics indicate.

But each of these points in direction.

A popular line of dismissal is that if a player who is not a good defender appears on these lists, then the entire metric is flawed and must be thrown out. There are going to be outliers in a data-only assessment, which is why you should never use a data-only assessment. If you want to evaluate a player’s defense accurately, you have to do the boring work of watching a lot of games and/or clips.

What I’m always trying to get to in any analysis is a triangulation of eye test, data, and stats. (Data being impact metrics, stats being box score.)

I’ll come back to the film side of things in March for DPOY talk.

For now, I wanted to share this chart that provides players in the 75th percentile of both total possessions defended and points per possession allowed via Synergy, along with their On-Court Defensive Rating and Defensive EPM.

The On-Court Defensive rating is heatmapped (green/yellow is good, purple/black is bad) so you can tell which players are playing individually great but their impact on the team’s defense is limited. This is the Anthony Davis case. AD’s individual box score stuff is terrific, and opponents are averaging the fourth-lowest points per possession when he’s involved. But the Lakers give up 1.14 points with him on the floor.

That number doesn’t prove AD isn’t a good defender, that would be lunacy. But whatever he’s doing? It’s not doing enough to raise the defense to an elite level. You can blame the teammates, and that’s fine. But if we are going to reward the best defender in the NBA, shouldn’t it be someone whose team, you know, defends well when they play?

(Maybe you don’t think so and think this should be in a vacuum, in which case, Victor Wembanyama is your DPOY because he leads in rebounds plus blocks plus steals.)

You can sort the columns to see the differences.

Takeaways:

• I’ve said this on Twitter, but Shai has a real case for DPOY. No, he’s not the best defender on his team (and that’s a pickle I don’t think I can get around), but he is defending the most possessions on the best defense in the league (by far). He’s got great help from everywhere, but opponents are also shooting 31% from the field against him and he’s third in steals per game.

• Notice how if you sort by on-court defensive rating, all the top five is all guards? And yet guards have no shot at DPOY? We are in an era where point of attack and closeout defense is the most important factor in whether you win or lose, where teams use smallball lineups successfully consistently, where no one posts up except Jok, Sabonis, and Embiid, and yet we’re still married to bigs for DPOY only. Make it make sense. The guys who are making the most difference outside of the very top tier of bigs (Wemby, Gobert, and Trip) are guards. It’s not that a big man shouldn’t win DPOY, it’s that a guard should be able to win DPOY.

• What Kris Dunn is doing for the Clippers is flat out incredible and he deserves a. a larger contract and b. a lot more credit. That was a sensational pick up by their front office. Kris is elite.

• Amen Thompson is already this good and impactful. As a second-year player. Coming off the bench. As a guard. The sky is the absolute limit for him and I think he too has a serious DPOY case. If you go through Houston and how it defends, you can’t point to Sengun’s rim protection or Dillon’s disruption (though both have been good). But Amen? Amen WRECKS teams for stretches of games.

• Can I once again say that much like Hartenstein’s metrics in his various stops were good, people don’t seem to understand the impact Goga Bitazde has? Goga’s not some all-world defender, but boy, it sure seems like he’s helping and not hurting in an area that should be his weakness.

• Speaking of iHart, he’s going to wind up on this list once he catches up in minutes and games. He’s unlikely to win DPOY because he’s already missed 15 of the 18 allowed. But just as Chet was DPOY before his injury, Hartenstein might be DPOY since.

• Damian Lillard’s on this chart and that seems ridiculous. I get it. The takeaway in context is that the Bucks have done a good job covering for him, and Dame’s given the requisite effort. This isn’t a chart of elite defenders. It’s a chart of guys where the defensive results have been good. Understanding how to contextualize it is important.

• How are the Nuggets so bad defensively? How? (Eye test does not match metrics for Nikola Jokic and not completely for MPJ though the effort has been there.)

• Who’s DPOY right now? I would have a guard. But if we’re going by the same way the award has been voted on, it’s Jaren Jackson Jr. Better defensive rating than Wemby for a better defensive team and better opponent scoring against him.