Upon Further Review: Defense vs Western Michigan

What in the hell is this? It's "Upon Further Review," MGoBlog's comprehensive, numbers-heavy breakdown of Michigan's previous game. It takes until Wednesday/Thursday because it's a lot of work.

A note on video: changes to the torrent format killed the thing I'd been using to cut clips out, unfortunately, and I haven't found a solution yet. I'm really trying to get this squared away and will add them to the post as soon as possible. For now, no video.

UPDATE! Video good to go thanks to askarpo.

A note before we start: attempted to assimilate the offseason knowledge I picked up about over, under, and 30 fronts and actually pick out which shift Michigan was in before every play. This did not go well, so take the below with something of a grain of salt. I do think I got the "Base 4-3" right but I was marking 30 fronts—two gap fronts akin to an NFL 3-4—4-3 under fronts most of the game. And I'm still a little hazy. This, I believe, is the 4-3 under we've talked about all offseason:

Roh is a stand-up guy outside the weakside tackle. RVB is in a 3 technique over the weakside guard. Martin is the NT, shaded slightly to the strongside of the center (a 1-tech). And Graham is in a 7 outside the strongside tackle. The linebackers have shifted to the strongside.

This is basically the same thing with Roh's hand in the dirt against an I-form tight formation; Michigan brings Mike Williams up for an eight-man front on short yardage:

And this is the under again against a balanced formation; you can see that Graham has taken a considerably more outside stance and the guard is "bubbled" over Ezeh, which means he can just run out and block him:

As for platooning: there was none. Michigan spent the entire game in a 4-3 if you count Roh a DE and Brown a safety. If you want to call it a 3.5-3-4.5 I wouldn't look at you funny. At least, not too funny. They did rotate in backups on the defensive line after the first few drives.

GERG!

Wait just a second, there's another caller on the line.

What do you make of the performance of the Michigan defense against Hiller and WMU?  My initial observation was that they just shut down a future NFL quarterback and that's a great sign for the rest of the year.  Upon further contemplation I got to wondering how much of a factor Hiller's health was.  He had off season knee surgery and when other players have ligament operations we tend to hear it takes them a full year to be completely confident in it again.  Is this the same for quarterbacks and if so, was this a contributing factor in Hiller's poor performance? 

Thanks,
Andy Heck

Hiller was not as advertised. He yakety-saxed a couple balls, airmailed a few others, and was considerably less accurate than Forcier. But the Michigan defense had something to do with that. They got a considerable amount of pressure for such a dink-and-dunk offense and usually covered Hiller's first read unless they were intentionally playing soft. And they tackled much better. It's a lot easier to see this in a—

Chart?

Chart.

So there you go: big plus days in both the metrics, and if I'd remembered I wanted to add "tackling" this year I can tell you that tackling would have been hugely positive as well. When you're looking at the numbers, keep in mind that a large section of the minuses are directly attributable to JT Floyd, which says a lot about Michigan's corner depth but not much about the starting eleven, and that Michigan got pretty soft late.

What about Brown at his new position?

As Greg—

GERG.

As GERG said, possibly apocryphally, he's "a hell of a lot better player there." Stripped of the requirement to find cosines and the like, Brown was free to be a super-athletic linebacker who's good in man and good at short zone drops without being a touchdown magnet. He's a modern OLB. We've yet to see if he can hold up against big pounders but with Notre Dame missing its starting fullback and starting a wisp of a pass-receiving TE, that potential weakness won't get tested just yet.

Speaking of, initial GERG review?

His defense makes a lot more sense than Shafer's even if it's almost the same formation. Michigan never got away from its base set but that's a lot less infuriating when the guy you're running out on third and long is Brown, not Johnny Thompson, and you've got the flexibility provided by a deathbacker. Michigan never, ever went to a three-man line except on obvious passing downs, never found themselves on the wrong end of a hugely speculative playcall, and rarely found themselves uncertain of what to do before the snap. GERG simplified the defense, made it flexible enough to function against spread teams without getting out of base, and put his players in positions to do what they do well.

Thumbs up after game one; game two will be a much stiffer test.

So Warren's day was… interesting. Argh Michael Floyd?

Short of being totally awesome in all ways I thought Warren's day was as encouraging as it could be given the three penalties and a couple plays that came at his expense. His game looked like that montage in a superhero movie where the hero performs a slapstick routine of smashing cars, punching through walls, and burning innocent pedestrians to death before he gets a handle on his newfound powers. Warren was hyper-aggressive in his first game free from bone chip soup in his ankle; the results were mixed-to-encouraging.

Plays marked "Warren" above:

• Busts up long route with bump; has better position than the receiver.
• Good position on a third and four slant that was fired too high and hard.
• Thumps ball loose on a hitch. (Or, at least, helps a receiver who was already dropping the ball finish dropping the ball.)
• Leaves two hitches open on Michigan's soft pre-half drive.
• Immediate tackle on hitch.
• Running a guy's route for him and going to be in position to intercept when receiver trips him (not in a penalty sort of way).
• Running almost inside a guy's jersey and gets called for interference, which I think is a crappy call.
• Gets deserved PI on a slant he was too aggressive on.

So… yeah, Warren had a couple incidents where Michigan gave up yards but the bulk of his day was running Juan Nunez's routes for him. Sometimes this got flagged and once he got tripped. But I'll take that sort of aggressive clamp-down coverage any day when the opponent is Michael Floyd. If Warren ends up a yard in front of Floyd the three to thirty times Notre Dame attempts to hit him deep, Michigan's going to be in good shape.

What happens when Notre Dame goes to three-wide?

Nothing. Michigan spent the entire day its base set and has no corner depth. They do have guys on the edge who can cover Robby "That's Racist" Parris or whoever; it's not like Notre Dame's backup WRs are speed demons.

Heroes?

Graham is the most obvious answer, and everyone shared in an all-around excellent performance before the D got backup- and vanilla-happy late.

Goats?

JT Floyd looked overmatched by a MAC team, which bodes very unwell for Michigan's corner depth. Jonas Mouton didn't have a strong game, though as mentioned it's tough to tell how good his zone drops were and the way the game went suggests they went pretty well.

What does it mean for Notre Dame?

I've sort of gone from thinking this is a bad matchup for Michigan to thinking it's an okay one or even good. Stick Warren on Floyd and Cissoko on Tate, give them deep halves help, spare the blitzing and let Michigan's diverse and sundry rushers attack the Notre Dame defensive line… I can see this working out. The prospect of a max-protect bomb still worries given what happened against Western, but if Warren's as ready to live up to the five-star hype—and he looked far more likely to in the Western game than any other to date—and Michigan can get away with shifting the coverage over to Tate and pulling up a safety into a robber zone to bracket Rudolph, I like Michigan's chances to hold Notre Dame into that 20-24 point range where victory seems a strong possibility. Notre Dame's run game has always been a finesse sort of thing heavy on screens and draws, which plays into the hypothetical strengths of Michigan's slimfast defense

I watched the Nevada game and a lot of ND's first half production was based on exploiting Nevada's "explosive pass rushers" at defensive end, which rushers also happened to be completely irresponsible. Graham isn't likely to be as exploitable, but Roh or Herron might be. I'd line up Graham on the strongside, which might induce ND to have Rudolph stay in to block, as they're going to double him lots anyway.

The key will be the safeties. Woolfolk is going to have to think deep first and not get caught flat-footed like he did on the Western touchdown; if Michigan loses to Notre Dame because of ND's ground game, well… that will be a surprise.