18-14, 11-9 Big Ten, #30 KenPom
|TIME||Saturday, March 19
5:15 PM ET
|ANNOUNCERS||Ian Eagle (PBP)
Jim Spanarkel (analyst)
Jamie Erdahl (sideline)
|SPREAD||Sinners: Tennessee -6
KenPom: Tennessee -6
Torvik: Tennessee -5.4
I have to admit that I was surprised by people responding to my post about how stress-free Tennessee’s game was yesterday by saying Michigan was next and they’re already stressed. Okay? I mean it’s March: after a certain point, generally the first round, you are going to have to play and beat really good teams. Sometimes you have to do that in the Round of 64. Tennessee didn’t, so they do now.
Anyway, Michigan was a preseason top 5 team that was extremely disappointing for a couple of months, then just became normal disappointing after recovering to make the NCAA Tournament and win a game. This is not 9-16 Kentucky, just a roster with serious construction flaws that took a while to gel. They’re now top 30 in KenPom for a reason, which is that they are pretty talented and pretty good, particularly on offense. Still, Michigan has an insane run of play going: for 11 games, they’ve alternated wins and losses. That means this is due to be a loss, which would send Tennessee to the Sweet Sixteen. Maybe it will be that simple. Maybe I am Sisyphus. There is simply just one way to find out: playing the game.
Let it be known that this is the significantly more fearsome unit on the average night. Michigan ranks in the top 20 offenses per KenPom for a reason: they consistently get good twos, they have guys that make tough shots, and they possess a certain level of crazy. That crazy can get them in trouble (an 11-minute stretch against Indiana with no made field goals), but more frequently, it gives them life (such as coming back from 15 down against a 6 seed). The other thing is that Juwan Howard is genuinely a terrific offensive coach:
The problem is that Juwan the Coach may be ahead of Juwan the Roster Construction Guy. Michigan’s main problem, which shows up more on defense, is that there are a grand total of zero wings among rotation members. Caleb Houstan is a 4; Terrence Williams is a stretch 4; Kobe Bufkin is a 2. Therefore: no real wings, nothing resembling a JJJ type. Now you know.
The obvious #1 guy is Hunter Dickinson (18.4 PPG, 8.3 RPG), one of the best bigs in America. You may see this and think “but Tennessee shut down the best big in America, Oscar Tshiebwe, three different times. Also, Kentucky lost to Saint Peter’s, which is the funniest thing to ever happen.” All of that is true. But Dickinson is a bit more than Just Big: a terrific post scorer who is also an excellent jump-shooter. Synergy rates him in the 82nd-percentile in jump shooting among all players, not just bigs. This is why Dickinson is an elite player: there’s all the usual stuff in the post, but he can just as easily pop off for a mid-range jumper or, worse, a three. This is a guy who’s shooting 58% on mid-range twos right now and it feels unusual when he misses.
So: Dickinson is a bucket and a problem, and I didn’t even get to him being a good passer for a big. The problem is that Dickinson can be all Michigan has on certain nights, and even then, it isn’t always enough. Dickinson went for 20+ in 14 Michigan games this season; M only went 9-5 in those, including 6-4 against Big Ten competition. That’s still >.500, but generally, when your best player has a great game, you’d expect a better win rate. This is because the rest of Michigan’s team is largely coin-flips that either come up golden or goose-egg depending on the night.
Michigan’s #2 in scoring, but #4 in terms of on-court usage, is the efficient guard Eli Brooks (12.4 PPG). Brooks isn’t Michigan’s most talented player or shooter, but he’s their most consistent one at 38.6% on threes and, per Synergy, 38% on mid-range twos. Brooks also has a pretty silky floater he unloads that he’s converting 46.5% of the time, so watch out for that. Unless Michigan really needs him to take over, which doesn’t happen often, Brooks can be generally expected to take 8-12 shots a game and occasionally get to the paint. Just don’t let him get off an open catch-and-shoot three.
The other main shooter is Caleb Houstan (10.6 PPG), the closest thing Michigan has to a wing that’s actually a stretch 4. Houstan is hitting 36.6% of his 161 three-point attempts, but that doesn’t tell the story. Here is one hilarious split:
- Caleb Houstan, home games: 47.7% 3PT%
- Caleb Houstan, everywhere else: 29.2% 3PT%
Here is another.
- Caleb Houstan, 59% eFG% or better (13 games): 12-1 record for Michigan
- Caleb Houstan, <59% eFG% (21 games): 6-13 record for Michigan
Wait! One more.
- Caleb Houstan, open catch-and-shoot threes (per Synergy): 47.1% 3PT% on 68 attempts
- Caleb Houstan, every other three-point attempt: 29% on 93 attempts
How a freshman shooter who only shoots well at home is this predictive for Michigan I don’t know, but that’s the case. If Tennessee forces Houstan into threes where he’s off-balance or literally has to move at all, it’s a win. If you leave him open, it’s bad news.
Lastly: Devante’ Jones (10.7 PPG/4.6 RPG/4.6 APG). Jones is the true point guard on this Michigan roster and has developed into a very good Big Ten starter after having a very slow start to the season. Jones is a relatively infrequent jump-shooter; he took 90 across 31 games, and most of them were threes (33.8% 3PT%). Jones’ main value add is as a passer and paint scorer. The problem for Michigan is that Jones may not play; he missed the Round of 64 Colorado State win due to a concussion. Howard seems optimistic Jones could be back for this game, but at the time of writing, nothing had been confirmed in one direction or the other.
The swing piece here is Moussa Diabate (9 PPG/6 RPG), a hilariously athletic center and poor shooter (8-for-31, per Synergy) who nonetheless will do 1-3 things a game where you’re like “that’s a first-round pick.” Diabate is 79-for-100 at the rim and 30-for-99 everywhere else; you cannot let him get open off of Michigan’s numerous pick-and-rolls. Others: Terrance Williams III shot 38% on threes and Tom Izzo got mad about it. Brandon Johns Jr. began the season as a starter but now plays 10 MPG or so. Kobe Bufkin: 16-for-26 at the rim, 14-for-53 everywhere else. Frankie Collins started in place of Jones against Colorado State and was fantastic, but prior to that, his season high was 8 points and he still has more turnovers than assists.
Here’s where I’d argue the lack of a true wing is most impactful. Last year’s Michigan roster had top-10 pick Franz Wagner, Isaiah Livers, and Chaundee Brown, three terrific players who could guard numerous positions and stay with guards. This year’s roster has Houstan, who was atrocious the first two months of the year on D, and…uh…well…
Here is the problem. At shooting guard is Eli Brooks, a 6’1″ player. Brooks is a gritty little dude who makes a ton of fun plays and hustles hard every time out. Brooks is also 6’1″ and regularly guarding 6’3″-6’6″ guys. This is a problem when you are playing 6’3″-6’6″ guys that can shoot well. This is a key reason why Michigan ranked in the 6th-percentile in defending off-the-dribble jumpers and 21st-percentile overall. Some of this is obviously luck-based, but some of this is just “your personnel cannot defend our personnel.”
The Michigan season can be defined by its hilarious 11-game win-loss-win-loss flips, sure, but it’s also more or less about a lack of wings that means having to panic into a zone mid-season (roughly 9% of the time, per Synergy, but more like 15-20% since January) to save the tournament streak. This worked out, and I credit Howard for doing it. Michigan did a terrific job of getting away with zone defenses against quality offenses by simply guarding them very well during the season; it is weird to see a 71/29 Guarded/Unguarded split in a zone but 67/33 in man. The point is that Michigan often has guys in place to force tougher shots. It just hasn’t always worked out.
Just like anyone else, Michigan has been wide-open to variance. The Wolverines are 4-8 when a team shoots 35% or better from deep against them and 14-6 when they don’t. Michigan is pretty good at running shooters off the line, and the fact that they’ve forced 20 games where opponents have shot worse than 35% is nice to see. Still: there will be some openings.
The problem with running shooters off the line with this roster is that it opens up a lot of driving lanes. Refer back to the lack of a consistent ball-stopper; refer back to how their two best defenders are a 7’1″ center that isn’t very fast and a 6’1″ point guard that may or may not play. If you get this Michigan team rushing at you on the perimeter, there will be lanes to the rim. That’s how a team with Dickinson can be giving up a 64.1% FG% at the rim, per Torvik, which is tied for fourth-worst among the 68 Tournament teams.
Michigan is also a very poor post defense team, but I think Tennessee’s odds of exploiting that are a lot better by drawing Dickinson/Diabate out of the paint than by playing bully-ball. Colorado State tried bully-ball and it worked when David Roddy was hitting crazy fadeaways; everything else was torture chamber stuff. I mean, watch this play: an extremely similar one to one Tennessee ran for John Fulkerson yesterday. Feels like basket cuts, once again, could come into play here.
Again: 78th-ranked defense in the nation that never forces turnovers, ranks 206th in 2PT% allowed, and has just the one somewhat-consistent rim protector who isn’t a foul machine. (Diabate is very promising, but 4.5 fouls per 40 leads him to play less minutes than you’d like.) Michigan went 0-12 in games where they gave up 1.09 PPP or greater, 4-13 when the opponent posted an eFG% better than 50%, and 4-8 when the opponent made 8+ threes.
How Tennessee matches up
The first instinct here is that Michigan ranks in the 23rd-percentile in ball-screen defense, Tennessee has Kennedy Chandler/Zakai Zeigler, and you think “okay, do that.” And that would probably work out just fine. Michigan has deployed a variety of coverages to try and make things work this year, but they’ve most often settled on putting Dickinson in drop coverage because of the mobility concerns. (I would also mention that Dickinson has played with nagging injuries over the back half of the season and still looks awesome despite it.)
Tennessee has not run a ton of ball screens this year or really ever under Rick Barnes, but it’s been effective when they have. Synergy has Tennessee’s ball-screen offense in the 68th-percentile; that number is obviously not adjusted for schedule, but take that number with the Michigan one and you can see the potential advantage. If Dickinson comes out to hedge or double, Tennessee will be playing 4-on-3; if Dickinson plays drop coverage, a good shooter like Chandler will have an open shot.
Along with that: gotta get stuff going at the rim, simple as. Michigan’s rim FG% allowed is horrible, of course, but being 8th-percentile in post-up defense is alarming no matter what type of schedule you’re playing. That being said, Tennessee’s post-up efficiency is similarly brutal, so I’d prefer to not see many of them. Instead, utilize Michigan’s perimeter aggressiveness to your advantage: when they rush to close out a shooter, drive to the basket instead. There, you can either go up for a layup or continue the power-play by finding someone open in the dunker spot. Either can work.
Defensively, this is first and foremost about slowing down Dickinson. You can’t shut him down – only three times this year did he fail to at least 11 points – but you can slow him down somewhat. Tennessee has to find a way to force Dickinson to his left, where he’s less comfortable. Per Synergy, Dickinson had 162 single-coverage post-up possessions this season. The 23 times he faced up were pretty obvious: most were jumpers. The other 139 are pretty interesting: 112 times turning to his right (his strong side), just 27 to his left. Dickinson was still very effective going to his left, but he’s not nearly as confident in it. If you can force him left, maybe you force a worse game? It’s not purely like last year where Dickinson could only go right but he’s still learning.
Beyond that, you’ve just got to force the toughest threes you can and hope the shooting variance gods smile on you Saturday. I did a study in late January of the Big Six teams to see just how much variance a team had in 3PT% from game to game; Michigan led the field with ease. At the time, anything from 19% to 54% from deep was within reason. That trend has more or less held for them: 10-3 when they shoot 35% or better from deep, 7-11 when they don’t. Or, if you prefer: 8-2 when shooting 40% or better, 6-4 when shooting 30-39%, 4-8 when shooting worse than 30%. I think Tennessee can survive Michigan shooting 35% or something. You have to do whatever you can to ensure that’s not 40% or better.
Starters + rotations
Three things to watch for
- Which averages hold? Tennessee versus top 40 offenses: 69.8 PPG allowed. Tennessee versus defenses ranked 40th-100th: 74.7 PPG scored. Michigan versus top 40 offenses: 74.4 PPG allowed. Michigan versus top 40 defenses: 68.6 PPG. Taken all together, that would suggest a 75-69 Tennessee win. But I know better than to trust anything at all in March.
- Shooting percentages. I mean if Tennessee posts their season average of 36.5% or better from deep, this should be a win. Tennessee is 18-1 when they shoot 35% or better from three. Of course, this is reliant on Michigan – a team averaging a 33-34% hit rate – not having an RNG game in their favor.
- Turnover margin. This may be where the game swings Tennessee’s way. Michigan ranks 336th in defensive TO%; Tennessee ranks 15th. Tennessee projects to have a +5 advantage in TOs, which would be tough to overcome for a Michigan team in need of every shot they can get.
Hunter Dickinson vs. Center Roulette. Tennessee has narrowed down their center rotation to Plavsic/Fulkerson/Huntley-Hatfield but frankly, this could be an Aidoo game. Dickinson will need someone to challenge him vertically as best as humanly possible. You’re also rooting for Dickinson to be relatively inefficient on his jumpers.
Eli Brooks vs. Santiago Vescovi. Brooks is the Gritty Gritstein (h/t MGoBlog) of Michigan: a guy who makes all the little plays and seemingly never stops moving. Brooks can reasonably keep up with Vescovi for a while, so it’s on Tennessee to set good screens to get Vescovi open.
If Devante’ Jones is playing: Devante’ Jones vs. Kennedy Chandler.
If Devante’ Jones is out: Caleb Houstan vs. Josiah-Jordan James. My guess is that Jones plays. In that case, Chandler has to be prepared to be hounded by Jones for all 40 minutes. You can fool Jones into touch fouls from time to time, but mostly, he’s just a pest. Any time you can force him into taking jumpers you have to. Houstan, meanwhile, is somewhat simpler: just make him move.
- Tennessee records a +5 or better advantage in the turnover department;
- Michigan has a run of 2-3 made threes followed by six straight misses;
- Tennessee 73, Michigan 67.