Glasswork: A deep dive into the New York Knicks’ rebounders
Just how much do individual rebound numbers matter? Are rebounds more of a team game than an individual one? How do the Knicks’ bigs rank out? Prez takes a deep dive.
The GOAT of Knicks beat writers, Chris Herring, has moved onto the luminous digi-pages of Sports Illustrated. One of his latest is a look into the paradox of Andre Drummond’s rebounding and Dre’s lack of paying suitors (he was unceremoniously bought out of the final $29 million year of his deal with the Cleveland Cavaliers). The central question is actually as much about the modern NBA writ-large as it is about Drummond himself: does rebounding matter in the modern NBA?
Dre is a rebound machine, leading the league in glasswork for years now. Herring’s article focuses mostly on offensive rebounds, where league-wide, teams have trended downward from nabbing 33% of misses 10 years ago to around 25% now. That’s due to the increasingly potent transition offenses brought about by the 3-point revolution, and by the proliferation of pull-up 3-point shooters, especially. Those same threes have also caused more rebounds to bounce long, erratic, and uncontested — further away from the mosh pit of the paint.
But Dre is no slouch on the defensive glass either… surely that has a place in today’s NBA? The short answer: kind of. Team rebounding is much more important than individual rebounding. It’s team rebounding that has always been a hallmark of playoff success. Pat Riley said “no rebounds, no rings,” and although not as correlated with winning as offensive or defensive efficiency, boards remain important.
The key nuance here: on an individual level, heavy duty rebounding numbers matter only as much as they help your team secure rebounds. If all a player does is gobble up easy rebounds while still sucking at securing the hard-fought ones, are they really helping your team? Or are they stat-padding?
If your whole team clears out and runs up the floor on uncontested rebounds because they know one player is going to both get the rebound and initiate offense (looking at you, Luka and Russ), then those uncontested rebounds — which are a lot of them, 55% of all rebounds in the modern NBA — are easy, and don’t really improve your chance of winning. From a team perspective, it doesn’t matter if Luka gets the board or if Maxi Kleber does, if no one from the opposing team is around.
Alright smartass Prez, how the hell are we supposed to measure a player’s contribution to team-wide rebounding then if not with how many rebounds they get?
Why, I’m glad you asked! Take a step back from the basketball card mindset and think of all the things that happen leading up to a missed shot and a rebound: someone usually has played solid defense; two, or three, or even four or five players are defending their assignments 25-plus feet away from the hoop; hopefully some of those five players stop “playing defense” and begin boxing out… and then the ball becomes available for a rebound. If the rebound path is predictable and there aren’t defenders nearby, someone can plan to go get it with little trouble. If it lands in a mosh pit, a player might have to claw and elbow and leap and stretch to snatch it out of the air.
Look at everything that goes into a successful rebound! This is why looking at a stat that assigns credit for all of that shit to one person is useless. If the best volume rebounder of the last 10 years has a mediocre impact on his team’s rebounding, why should we give a shit about him averaging 15 rebounds a game? It’s all about the W at the end of the day, and if him hogging rebounds doesn’t help towards that, why should we glorify that number?
Instead, as friend of the site @Tim_NBA implies above, it’s better to look at whether a player gets contested rebounds or how often they box out.
Another useful question: does the team rebound better with the player on the court, or worse? The Lopez brothers famously sucked at rebounding but were tremendous at boxing out, and their teams were good rebounding teams consistently because of it. This was before we could easily track box outs and track team REB% on/off numbers. But even with that data we need context: what happens if both Lopez bros play together? After all, if they both box out and contribute to their team rebounding better, the team won’t rebound better or worse with one on the court vs. the other. This is why it’s also important to look at who else is doing the rebounding/boxing out on a given team for context, and why we will look at all of the Knicks’ bigs, and not just Mitchell Robinson or Julius Randle.
OK PREZ, we are TEN-PLUS PARAGRAPHS IN with LITTLE TO NO WORDS on how good the Knicks are or are not at rebounding. Please get to the juicy stuff. This is The Strickland, not your WordPress rap music blog from 2005. We need an intervention. Don’t you have editors?
Harsh, but fair, dear readers. LAY THE BLAME ON ME AND LEAVE THE EDITORS OUT OF IT!
Now that we have the context thoroughly exhausted, here is what we will explore for a granular look at how well Randle, the depressingly foot-fractured Mitch, and Nerlens Noel hit the glass:
Total Knicks team rebound percentage
Rebounding on/off numbers
Box outs per game and per 36 minutes
Contested rebound percentage
Rebound chances per 36 minutes
Team Rebounding
The Knicks are 10th in the league in REB% at 51.3% .For reference, the best rebounding team is the Pelicans, at 53%. The top 10 contains some contenders: the Jazz and Sixers are second- and third-best at rebounding, and the Bucks and Nuggets are behind them, with the Lakers and Clippers also in the top 10.
For the Knicks, the high ranking makes sense intuitively when you think about their roster: their top seven minutes guys are good for rebounding. Elf, for all his flaws, hits the glass. IQ sneakily does too. Bullock and Burks are both unafraid to mix it up and 6-7 rebound games from the 2 spot for them are not uncommon. RJ has been a plus rebounding wing since his time at Duke. Julius has been a plus rebounder all of his life. And Mitch, who came into the season weak at rebounding, has — as we will explore below — turned this weakness into a legitimate strength this season.
Rebounding On/Off Numbers
Julius Randle:
Knicks’ DREB% with Randle on: 77%. With Randle off: 72%. Five percent may not seem like much, but it’s actually quite a bit. Steven Adams and Rudy Gobert create a 1% difference; Jokic creates a 2% difference; Bam creates a 5% difference. To be fair, it probably also helps that Bam and Randle’s backups are like… Kelly Olynk and Obi Toppin, LOL. The point is, anything more than a few percentage points is pretty legit.
Knicks’ OREB% with Randle on: 27%. With Randle off: 24%. Part of that probably has to do with Randle’s own proclivity for missing at the rim and getting his own board.
Mitchell Robinson:
DREB% on: 79%. Off: 75%. Remember when Mitch’s rebounding was a big question mark? His differential was negative last year.
OREB% on: 30%. Off: 24%. We know Mitch is an elite OREB gawd, so this makes sense.
Nerlens Noel:
DREB% on: 73%. Off: 79%. Rebounding isn’t Noel’s strong suit, but this is also impacted by him sitting while Mitch plays. He had a similar dynamic in OKC with Steven Adams, another great rebounder. In Dallas prior to that, his numbers were more muted — his team was about even on the glass with him on and off.
OREB% on: 25%. Off: 27%.
Box Outs
Mitchell Robinson leads the NBA (!!!) in box outs per game at 5.1 in his light 28 minutes per. That would be about 6.6 per 36, which would be a very large lead over the second-place Bam Adebayo and his 5.2 per 36. Remember when rebounding was a big question mark for Mitch?
Julius Randle has two per game in 37 mpg.. His being lower makes a little sense, as he’s away from the hoop much more than the true centers, and also because he’s definitely the designated rebound collector more than he is a box out guy. Still a solid number, though.
Nerlens has 2.1 per game, but in 22 minutes per game, so about 3.4 per 36. Good!
For good measure, OG Taj Gibson averages 3.3 per game in 20 MPG, or 5.9 per 36. Taj is good at boring stuff, confirmed!
These numbers may seem low and trivial, but consider that a given game has 40-50 rebounds, about 10-12 which are probably offensive rebounds. So 30-40 defensive rebounds, and about half are uncontested. So there are approximately 20-30 “up for grabs” rebounds a game, and New York’s bigs have to either box out, go for the ball, or both on all of those when they are near the ball. You want your bigs to be able to box out, and also just go for the fucking ball like a savage, and to balance that as a team — you can’t have everyone box out and no one actually get the ball, and you can’t have everyone go for the ball and no one box out.
Contested Rebounding/Rebound Chances
Julius Randle:
3.4 contested rebounds per 36, contested rebound percentage of 31%, 17 rebound chances per 36.
What I get from this: Julius gets most of the uncontested rebound chances — which we know based on the eye test. The 32% figure on contested boards is pretty good for a power forward. Julius is at 32% on those, Steven Adams is 35 (very good), Zion 25, AD mid-20s, Sabonis 29, Turner 35… just for comparison.
Mitchell Robinson:
Seven contested rebounds per 36, contested rebound % of 65% (!!!), 18 rebound chances per 36.
TL;DR his insane athleticism, length, and instincts create contested rebound chances. He also fights hard for them. What a specimen. What a marvel. I hope his agent doesn’t use this piece in negotiations.
Nerlens Noel:
4.1 contested rebounds per 36, contested rebound % of 42%, 12.7 rebound chances per game in 22 minutes (!).
I suspect Nerlens’ high rebound chances number comes from his playing with the bench unit so much, which has no designated uncontested rebounder snatcher, so he becomes it by default.
Taj (for fun):
4.3 contested rebounds per 36, contested rebound % of 48%, 15 rebound chances per 36. TAJ STRONG!
So what have we learned?
Well, I think Thibs would be the first one to hoarsely harp on how rebounding, like defense, is less about the flashy 1-on-1 sequences or volume stats and more about the stuff that doesn’t make it onto box scores. The communication, the box outs, the physicality. The team comes first, and the players who embody that don’t necessarily get rewarded with fantasy stats.
We now have enough knowledge, both of the data and eye test varieties, to show that we have to be very careful when deploying rebounds per game and even REB% as useful stats for individual players. They have some shorthand utility, but also some risk, and can cause narrative to diverge from fact. Better to use those as team stats.
I propose we drop the descriptor term “plus rebounder” and exchange it for the more useful “plus rebounding impact player.” The Knicks employ only one real “plus rebounder” as traditionally understood (Julius) but at least two-plus “rebounding impact” guys in Julius and Mitch, and a bunch of players who would be considered solid rebounders for their positions. Put simply: our team is good at rebounding because they play many guys who are good rebounders or rebound impact guys.
In the future, I am sure someone smarter than me can create some kind of awesome weighted metric incorporating box outs and impact on team rebounding relative to positional norms and controlling for lineups, but I ain’t the guy for that.
(But if you happen to be that person, please get in touch with me ASAP tho.)
On a more Knicks-specific note, I think we have enough data to conclude Mitch is not only not a weak link on the glass, but REALLY, REALLY IMPACTFUL at helping our team rebound even if he doesn’t vacuum boards like Joel Embiid or Nikola Jokic. His ability to box guys out a TON, and fight for contested DREBs at an ELITE clip, and be an elite offensive rebound threat make him one of the most impactful glass forces in the NBA this year!
Out of curiosity, I asked Tim of BBall Index if their metrics lined up with the observations I made, and sure enough, they did. See that last stat, Real Adjusted DREB Rate? That’s basically impact on team rebounding, and incorporates box outs, rebound chances and rebound conversions, contested rebound rate, and some other tracking info. An A rating. Say it with me: ELITE REBOUNDING IMPACT CENTER MITCHELL ROBINSON.
Obviously this is all the more bittersweet knowing that our young phenom is out for the season. Get well soon, Meech! I hope he comes back in good health next year, and can continue joining forces with Coach Thibs to spread the gospel of boring box outs and team rebounding. Until then, may we never cite individual rebounds per game or REB% again, and bask in our newfound enlightenment.