Upon Further Review 2017: Offense vs Rutgers

SPONSOR NOTE. Man, it feels good to get something out of the way, like a Rutgers linebacker, or a Rutgers defensive end, or a Rutgers defensive back, or that mortgage refi you've been thinking about. HomeSure Lending can get that out of the way for you, and then it's smooth sailing until the endzone, or at least a safety.

FORMATION NOTES. Hello, manball. Michigan's approach in this game was downright neolithic, featuring 32 snaps with one or zero WRs. Feel the Harbaugh goodness as Michigan goes with a goal line set on first and ten on their own 34 (the "WR" is Gentry and he will motion to a TE spot presnap):

Michigan did suffer to allow two wide receivers on the field 22 times; three WRs managed to get out there on 14 snaps.

Note that Michigan's formations were slightly less neolithic than the personnel: Michigan was happy to spread their TEs out as WRs for a half-dozen or so 2 WR snaps.

SUBSTITUTION NOTES. OL was the usual until Onwenu got dinged late. Runyan got 7-8 plays in his place. Peters replaced O'Korn about halfway through the second. TE was the usual: mostly Gentry and McKeon with Wheatley emerging into a more prominent role but still third; Bunting snaps were scanty. Michigan did use a version of Spanellis wearing 97 as a bonus OL on a number of snaps.

WR was mostly nobody, but when it wasn't nobody it was Schoenle or DPJ on one-WR plays. Perry was always in when there were multiple WRs. Collins got a little run and a catch. Crawford was out after being spotted in a boot this week.

RB was a Higdon-Isaac-Evans-Walker production, in order of declining carries. Samuels burned his redshirt for good in the last few plays. I try to not get peeved about RB redshirts.

[After THE JUMP: hundreds of pounds of ground cable subscriber]

Tell me about the babe you saw in swaddling cloth.

I mean... he was fine?

FINE?

The move had to be made and Peters should be the guy going forward because he's at least as good as O'Korn right now and will be infinitely better next year since he'll still have eligibility. But the training wheels were on, and Michigan was clobbering Rutgers on the ground, and we just saw this happen with O'Korn in the Purdue game. There's a lot left to prove.

Please tell me you're about to say "but"

Okay, yes, but: good start, demonstrating the ability to alternate touch throws and blast it in when necessary. His first two throws were soft tosses that were almost casual pickup stuff:

That latter is straight butter, perfectly placed and easy to catch for a fullback. It's not a DO because it's like a four yard throw but it's the kind of short pass that makes it cross my mind.

He ramped it up for a ten yard hitch...

...and continued to fire it hard when the situation called for it. He was accurate and confident for the most part.

He also demonstrated that he has the ability to break the pocket and throw on the run, once hitting Isaac for a first down...

...and once converting third and long with a tough ball to McKeon.

Do that too often and it's a crutch, but on both of these plays Peters had a very good reason to leave the pocket and didn't appear to miss a read. Both deep routes seem covered on the Isaac catch and there's nothing much short; Cole gets beat so quickly on #2 that he doesn't have time to read until scramble mode sets in. Note also that both of these were blindside rushes. He got a shout, I'm sure, but that's good pocket presence. This also felt like good pocket presence:

Buuuuut: when O'Korn did that against PSU he couldn't go as far because the pocket had gotten pushed more and the RB block wasn't good so he got it stripped; the margins on these things are small and maybe in another situation this becomes a turnover. Not enough data yet.

Here is a chart. Two charts, now.

JOHN O'KORN

BRANDON PETERS

That is a very clean chart, save for the one asterisk.

Can he read defenses and come off reads, though?

Maybe? Kind of? Probably not real well right now? Peters didn't have to come off first reads much, if at all, and there were a couple of plays where he delayed his decision a split second. The most obvious and potentially disastrous of these events was the near-INT. This isn't a bad read; it's a late read.

If Peters throws that in rhythm as soon as his foot hits the end of his three step drop it's a completion with the safety tackling immediately after. The beat he waits gets him in trouble, and that's the #1 candidate for "reasons Peters was not tried earlier."

Also: Perry sits down, feeling zone. Peters appears to throw it as if Perry was going to continue at the same speed. One of these guys is a junior with a ton of playing time; one is a freshman. Once you're looking for it there's a moment of hesitation on a number of Peters's throws. I'm not sure if this is Peters waiting for the right moment or having a moment of indecision despite Perry being wide open:

Again there's a second after the drop is complete before the ball is out. Here it's totally fine because Perry's wide open. In the context of the near-INT I'm a wee bit leery.

OTOH, redshirt freshman instead of redshirt senior, and there were some blips. This appears to be a second read throw as Peters comes off of Perry to find Gentry:

What Peters would have done if the LB behind Perry stays in his zone is an open question. Hamilton did say the soft toss to Poggi was a checkdown; if so that's a quick and good read. His incompletion to Wheatley sees Peters read the right side of the field the whole way and throw when Wheatley runs into it, open:

That's not staring down a guy, that's reading a flood concept, right?

Okay, but what about the other babe in swaddling cloth?

My large adult babe?

That would be the one.

Yeah, Michigan's run game was an epic asskicking all around. Here's a slightly different version of last week's parade of one guy plays, this one featuring various Michigan players giving Rutgers folks the business. Ty Wheatley on the opening snap:

#17 TE to bottom

Juwann Bushell-Beatty on an otherwise unsuccessful run that we'll talk more about in a bit:

#76 RT

Cole stands up a charging LB and seals him inside on a weird pull-reach that looks super tough:

#52 LT

Gentry pancakes the playside end here:

#83 TE to top

Ben Mason puts his guy in the endzone from ten yards out, admittedly with some help from the guy himself:

#42 FB in a wing TE position to top

Mike Onwenu uses the power of his mind to FUS RO DAH a Rutgers DB:

(More seriously, McKeon does work on the playside OLB.) McKeon would later provide Walker a walk-in touchdown by blowing a 245 pound guy out:

#84 TE to bottom

Rutgers got paved. Even a lot of the runs that didn't work featured Rutgers getting paved. The Peters botch on the handoff featured an absolute cavern that Evans probably felt sick about:

all my beautiful yards, up in smoke

Wave your hands about just Rutgers all you want but this was the best run blocking performance Michigan has had under Harbaugh, no matter the opponent.

Sounds like a chart intro.

Yeah, this one can be Rick this week.

If you're like me the first thought you have is "that's too good to be true." Rutgers's statistically decent rushing defense seems likely to be a mirage that will evaporate by season's end. This is already in process: the Michigan game cost them almost 30 spots in the S&P+ rankings; they fell from 48th to 77th.

On the other hand, that only emphasizes what a paving this was. I have never charted a Michigan run game as dominant as this against someone other than the likes of Delaware State. 7.6 YPC on real carries, man. That's nuts.

How...? I don't...? I mean...?

This did seem to be coming as Michigan matured into the season. The Mathlete's got another demonstration of Michigan's increased run prowess for you:

Similar to the Line yards, this is OL rush success rate, odds of getting 3 yards or a first down on a run play by week. Ohio St is #1 at 73% on the season. Michigan was 86th after the first 5 weeks, up to 43rd for the season now.

In addition to paving a bunch of dudes, Michigan seems much smarter as a run unit. Let's go back to that first play. Check Gentry, who wants to seal his guy inside and finds that impossible because dude is booking for the corner. Okay, I'll just get this other guy.

#83 WR to bottom

Michigan isn't at the point where the second and third pullers see this and adjust themselves and may never get there, but there were a number of plays on which I thought Michigan was smarter about what they're doing. Here Cole checks to see if Bredeson needs help before hopping downfield for a second level block:

#52 LT

That's rare agility from an OL and a correspondingly rare +2 downfield block; if Rutgers did slant to this Cole would likely have cut it off for Michigan's profit.

Isaac's big run on third and long was a classic Never Turn Upfield example, because Michigan gets got by a (slot) corner blitz. Nobody can help Ty Isaac, so nobody tries:

And once Isaac beats that guy Michigan's a blocker up for the rest of the play. Never Turn Upfield has rarely been executed in my experience, so that's a thing.

Also, a couple weeks ago I complained that Michigan's OL didn't check for slants well, citing a chunk play from a Rutgers game a couple years ago on which Kalis perceived he needed to cut off a guy in the gap behind him. Kugler managed that here:

#57 C

This was not a universal or even common occurrence; slants did defeat a number of Michigan rushes. It's still progress.

In contrast to last week, big individual errors were rare. When runs were unsuccessful it was usually Rutgers putting on a good call, whether it was a line slant that got their LBs to a power play...

...or a backside stunt that sent an OLB directly to the gap. Other than an ugly power play where Kugler couldn't lock out a guy on the backside and Ben Mason getting hyperactive and tripping over his own guy

Onwenu and JBB both end up with no one they can reasonably block. Every offense in the world gets hit like this from time to time. Michigan had particular trouble dealing with corner blitzes, which tend to make all OL assumptions about how the defense intends to fill their gaps null and void. I had to look at this play a few times before I realized it was just an iso play that got a CB blitz.

Michigan, particularly JBB, blocks that one well and may give Higdon a lane to the safety but Higdon's cutting off Poggi there, probably because he's not expecting the corner to come and whack him from the side. CB blitzes are 100% run blitzes that suck at getting to the QB; Rutgers was able to call so many of them because Michigan had no passing game with O'Korn in and not much of one with Peters.

Anything as Harbaugh this week as podcast with Judge Judy or having a Halloween candy weigh-off?

Yes? I think yes. I'd been wondering if we'd see anything like a subtle power cutback play that Smart Football linked to on twitter a few weeks back. The upshot is: team runs a regular power play except that a couple linemen are actually trying to open up a backside hole. On the example he cited the Chiefs ran a power while their center hopped around a DT, opening up a lane to the backside.

So. Michigan has that "split flow" counter play that got them a critical TD against Indiana. That's on film and Michigan's run it a bunch in the Rutgers game and then suddenly there's this spate of plays that look like the Indiana counter and feature huge holes up the middle away from the nominal point of attack.

It's possible that this is just Rutgers overplaying it and a cutback but I think this looks intentional by Michigan. Kugler checks Bredeson's guy and when he moves away he goes to the second level; if that guy needed to get hit by Kugler, Kugler is hitting him to the playside to open up this gap. Also the RB's angle of attack is more vertical than it was against Indiana. Contrast that angle of attack with Isaac's on an actual counter a bit later. Michigan got two more gaping holes up the middle that the running backs looked totally unsurprised to see, so... yeah.

YMMV, but this one's pretty clear as Runyan and Kugler are splitting two guys in preparation for a jet up the middle.

Also, Mason does an otherwise crazy thing by helping Cole shove his guy way down the line.

Similarly this feels like an intentional cutback, where Cole's forceful shoving of his guy to the playside and Kugler seemingly taking an angle to the backside of the play, where he would be able to cut off a recovering linebacker, make me think this cutback is planned:

Also the angle here is vertical and actually bending to the backside before Higdon makes a decisive cut.

And then there was this weird thing, which sort of worked and only sort of worked. Check the tackle to the top of the screen pulling into nothing.

That was as designed. Every time Michigan ran this that dude did that and Michigan went off tackle the other way. I imagine that's supposed to hold the backside linebacker; it didn't, really. Here it did enough to get Michigan five yards, and that's about what it got the rest of the time. It was mostly a short yardage ploy (3/4 attempts were X and 1), so maybe that's intentional?

The lone first and ten edition of this play gives a hint as to the backside tackle pull's purpose. Watch the Rutgers OLB to the bottom of the screen widen out on the snap:

He's anticipating the crack and Michigan does pull that backside T on it. (For what good it does. Usually he runs behind the play and observes it.) He gives Hill a free kickout block and that's enough room for the RB to run away from that unblocked WLB.

This still feels pretty mysterious to me. It'll be interesting to see if Michigan keeps doing it and if they add some other stuff to play off this tackle fold more productively.

Perhaps more important than the modest yardage this play acquired was its contribution to a run game that was a flashback to that of the last couple years. It was one of a blizzard of different plays featuring virtually no zone blocking. Setting aside a FB dive and a couple of QB borks I had Michigan running

• 12 edge runs: seven crack sweeps, a jet sweep, and four pin and pulls,
• two zone plays, one a split zone to Isaac and the other an old style Rodriguez "belly" play, and
• 30 gap blocked plays in eight different flavors.

That's 13 different things to deal with, and I'm pretty sure the botched Peters handoff was #14. Michigan mixed them such that Rutgers was frequently running itself out of position. If we can go back to the Onwenu shout run, check the LB Bredeson engages with:

Rutgers MLB blocked by #74 LG

That guy is looking at a standard power play and once Bredeson gets to him he thinks "oh crap" and looks back inside because he's frozen in fear that Michigan's countering him. Bredeson has no angle to cut him off but because of that indecision Higdon flies by.

Also in standard power plays that baffle Rutgers linebackers, the long Higdon TD sees a Rutgers linebacker juke at nothing mid play because he thinks he's going to eat a cutback.

Rutgers #15, ILB to bottom of the screen

This was the play right after that power cutback thing. Onwenu decides not to block that guy, which is a mistake, but the delay still gets Higdon through.

Michigan didn't just  grind Rutgers down physically, but also mentally. By the second half their linebackers were chasing ghosts, and... I mean, I get it. There's a reason that Harbaugh's been able to keep ahead of the curve on the ground and this kind of stuff is how he did it at Stanford.

Well that's certainly the largest RPS sec—

Also in the realm of RPS wins, hello wheel route.

Where have you been all my life. Rutgers is clearly shifted away from this and likely to drop the DE in coverage on Evans; it is so and it is a touchdown.

This was a good offensive gameplan.

Receivers?

I just want you to know this was a really annoying chart to type out, with so many catches spread over so many receivers.

[0 = uncatchable, 1 = circus catch, 2 = moderate difficulty, 3 = routine]

ROUTES: McKeon +, Wheatley +, DPJ -

I just wanted you to know that. Also, McKeon hit for the cycle.

So... Kareem Walker now exists?

Walker gives off a distinct A-Train/Brandon Minor feel where he "runs angry," which is an idiom that's never made any sense but... I mean. Runs angry. He had a nice gain on a power play where he ran through an arm tackle and lowered his shoulder into a safety.

A couple of less successful plays also saw Walker finish his run like that, grinding out a yard or two on contact. He seems like he's all legs and hair but he brings it.

I have a bad feeling you're about to link to something you clipped just because you want to justify your occasional statements that Matt Millen is good at anything, in this case being a color guy.

Look... okay, I remember "follow the fullback" but the guy points out subtle stuff with some regularity, and has a knack for immediate ID of things. This is good analysis!

Nobody cares about the Lions! Get over it!

Heroes?

Literally everyone involved in run blocking. Or running the ball.

Maybe not so heroic?

I mean... the QB change was for a reason.

What does it mean for Minnesota and the future?

Please be 80% as good as this. This is going to be Michigan's best ground output of the season. I accept that. I am serene about it. But if Michigan can maintain a level of play a notch or two down from this they will go into Wisconsin at 8-2 with a real shot.

Brandon Peters, welcome to now. All you have to do the rest of the season is run play action and convert the occasional third and eleven. Maybe this one is the one.

Boy, they've got a lot of guards. JBB would probably be just fine at guard. Runyan and Spanellis are looking functional. My kingdom for pass protection, though.

Harbauffense is here. Yeah that zone stuff from early in the year is completely gone now after being 40% of the offense even after a shift midseason. That may be a matchup thing: on the Friday podcast I speculated that a team with big solid DE/DT types and a weak edge would be one M could attack with power concepts. I assume that IZ practice they put in is going to be at least somewhat useful going forward.

If McKeon and Gentry can block... I still think the jury is out on them a bit since their opposition has been weak. They both have the potential to be Michigan's best two-way tight ends in a minute here—Butt was always an average blocker.

Ben Mason's going to be just fine at FB next year. Nice to check that off the concern list.