SPONSOR NOTE: Upon Further Review is sponsored by HomeSure Lending and Matt Demorest. Rates are the lowest they've been in three years so it can't hurt to check whether you can save money on a refinance. Or you could buy a house in Ann Arbor! Good luck with that!

Matt's relocated the bus to Pioneer this year, BTW, and invites everyone to stop by and say hi. There's beer. I mean, obviously. Matt. Matt and beer: a good pairing.

FORMATION NOTES: Another close to 50-50 split between one-high and two-high looks, which was generally successful at keeping Clifford off balance except when KJ Hamler was wide open downfield.

Otherwise, I hope you like 3WR 1 TE sets.

SUBSTITUTION NOTES: Hutchinson, Paye, and Kemp got every snap that wasn't the jetpack rush package; Paye and Hutchinson are out there for that as well so they never came off the field. Danna only got jetpack snaps; Dwumfour got about a third of the game; Uche got ~two-thirds.

Hudson/McGrone/Glasgow with no rotation at LB. Hawkins and Metellus got every snap; Lavert Hill's return meant the CBs rotated, but there was a clear preference for Hill and Thomas over Gray. Dax Hill only got jetpack snaps.

[After THE JUMP: shots shots shots shots]

283 yards!

28 points

283 yards!

I kind of thought this was a bad performance? Or at least a meh one? 

What?

When Seth compared Clifford to Shea Patterson that was pretty accurate. He hit a few deep shots and then submarined a lot of the offense with bad passes, clean pockets exited, and a few different runs on which he either went the wrong way or should have pulled. It is abundantly clear why the PSU offense has been so stop-start all year: they have some crazy skill position guys and the guy pulling the trigger is a sophomore who's still growing into the role.

So a lot of things looked like PSU errors rather than Michgian forcing things. Both the pressure and coverage metrics, which are intended to give an overall picture of how comfortable the idea of passing is, checked in moderately negative. That hasn't happened in a long time. Michigan got a little pressure here and there but not much of it came from the DL. The starting ends both checked in with nice days largely gathered in run defense, and then the rest of the crew was meh:

This was about six inches from being a 35-point day for the defense based largely on deep shots when Michigan couldn't get to the QB and couldn't cover a post.

The DEs, who were Michigan's best performers by a long shot, had the kind of day where I barely clip anything and they keep racking up +0.5 or +1 by being reliable and getting a little push to make two yard runs happen. Paye did make a nice play to rescue a run that could have gone badly:

Their numbers do not include 16 team pass pro minuses, about 13 of which I didn't assign to individual players. That's probably another –2 or –3 for not  getting rush wins, and that brings them down to just above Just A Guy Territory. I was expecting a little more against an OL that looked a little wobbly going in.

I still don't see how it's possible Paye is the #2 P5 player in the country at pass rush win rate. I had basically nothing as a rusher from him in this game. He's good as a team rusher who doesn't get out of lanes and can be useful on stunts, but he's not an edge guy at all.

I wonder if you confirmed or dis-confirmed your wonderings about that Hamler touchdown?

As a reminder: I posited in the game column that the multiple guys running free downfield were a cost of Michigan neutralizing drag routes. I think that bore out. It was the crossing routes that eventually got Michigan even though they got covered. For review, PSU exits the field on their first drive because Michigan swaps coverage on Hamler's crossing route:

This is mesh with a couple of fade routes down the sideline so there's no real reason for a safety to be in the deep middle, and indeed there isn't one:

Metellus is moving up to the LOS, as is Hawkins.

Flip a fade to a post, though, and trouble beckons. PSU just missed another long TD before halftime:

Metellus starts that play as the deep S, and by the time the ball gets launched he's a yard from the LOS covering Hamler on a drag:

Hawkins is tooling around about 15 yards downfield in what seems like an obvious bust, after both guys were talking and pointing on the snap.

Hamler's touchdown was a bust, per Harbaugh, but also if a guy doesn't get the call his natural thought is probably to play a deep post safety; instead Hawkins creeps up to the line of scrimmage, reacting to nobody.

S #20 to top

There are no underneath routes:

That's a guy operating on instinct because this is Yet Another Tempo Bust, like Dwumfour going straight upfield. His instinct right now is to go find a drag route instead of playing the deep middle. So there was one deep shot that seemed explicitly about M addressing drag routes with their safeties, and a second where the bust seemed related.

So we got RAHNEd on?

I mean… I dunno. RPS came out even and there were a number of plays on which Michigan had PSU downloaded. Here's Hudson running under the RPO slant he can't see but knows is behind him:

LB #7 to bottom

Clifford has to scramble when Hudson delays his decision-making process even a little because OL will be getting downfield.

Michigan did get lost on the opening snap, as the TE moving away from the play befuddled both linebackers:

That's a blast from the past: this play is commonly known as "power read" among coaches because it's power with a read—huh—but was popularized by Chris Brown as "inverted veer" back when TCU invented it and was a major feature of the Denard years. It's fallen out of vogue—can't remember the last time I saw an opponent use it—but returned here, with that tweak.

It was less successful later, but that was in part because of Clifford's wobbliness. He has an opportunity here with McGrone flying out to contain the back, but he goes into a wad of bodies instead:

This goes back to the thing I was saying about Clifford being a source of unforced PSU errors that generally kept the grading down even on successful plays.

McGrone had a more normal day.

It was all in about two drives, the bits that stood out. So that's still a little weird.

McGrone had a six play sequence in which he was the guy I put in the "player" column—used to designate the player most relevant to the success or failure of the D on a particular play—five times. Time #1 was just red meat for a sideline to sideline LB:

MLB #44

PSU punts; next play is a run up the middle on which McGrone is able to power through an OL block and jam the play up at the LOS:

MLB #44

Next play he gets a free run that he can't quite tackle on:

MLB #44

This was one of only a couple times Michigan schemed up a free guy; note that presnap Uche backs out and then Glasgow threatens before backing out; feint within a feint.

He did pay off the free run later:

McGrone wasn't given the opportunity to run under any blocks and didn't. I didn't think he had much to do with the long run.

Well, then, who did?

The big run was mostly a thing that's bitten Michigan from time to time over the past few years: Khaleke Hudson acting as an ILB. Dwumfour is able to burrow into a gap on the backside and occupies two guys for the whole play; nobody blocks Hudson. Hudson puts himself in the wrong gap:

If both OL are stuck at the LOS you have to be able to get over to the hole the RB is in. Maybe if it's a gap or two over you miss a tackle or whatever, but you have to get there. Hudson doesn't even get a tackle attempt in. Then you have no safety because Metellus is containing the QB and Hawkins is shaded to the slot because Hamler. Turns out having really fast guys who are good at football is helpful.

That's also a big gap ceded by Kemp, who is not able to keep his other OL off of McGrone and still gets put definitively to one side by a guy who's able to shove him a couple of yards laterally.

But mostly it's Hudson, who plays in the slot and at deep safety and as an overhang defender and has never and will never get sufficient reps to be a reliable ILB; it's just too much to process too fast if you don't go full bore with it.

This secondary has a lot of explaining to do.

Uh… yeah. Each one of the four starters was directly responsible for a swing play and they didn't do a whole lot to get it back. Note that Thomas's swing was not the actual one with yards. I underestimated how close Thomas got to the fade completion:

I mean, jeez:

You wish he'd been able to push that route to the sideline a bit more but on anything other than a dime that hits the WR in-stride he's got a play. Freshman Jourdan Lewis-esque.

Thomas's big error was getting beat to the post on the near-TD mentioned above.

Meanwhile, we should probably talk about Lavert Hill grabbing guys when it is absolutely not necessary. He got hit with a flag against Army, when a WR who plays for Army ran a fly route:

Here he wrapped both arms around Justin Shorter, a true freshman who had not been targeted all day, to draw a critical holding call that kept PSU on the field so they could score their fourth touchdown:

A fair bit of Hill's grabbiness is handfighting or generally warranted—he took a holding call against Iowa in a bad situation—but sometimes he just can't help himself. I wonder if that's why he's not as attractive to the NFL as you might expect.

And the safeties?

Metellus was actually okay. He was not really at fault on the Hamler TD—he's playing outside leverage expecting a bracket that's not there—but that frustrating third and eight QB run that got PSU down to the eight on their third TD was Metellus just whiffing on a not very mobile QB:

That should be fourth and three and is instead first and goal.

Hawkins… was not good. In addition to Yet Another Tempo Bust Hawkins demonstrated considerable limitations trying to stay on the same planet as Hamler, not only on the TD but also a later chunk play where he ends up getting passed by Glasgow because he bites on an inside jog from Hamler:

Note that Dax Hill was about a fingernail away from getting a PBU there.

Honestly I'm not even that mad about the TD, which sees Michigan go pure cover zero and asks Hawkins to cover the whole world.

Just asking him to do something he can't. Silly to do this and not flip Thomas or Lavert Hill—not even on the field on this play—into the slot. But that thing with Isaih Pacheco outrunning him last year turns out to be a realistic depiction of his speed. Michigan can't get away with the stuff they want to get away with because Dax Hill is not operational.

Are we still mad about Chris Hinton not playing much?

Uh… no.

#15 DT to bottom

Freshman DTs suck. Maybe next year.

Are we still mad about the officials?

Hell yes.

The Hudson OPI incident only gets worse once you look at it a lot, because Hudson is in the guy's back pocket and the ball hits him in the facemask:

Hudson gets the shove just as he's getting his head around for a ball he's in better position to catch than the TE. For that to go uncalled when Michigan gets a garbage OPI on Collins would be tough to take if we had not reached a stage of serene acceptance about the meaningless of fleeting organic life in an unfathomably vast and ancient universe.

Heroes?

McGrone and the starting DEs, although nobody really had an All-American level game.

Maybe not so heroic?

Hawkins. The pass rush as a unit.

What does it mean for Notre Dame and beyond?

TEMPO. Another week, another failure to get lined up and run something sensible. This time it leads directly to a 53-yard TD and the winning points. Will this get fixed? Not likely.

Hawkins is slow. KJ Hamler is not.

Michigan's DEs are both SDEs. They're good against the run and push pockets closed; nobody's getting around at eight yards. Even Hutchinson's ability to flash pass rush has mostly been on inside moves.

Defending drags has popped open holes on deep routes and Michigan is going to have to adapt again. Some of this is fixable by being in the right D.

Hill could be a little less grabby. What it says on the tin.