Upon Further Review 2014: Offense vs Miami

FORMATION NOTES: Hello tiny TEs. Michigan used a lot of formations where they would bring a wide receiver tight to the line to act as a blocker. Here's Chesson in what I called "pistol biggish," because it's only big-ish.

For its part, Miami ran an under front whenever presented with seven blockers for the opposition, and about 90% of the time brought a safety down late or just lined him up in the box.

This press look was not common.

Miami would roll that safety down before this snap, FWIW.

Michigan used a lot more under center stuff in this game. Under center stuff was approximately 55% of the offense after being maybe 20% against ND, and there were a lot of tight ends. Only about 40% of Michigan's snaps had 3 WRs, again way down from ND.

SUBSTITUTION NOTES: Line static: Cole/Magnuson/Miller/Glasgow/Braden. I saw Kalis in for the last drive, and I thought I saw him earlier in the game live but either I missed it in the film review or my mind was playing tricks on me. 61/67 are not easy to distinguish. Burzynski got in at the tail end at left guard.

Gardner QB; RB mostly Green, with less Smith and Hayes relegated to third down duties and some late stuff. Johnson did not appear. Mo Hurst(!) got a goal line FB snap. Showy, but a dollar says Kerridge is more effective. At TE, Butt got a little bit more time but it was still mostly Williams and Hill, with Heitzman again appearing sporadically.

Without Funchess, Darboh and Chesson were the main guys at WR, with Norfleet marginalized with a ton of 2TE sets. Damario Jones got about as much playing time than Canteen, making the first catch of the game.

[After THE JUMP: yards, eventually, and yet more infinite RB discussion.]

WHY WE NO POINTS

Oh, some bad luck and some continued teaching moment whatnot.

Bad luck?!?!

Look man when a ball gets deflected right to a DB because someone on the line gets a fingertip to it, that's bad luck. And the punt thing was something Michigan has to fix but not on the offense; the Darboh fumble shot down a drive that had just cracked field goal range. I think wide receivers should not fumble, of course, but the fact of it is that sometimes they do not take my advice and there's no way to predict it or plan for it or really even coach it—Darboh was high and tight until the ball came out a fraction of a second before he was down.

So I'm looking on a down to down basis and it looks… fine.

Fine?

Are you mad that I think it's fine and not horrible or mad that I think it's fine and not better after running for like a million yards at a good clip?

Both?

At least you're consistent?

I guess?

So to address both outlying complaints: Michigan mashed Miami off the ball, a necessary but not sufficient condition to approach the rest of the season with anything other than gallows humor. Check the Miami DTs on this stretch play, which is not always a situation where you need to blow guys downfield:

But that's three yards. It's three yards because

1. Cole doesn't gain any ground and makes Green's decision non-obvious.
2. Green doesn't make either decision quickly, cutting in-out-in.
3. Michigan doesn't do much of anything to hold the backside end, who is able to shuffle-pursue effectively.

1 and 2 are recurring issues. The line is better but still experiences minor breakdowns more often than high-octane OLs, and the tailbacks were better but still felt like they weren't acting instinctively. #3 is an RPS thing that they seemed to be aware of and had a play for (about which more later).

So if you're frustrated about the early bit of the game where Michigan kept eking out two and three yard gains, you should know that Miami didn't change their approach one iota—it was eight man boxes all day—and the minor problems that hampered the early runs were solved, whereupon Michigan did what you would expect them to do. Except, of course, we don't expect them to do much of anything right now so even progress against Miami is progress.

Por ejemplo, Michigan came back to the stretch to the left on the next drive; Cole's block was more emphatically sealing and a yard further downfield, so Green's read was clearer and decisively taken:

Those two plays are a summary for the game itself. At first I was like ugh, but then I was like… okay. I mean, Miami guys were resurrecting techniques from the days when defensive linemen looked like mean pencils. On one play the only thing preventing a big gain was a defensive tackle figuring his best bet was to fall over and make a pile and turning out to be right.

If that's the best they've got, okay. The line is doing some very nice IZ things. This Green rumble is textbook:

Those combo blocks are how it's done. M gets movement on the nose and Magnuson pulls off to intercept a charging linebacker; they get a different feel on the other double and end up scooping him to get to the other LB. Consistent? No. I'd be surprised if you showed me one play from last year's line that looked that good, against anyone.

So this is why I am relatively happy. The line seems a lot better, and I think that'll show over the course of the year. HERE IS A CHART

dammit, no fair

You are the second-slowest bolded alter-ego in the country at getting the chart interruption off.

So this is a ground clubbing, but that protection number is not great against a bad team. There was one 0/4 sack in there, but it was really an 0/4 sack, featuring a totally whiff by Braden and a second guy coming through a double thanks to bad communication.

Tailback evaluate-o-rama time.

Obvious progress, and lingering frustration. Here is a digression about the thing I know best about running back patterns since I've seen piles of it under DeBord and RR: the stretch.

I thought we were an inside zone team.

We are but we run enough stretch to keep people honest. And this is about RB decision matrices.

Oh you fancy.

I AM FANCY

Anyway.

I'm not sure if Michigan teaches the stretch the same way Rodriguez did but it seems likely that they do. Rodriguez told his backs to follow this pattern. First you run hard to the playside tackle and then you try to cut as close to him or your next option as possible. Your matrix, ordered by priority:

1. BOUNCE. If the tackle gets his guy sealed inside head to the corner.
2. BEND. Check the gap inside the tackle if the defense has forced it back.
3. BANG. Cut behind the next OL on the line (usually the center) and get what you can. (Green's cut on the embed above is a Bang.)

Since a stretch is a stretch I'm guessing that Michigan's pretty much the same. "Bend" is your goal as an offense. Bounce only happens in cases of defensive incompetence; Bang is bringing a lot of backside players into play. Bend runs are those runs where the center steps around a guy lined up a gap to the playside of him and you get a gashing run. Bend runs are Molk runs.

In that context it seems like Green is not always executing per spec even when he's getting yards. The bang embedded above was possibly a bounce situation and his hesitation cost Michigan a yard or two—every false step taken is one step closer for the backside end.

This was not isolated. Michigan had an odd tendency to run stretch on short yardage, and here Green cuts up for the first down:

Hooray for moving the chains; I wonder if Green got a talking to about his approach here. Ideally he runs harder to the tackle, checks to see whether the gap between Cole and Miller is viable, and if not cut right off the back of Miller when his guy fills the gap. Here he picks a side before the guy blocking Miller does. And it works out, so hooray and all that.

Green pulled it wider in the fourth quarter, FWIW:

That's a bend in which the tackle has not sealed the guy inside (which should never happen unless the defense goes derp) and Miller has gotten the DT delayed enough, if not exactly walled off.

The point is: going that wide expands the range of blocks that are good enough to get the job done when you do cut (hard and decisively) upfield. Miller loses the guy in the most recent embed but by the time that happens Green's already gone; in the first one a tackle gets initiated by a player who flew upfield of Glasgow, something that should mean you have just decided not to participate in this stretch play. I may not be picking at this if I hadn't just seen the tailbacks with the equivalent of five airballs in the last game, but I dang well did.

But this is wishy-washy stuff about how plays that worked were not working per some imaginary spec that Michigan may or may not use.

Okay:

[Side note: check Darboh forming up for a quick pass like the one Funchess grabbed in week one. Presnap abort check, that is.]

That's bad yo. Green ends up with a massive frontside gap and cuts to the unblocked guy for zilch. Seems like he sees that tackle rip inside and thinks OH NO NOT AGAIN and runs away from it because he's still assuming this is doom; Magnuson calmly pancakes the dude and now there's an inviting hole. This happened more than once.

Neither of these plays is as frustrating as the ones from Notre Dame because there are some yards on the backside and the holes aren't as gaping. I do want Green to favor blasting playside because anyone contacting him is going to be harassed by an OL and pile-lurching is in the cards. This was relevant on the third and short that Miami stuffed up. A LB hit Kerridge hard in the hole, Kerridge stood him up, and Braden ripped his guy to the outside to provide a crack; Green cuts back into a pile of anger.

Sometimes he needs to just run the damn play instead of cutting to the backside. Driving your legs forward with Kerridge pushing is a better bet than heading back to the writhing pile of limbs behind. I didn't minus him for this because it was by no means obvious, but it's another example of Green preferring to cut away from the intended path of the play.

De'Veon Smith was coming off a game where he was less obviously frustrating than Green but in this one the roles were reversed, to the point where I think Smith not taking the right cuts saw Michigan bench him until the fourth quarter. Thus the highly imbalanced carry distribution. The early TFL no doubt set teeth on edge, again on a stretch:

Braden set you a bad table, but once that guy is that far in the backfield you have to cut behind him and eat whatever's waiting for you. In this case the answer is a bunch of dudes slashed to the ground and some yards instead of no yards; sometimes it's going to be no yards. This goes double for a guy like Smith, who is not fast but does run through tackleslike a mofo.

Smith's last carry of the first half was similar to the above, another stretch where he tried to go outside when presented with difficulty. This one was mostly RPS with a Miami LB blitzing his way into the running lane but Michigan actually gets this blocked for the most part and if Smith doesn't try to run outside of blocks that have defenders in those gaps, a MAKE PLAYS play is there.

Tough, yes, and if his next carry wasn't almost 30 minutes of game time later I wouldn't read anything into it but it dang well was so I am.

All of this is of course moving too fast for Smith to make actual decisions about this; it's why people talk about feel and instincts when discussing this stuff. Instincts can be learned—they're what happens when you're executing things so familiar that they no longer rise to the level of conscious thought. Reps reps reps reps. Hopefully things start feeling more natural as the season goes along.

But MAKE PLAYS?

Yes, there was some MAKE PLAYS in this game. The main one came on a weird run that featured a fullback, a rarity. That fullback hit a backside gap on a play that looked like it was going to the other side.

Unblocked guy showed up in the frontside gap, Green cut backside hard and ran through a tackle. I don't know what that is. There is no backside end left alone; the opponent doesn't have to account for the QB. The free guy is just waiting for Green, and then Green beats him. Wave your tiny flags, people.

BONUS: I'm counting a third down conversion on which he got nailed behind the line and spun out for the first down as a MAKE PLAYS as well.

Was it just me or are there a ton of plays labeled "counter" above?

Not just you. Michigan ran a ton of 'em (nine, almost as many as any particular base play) in two primary flavors. One was your regular pull-this-guy-backside counter. Braden can't get out to the second level on this one:

(Note that this was one of the rare plays on which it seemed like the extra Miami guy in the box had much impact.) Michigan had better success with it on the very next play from a three-wide formation with Miami overloaded to the IZ side.

Michigan is also definitely doing some sort of subtle-ish belly type thing which doesn't have a pull but is definitely headed for the backside of the play on purpose. One of those was mentioned in the Picture Pages;  they went back to it for the Green touchdown that was bounced to the corner:

The line fires out in an inside zone-ish fashion; Michigan hands to the backside of this hypothetical IZ. This was even more apparent when Michigan ran it from under center:

Also on this play: I'm hoping Michigan can find a way to screw with this shuffle technique DEs do these days. I think they tried something like that on a Gardner keeper that didn't quite work out. Watch Darboh start to crack down and then change his mind:

Continuing would be a violation of the hallowed Never Turn Upfield blocking strategy, but in this case I think we need to allow it. Michigan caught a scrape exchange here that may have confused Darboh. If he does get that crack down Gardner has  a lane upfield for a nice gain.

Norfleet block of the century of the week?

What an excellent idea for a thing, bolded alter-ego. Glasgow your winner.

What was with this tiny TE business?

I assume they think Chesson, Norfleet and Darboh are better run blockers than any defensive backs are run defenders and are hoping to seal edges and the like, creating piles that are hard to run around. It was… iffy. I'm a spread zealot so I just see piles of guys who can screw something up when teams go big, and freshman tight ends and wide receivers don't make for confidence.

We'll see.

I've been a bit frustrated by this shuffle-DE action opponents are showing on the zone read that induces handoffs and allows the backside end to pursue hard. Some of these cutbacks are nice gains if the DEs are pursuing upfield to contain, and this may be a way to address that by mucking with who exactly the unblocked guy is. It also creates the potential for plays like one Michigan didn't quite make but looked like they could be tweaked to aim at the DE shuffle. Check Darboh:

He's cracking down on nobody as Miami sends an exchange. If he does pick off the LB, or a shuffling DE, Gardner has a lane he can go upfield in between that and the Butt block. Finding a way to punish the shuffle is recommended, because it seems like everyone does it.

We haven't actually talked about anything other than running.

Right. Right. Gardnerchart:

Devin Gardner 2013

Devin Gardner

Shane Morris

Gardner was basically fine. The bad reads were mild varieties: he did not throw to Darboh when he seemed open and then after a bunch of running around threw a much tougher ball to him on one particular third down, and he threw at a covered Hill for a PBU when he had time and a running lane on another. Neither was mindblowing.

His interception was tipped, and while the throw was going to be a bad idea either way, to me that's more on the play design than Gardner:

Kerridge is the flanker on second and nine; Miami deploys a robber safety who is headed for Chesson's ribs on the snap. The resulting hospital ball is asking for trouble even if it's not tipped. Meanwhile Kerridge is totally open one receiver over… and a fullback.

I'm fine with this SPLIT OUT ALL THE BLOCKY GUYS as a tactical gambit on standard downs; on second and nine or even more extreme passing downs I say thumbs down. Also I'm hoping that Butt's increasing involvement will put him out at flanker, where he's viable enough to prevent these kind of overplays.

The other dodgy throw was thoroughly so. Gardner found Jake Butt open, moved up into the pocket, and threw a weird jump pass well short of the right place. Things looked grim until Butt rescued it for the rare IN-marked catch. We saw Gardner execute a throw like this to Funchess in the opener and the right throw here is the same kind of looping pass that was, but the reason they mechanics you is to make you consistent. That was not consistent. But I think we're just stuck with that now.

Wide receivers?

Not a ton of note other than Darboh flashing those slant merchant chops when Miami loaded up:

Here's the chart:

ROUTES: Darboh +1, Hill –1.

Heitzman turned in Michigan's first flat-out drop of the year. So it's been pretty good so far when it comes to hands.

Heroes?

Interior line. How about THAT. Green.

Maybe not so heroic?

The tackles were a clear step behind the inside guys when the pressures they allowed are factored in. Darboh did fumble.

What does it mean for Utah and the future?

The tailbacks are working on it. Gradual improvement, maybe Green getting some faith that hole is not a mirage, are goals for the next few games.

The tackles showed some holes. They were still positive overall, but here and there they showed some freshman issues. I assume this is going to bite Michigan at some point.

The interior line was coherent and good. Level of competition yes. Progress undeniable.

I don't know what the deal was with all the TEs and under center stuff. It wasn't a gameplan to defeat Miami, in all likelihood. I prefer Norfleet to Hill and Williams at the same time. Butt and Williams is more plausible but still let's just get him the dang ball five times a game.

Darboh does feel Avant-ish. He is not a blazer; he'll have sufficient assets other than that to be a solid #2.

We aren't going to do the same thing over and over and say "stop it" and then say "oh I guess you stopped it." Thumbs up to the frequent counters in different flavors and liberal sprinkling of stretch. Michigan might have to whittle down if they can't run their alternate stuff effectively, but this has the shape of an offense with a thing and then alternatives that burn you if you overplay the thing, which Borges offenses never really did.