Upon Further Review 2016: Defense vs Maryland

SPONSOR NOTES: A gentleman named Matt passes this along:

Just picked up the UFR and saw your note about non-astroturfed comments. 

I'd like to add another - I just  bought a house using Homesure as our lender (because of a bad experience with our previous lender and the adds on the blog, of course).   Matt and his crew (Amy and Gina) were supremely professional, knowledgeable and effective.  

Without exaggeration: every other real-estate professional related to the sale of our old house and the purchase of our new house dropped the ball somewhere (losing documents, failing to transcribe information, etc...) - and Matt & Co. were there to bail them out at each turn.  This iteration of house buying was shockingly easy, and it was due in great part to Matt's expertise, acumen, and responsiveness.

'Thank you' to you and your fellow bloggists for turning my wife and I on to his company.

This guy said bloggists so you know he's a real person.

In addition to being a gentleman replete with Michigan tickets, Matt is also a good man to know if you need a mortgage. It's striking that we actually get non-astroturfed comments about positive experiences with Matt not infrequently.

If you're buying a home or refinancing, he's the right guy to call.

FORMATION NOTES: Maryland was mostly a rote spread team; they did spend a big chunk of the day in this Rich-Rod-esque super wide formation. There are two WRs barely on-screen to the bottom, and you can see the two guys lined up outside the numbers.

This was an effort to clear the box and limit the things Michigan could do; it was reasonably successful.

SUBSTITUTION NOTES: 60 snaps before Maryland's final uncharted drive, and as per usual the starting secondary—including Peppers—got all 60. Gedeon missed one snap.

After that there was a lot of rotation. Mike McCray only got 36 snaps due to a minor injury and some iffy play; Devin Bush got 28 in his stead. Michigan eased off first team DL snaps: Charlton, Glasgow, Wormley, Hurst, Mone, and Godin all got between a third and a half of available snaps; Winovich was your DL leader with 40 and Rashan Gary was close behind with 36.

One item of note in the secondary: Lavert Hill got 17 snaps, more than Watson(14) and that competition appears to be on. Kinnel, Metellus, and Hudson all got around ten snaps, with Kinnel the guy getting PT during non-garbage time, insofar as that existed in this game.

[After THE JUMP: edge edge edge edge edge edge.]

Soooo... are we concerned yet? You dismissed the fourth quarter of the MSU game pretty hardcore a couple weeks ago.

This was more concerning than the MSU game. Reasons why:

• Maryland had some stuff that I assume was new but they repeated it over and over again; their ability to move the ball early wasn't based on a meticulously scripted drive and intricate motions that petered out as soon as they went off script.
• They moved the ball consistently through the first three quarters, unlike MSU.
• They were not the beneficiaries of ludicrous, drive-extending penalties and did not hold their asses off to get most of their yards after the first drive.
• The tackling issues that were new, and potentially a fluke, against MSU cropped up again and are likely a real problem.

This isn't little spinning panic guy time—Michigan gave up 3 points and had five turnovers, three of them on downs. They were always going to come down from their unsustainable, should-give-up-less-than-a-point heights from early in the season.

But, yeah, man. There are some holes.

First let's check the

chaaaaaaaaaaaaaaa

...

aaaaaaaaaaaaaaaaa

...

aaaaaaaaaaaaaaaart

was that you being an edge run

YES

deplorable

aw, cumong man

Anyway:

[NOTE: PFF grading works out differently than mine. In UFR zero is bad for DL; it's average in PFF. Their scale seems to be about a third or a half of mine. PFF has a "penalty" section that I've elected to omit for space's sake, which is why the individual components might not add up to the "overall." number. I'll note when it's a big discrepancy.]

Relatively thin numbers for the DL and a bunch of stuff for the linebackers as Maryland correctly decided any attempt to block a Michigan DL was likely to go badly. That +/- ratio is unprecedented. 

What happened to the +17 and +20s of my youth? I mean, +12 is fine. But I mean cumong man.

Can I assuage your hurt with clips?

Taco Charlton built on an excellent game last week with a bunch of plays that didn't show up in the box score but were nonetheless key. He provided all manner of back seven gents with a free flight path on the second and goal TFL that preserved M's shutout for a long while:

On the ensuing third down Charlton blew up a screen by going after the back so he couldn't get in his route. This game was validation of the UFR concept, because a half sack was his only contribution in the box score when he was Michigan's best DL on the day.

Unless it was Mo Hurst, who remained absurdly disruptive. His ability to shoot into the backfield has been destructive to opponents even when they're running away from him.

Also when they're running towards him. Trying a pin and pull when he's on the field is waving a red flag in front of the bull.

Get in his belly.

He had a ton of production in just 31 snaps; when crunch time comes I think he'll be out there with the first team line. In a walkover against Maryland, fine. Even early against OSU, fine, he can be second team. In the second half he's going to be next to Glasgow and those guys aren't coming off the field much, if at all.

Finally, Chase Winovich added to his burgeoning highlight reel.

Winovich got more pass rush opportunities than anyone else—22—and had a sack, three hits, and three hurries. That's excellent production. Jake Ryan as a DE: on track.

Uh... so you hated Stribling.

Stribling picked up some scattered negative half points on edge runs where he didn't get off a block; he got a –3 on the reverse because he was set up in the backfield unblocked and let the guy outside for a first down on what should have been fourth down.

On the previous play he was tasked with replacing the safety after his guy cracked down, and he just about gave up the same gain. His tackle was a shoestring thing that turned a should-be TFL into three yards. His run support is an issue. His guy was also open the two times that Maryland threw at him.

Sometimes this just happens. Because I'm grading a subsection of players on each play variance can be severe, especially for DBs. I assume he was in good coverage on various other throws that I cannot grade because he's not on the screen; I still think he's a good player.

Also, the PFF grade is considerably different. Some people have left comments to the effect of why I bother with them if they have a different opinion that I do and I justify mine. As always, I am an amateur and I welcome any and all sanity checks. PFF and I are in close agreement about 80-90% of the time, so the differences are interesting to me. They might be right, after all.

But yeah man that reverse was WTF.

I didn't know you could build the whole plane out of edge runs.

I mean... you can't, really. Michigan didn't play those particularly well but since those were entirely reliant on one-on-one matchups without blocking they were more about Michigan than anything Maryland is doing. Maryland drives petered out because eventually you're going to get tackled at or near the line of scrimmage, and then you've got to do something else.

Maryland's tendency to run the same thing from super-wide formations repeatedly isolated the same Michigan players in the same situations, so we got to see Mike McCray undergo The Evolution of Man on various edge runs.

Edge run #1: McCray takes a crappy angle, gets edged, recovers well, tackles after eight yards.

Edge run #2: McCray heads to the edge much quicker, tackles after two yards.

Edge run #3: McCray takes an even better angle and is about an inch from a tackle for loss.

He misses that tackle and his buddies got a little complacent about it so the guy can get a few yards, but most of the time that play is going to be lucky to get back to the line of scrimmage if McCray plays it like that.

At this point Maryland ditches running inverted veer from these super-wide formations and moves on to a different thing, a bunch of edge pitches. This switched the linebacker Maryland was testing on the edge from McCray to Gedeon and again finds that LB successfully flanked:

It's the first step here that gives Maryland five yards or so. Gedeon's going vertical; Maryland is going horizontal. The next two plays were the same thing; Michigan got severely RPSed on one and then McCray was like eff this, man.

The linebackers got better and demonstrated they have the athleticism to get to the sideline against a super quick scatback. If they rep it and rep it, as I imagine they will over the next few weeks, I think they can get it down and mitigate many of these issues. That's not to say they won't get got on these plays—when Maryland returned to testing McCray one on one on the edge in the second half he got beat for 21 yards. I expect OSU to run a bunch of edge stuff and get chunks of yards. I expect they'll be less consistently successful.

They'll rep both LBs and Stribling constantly on edge run support the next few weeks and insert various changeups to defeat the kinds of things they struggled against in this game. Hopefully it pays off. I mean, this could be a blessing in disguise.

Wait, what?

I'd rather not have the issue. If there is an issue I'd much rather have it pop up now than in three weeks. I enjoyed watching Ohio State get gashed by Wisconsin's various manball shifts, but also hated it because the Wisconsin game gave OSU a very clear picture of problems they need to fix before Michigan gets to town. This is similar. 

And then they didn't actually gain many rushing yards.

No, they didn't. Remove three sacks for 15 yards and Maryland acquired 93 yards on 35 carries, 2.7 an attempt. It felt much worse than that because Michigan gave up chunk plays a bunch. Those were offset by a whopping 38 yards of TFLs. That is an inevitable cost of going wide all the time. You are exposing yourself to Tyree Kinnel getting on his horse.

Or a linebacker getting it really right, or a corner blitz, or whatever. If you run side to side you will eventually get TFLed.

And when Maryland went anywhere other than the edge, this was the inevitable result.

Maryland's gameplan was predicated on the absolute fact they could not hope to block Michigan's defensive line, and it did relatively well in those circumstances. That fact remained a fact, and eventually the edge stuff peters out.

Once it did peter out, any passing down resulted in the quarterback being overrun by rushers.

This was so severe that Maryland started punting up fade routes at Jourdan Lewis because that was their best option to get anything on a passing down. Sometimes they punted up fade routes at Delano Hill because two different guys were arguing about which one got to eat Caleb Rowe's bones.

How did those shot at Lewis go?

You know the answer to this.

Nuh-uh.

They went the same way they always do, man.

No sale, no sir. One time the receiver almost got a catch because he pushed off—that offense was more severe than the OPI on Drake Harris—but Dymonte Thomas came over the top and separated the WR from the ball.

Even when it wasn't Lewis, Michigan was crushing downfield attempts. This is probably Brandon Watson's best snap at Michigan:

That's what he did throughout high school and it's a little odd he has been asked to do things that aren't that.

But they passed for ten yards an attempt nonetheless?

Hills and Rowe had 276 yards on 27 attempts, yes. 103 of them came on four tunnel screens. 72 came on two plays that hit against linebackers: Gedeon got beat badly on a wheel route and Devin Bush was unable to keep up with one of their running backs on a seam route. Everything else averaged 4.6 YPA.

Here again is a clear issue of limited scope: Michigan has frequently been gashed in the passing game when their linebackers get stuck in downfield coverage.

Gedeon fruitlessly chasing a wheel route has been a theme on the rare occasions something does go wrong for the defense.

Unlike the edge stuff, I think this is a permanent issue that Michigan needs a tactical response to. That wheel route sees Delano Hill run up past everybody into the flat; ideally Gedeon has the flat and Hill pops back over the top to deal with the wheel.

Meanwhile, Michigan got gashed on three first half tunnel screens. These were mostly Michigan calling rock against paper. The first saw a six man box and five guys blitz:

That's ten yards and then poor safety play extends it. The second tunnel was almost exactly the same thing; the third ended up with Michigan sporting a five-man box since a running back motioned out, taking Gedeon out of the equation.

I don't really know what to do with that one. There are seven seconds left in the half so Michigan's not expecting a screen, and Maryland again has three blockers for two guys anywhere near the LOS. McCray needs to push the panic button here and give a ton of ground to get over the top; instead he goes under the block and is fortunate to catch up at the one.

This obviously has to get fixed. It should, but since the answer is likely the defensive back in off coverage reacting much more quickly that will leave Michigan vulnerable to getting hit over the top. The one tunnel in the second half saw Thomas react much more quickly to that action, which helped Bush out by opening up some extra room he could attack.

Remember the jet sweep against Penn State? PSU had been atrocious at defending it. Michigan ran one and lost five yards; they then got a big play from Peppers after a jet fake took multiple PSU defenders out of position. Michigan will rep these screens and defend them much better; it's the counter to Michigan's counter that worries me.

Did we get any cool Don Brown things? I feel like the defensive UFR needs one of these sections as well.

Michigan briefly returned to the 3-3 package they've been playing with against spread teams, getting an RPS-heavy TFL from Gedeon on a blitz that crushed inverted veer:

Michigan went away from the 3-3 shortly afterwards, sticking with four man lines most of the rest of the day. Brown still schemed up the occasional large TFL. Here Hurst and Gary stunt; Hurst ends up totally unblocked and the results are pleasing:

Michigan is getting many more TFLs this year because of unblocked guys teleporting into the backfield, and seeing these items that seem directly targeted at OSU staples is heartening.

Heroes?

Charlton and Hurst were first amongst destructive equals on the DL. Gedeon had a very solid performance as Maryland sought to test him again and again. Someone threw fades at Jourdan Lewis.

Maybe not so heroic?

McCray wasn't terrible but he did end up slightly negative in my grading; Stribling missed one very consequential tackle and was beat twice when targeted.

What does it mean for Iowa and beyond?

Taco taco taco taco. The Wilton Speight of the defense; he's blown up the last few weeks. He has unlimited NFL potential; would be real nice to see him start hitting that down the stretch here.

Chase Winovich crushes bad OLs. He had a game like this earlier in the year where he was like +14 or something despite playing only half the snaps. He's been up and down since; the potential is clearly there.

Michigan has work to do on the edge. This is mostly a Mike McCray and Stribling thing; while Gedeon gave up a little bit here and there it's telling that they only tested him a couple times and then gave up on it. McCray has the ability to do it, as we saw. He doesn't have the margin of error Gedeon has.

I'm going to be worried about Stribling making a tackle until I see him do it repeatedly. He's reached Early Season Kenny Allen status.

Maurice Hurst is just waiting to be unleashed. Expect a big role for him as the opponents ramp up at year end. He's killing it, and can take another chunk of snaps.

Rep those tunnel screens man. That's a big flashing DO THIS sign on a defense that doesn't have many others.

Linebackers are vulnerable in coverage. I don't think they can be anything but; Michigan needs to design some responses to wheel routes that involve safeties instead.