The-Power wrote:Are we sure the defense is led by Mobley to a degree that we can ascribe their #1 defense primarily to him? After all, the defense is even better with him on the bench and the on-off numbers for the entire Cavs roster are pretty steady. To me, it seems like it's been a team-wide effort. Of course Mobley is a key factor here, but so is Allen and possibly some under-appreciated role players in addition to just a great defensive culture. If we want to honor one Cavs player in particular: what makes Mobley so much more important than Allen? And that's a genuine question, not me implying that Mobley doesn't have a strong case.
This is going to be a long reply and it also functions as my argument for Mobley being the DPOY.
Mobley suffers from really bad opponent 3P shooting luck that's nuking his defensive on-off metrics. Opposing teams shot 38.5% from three with him on the court but only 32.8% from three with him on the bench, so they go from shooting like the #2 ranked Warriors to the #30 ranked Rockets. And that causes an enormous swing. The fact that, in spite of that near 6% 3P disparity and while considering the Cavs still have a very good defensive big in Allen always on the floor when Mobley is off it, the Cavs were
only 1.03 points per 100 possessions worse with Mobley on the floor this season is actually a testament to just how great he is on the defensive end, funny enough. It's been shown that opponent 3P shooting is highly variable and takes like 160 games (don't quote me because I can't find my source but it's close to that number) to stabilize and I think that might actually be at the team level too, so it would take even longer for an individual player. But I can break down even further why it's clearly just really bad luck and not something Mobley is doing wrong.
For one, the Cavs allow the least number of threes per game in the league and the 2nd least number of threes per 100 possessions. So they already did a great job of actually limiting three point attempts. When Mobley was on the court, opposing teams had a 3PAr of .377 yet when he was off the court, it was only .341. Understandable since when you have a rim protector out there, the opposing team will naturally take more threes. The number of mid range attempts by opponents stayed virtually the same with Mobley on versus off. But with Mobley on, teams attempted 30.3% of shots at the rim, shooting 61% on those attempts, and with him off, they attempted 33.6% of shots at the rim, shooting 66% on those attempts. And so they were trading those rim attempts for more threes. And yet Mobley also led the league in contested 3P shots this year. It's not a perfect metric, but this is where it's good to supplement with the eye test. I don't see Mobley doing anything bad on a fundamental level that's causing his contests to not be effective.
So when considering those stats above, with Mobley on the court, teams are attempting about 33.8 threes per 100 possessions, and with him out, it's about 30.4 threes per 100 possessions. If just for the sake of argument we assume that it's true that individual defenders have a negligible impact on opponent 3P% and that that Mobley's disparity is entirely noise, and we then adjust the opponent 3P% so that it's the same constant 36.8% (Cavs' opponent 3P% on the season) with Mobley on versus off, that changes the Cavs' DRTG from being 1 point
worse with Mobley on the floor to being almost 5.4 points
better with Mobley on the floor. And again, considering Allen is the backup, that's really great.
So it really comes down to whether you believe Mobley's opponent 3P% variance is his fault or not. Considering last year, the disparity was only 1.6%, with opponents still shooting better from three with Mobley on but the difference is much smaller. And taking into account that about 35% of Mobley's total minutes off the court this season came with Kevin Love at that 4 position instead, in which opponents shot only 32% from three, and Love doesn't even bother to leave the paint to contest outside shots at this stage in his career. It's clear to me from watching it's nothing he's doing wrong and he's just an unfortunate victim of statistical variance, while guys like Love, Cedi, and Wade benefitted.
And so depending on how much they incorporate the lineup stats and whether or not they adjust for those unstable factors, it's not a surprise the impact metrics are all over the board for Mobley. And for the ones, that incorporate a lot of box-score stuff, Mobley's not exactly a stat-stuffer either. He ranks 1st in Defensive RPM, 1st in Defensive Win Shares, 4th in Luck-Adjusted D-RAPM, 4th in D-LEBRON, 12th in D-DRIP, 16th in DBPM, 32nd in D-RAPM, 56th in Def RAPTOR, 75th in Def EPM, and 91st in Def DPM. The catch-all metrics and lineup stats are good starting points for any analysis, but in some cases like this, they can miss a whole lot.
I think Mobley's argument resides on being able to plug so many holes for a very flawed defensive roster en route to the #1 defense in the league (and by whole 0.7 points), while leading the team in minutes by far. How many people prior to the season would've projected that? I saw most arguing it wouldn't work with too 6'1 weaker defenders in the backcourt, and then you have Okoro (who's probably closer to 6'4 than his 6'5 listing) manning the SF spot, making for a very tiny collection of defenders on the perimeter. Okoro, while a good defender, isn't exactly providing much in the form of help defense, steals, rebounds, or switchability either. He's a limited 1-on-1 stopper of guys who are similar in size.
And I don't want to take anything away from Allen who's an awesome defender in his own right. Or deny that guys like Mitchell and Garland and Caris have been putting in great effort on the defensive end. Also, JB has definitely instilled a culture that gets guys playing very hard for him. But it's important to keep in mind the Cavs in the 2020-21 season were the 25th ranked defense. And actually after acquiring Allen a couple months in the season, the Cavs were 28th in DRTG in the stretch of their final 58 games. And that team had:
Garland, Allen, Okoro, Cedi, Love, Wade, Stevens, and Love were all still rotation players for that roster. And JB was still the coach. Ignoring Mobley, the other major guys we have on this team now are Mitchell, LeVert, and Rubio. Again, while LeVert and Mitchell are certainly putting in a much better effort now, you have to also realize they're coming from being in the 5th percentile and the 1st percentile respectively in the three-year Def RAPM dataset (2019-22). And then Rubio is obviously good defensively, but he played just 566 minutes this year and I think post-injury he's a step slower than he was last year. When you look at the bigger picture, it's clear Mobley is the biggest factor that led to the turnaround. If he were to miss extended time this season instead of playing 79 games, which would've better shown how much the Cavs miss his presence, and if the opponent 3P shooting luck weren't so slanted against him, that would be more reflected in the lineup data and the impact metrics. But anyone who's studied him closely can see just how many advantages he's closing off every possession. It's almost every trip down the floor, one of Garland or Mitchell is dying on the screen at the point of attack and it leaves Mobley having to cover a ton of ground on the court, hedging and recovering, switching, guarding the help-side, and protecting the rim. He just covers and threatens so much space horizontally and his positioning is always on point. He knows which lesser efficiency shots he needs to concede to the defense, which typically takes bigs longer in the league to learn. He uses verticality well and keeps blocks in bounds. He's switchable onto virtually anyone in the league besides a few like Embiid and Jokic who are just too big for him. Allen is awesome and I'd say he's slightly better as a post defender, but outside of that, I honestly think Mobley is better at every other aspect of defense.
Also, Mobley has manned that center spot without Allen very effectively this season, which was a huge critique of him last year. In 13 games without Allen in which Mobley played, the Cavs had a 109.6 DRTG which would rank 1st in the league above their already 1st place standing as a team. Furthermore, in 1253 Mobley at center minutes on the whole season, the Cavs have a 110.3 DRTG. Remove old man Love from the equation, and those 873 Mobley at C minutes have a 108.6 DRTG. With the added strength and learned knowledge, he's done just as great a job as the backline of the defense this season after only really being optimal in that help defender role last year. This video by Ben Taylor kinda showcases some of that where Mobley completely dominated defensively in a game against the Jazz without Allen. It also mentions more about those misleading on-off numbers.
Finally, one last thing I want to mention is that impact metrics and lineup stats can't accurately capture what a player means to his team in terms of the lineups he allows a team to actually put on the floor. Mobley's versatility allows the Cavs to play a variety of defensive lineups and in different coverages, catered to the opposition, and he can seamlessly switch between that 4 and 5 spot, with Allen at center next to him or with guys like Wade/Stevens/Cedi at the 4. You compare him to a guy like Lopez on the other hand who's stuck at the 5 spot and moreover has to be in that drop coverage, which he's awesome at by the way. But it's context worth considering when he has Jrue Holiday and Jevon Carter blowing through screens at the point of attack like they're nothing and Giannis on the help side shutting off passing valves and threatening with rim help. Additionally, when opposing teams offer a lineup that's more mobile and spaced, the Bucks will go with Giannis or Portis at center. Models that take lineup data into account will look at that and say, see how much the team misses Brook as the defense gets worse? But that's not entirely accurate because there's value in being able to be on the court in those situations against those types of lineups at all, and that's what Mobley offers, which partially explains why he played 4 more mpg than Brook. If a team goes small, JB will put Mobley at center surrounded by a bunch of small guards and small wings, he can also be a 4 in a traditional twin towers lineup, and he'll even be deployed up top in a zone. It's that kind of effectiveness in being the key to countering virtually any style of offense that has allowed this team of oddly-fitting parts, most of whom have never carried reputations as good defenders, to be so successful on that end.
Besides Brook, JJJ is the other obvious candidate. I've said I think on a per-minute basis, he definitely is the most impactful defender in the NBA. But I just can't get on board with awarding a guy who's played 45% of his team's total minutes this season, so he gets docked for that. Draymond has been great but I think he's performed at a lower level than last year and isn't quite at DPOY standard. Adebayo and Claxton are in the conversation, but I don't prefer them to any of the others I've mentioned.
So yeah, I'm biased obviously, but I think Mobley does deserve DPOY this season. If he doesn't get it though, he'll obviously have plenty more opportunities of course. And he can still be the youngest winner in the history of the award if he manages to win it next year, which would be a cool honor. Either way, I'm very excited that he's already at what I consider to be DPOY level at 21, and I hope added mass can make him generational.
