Upon Further Review: Defense vs Michigan State

Sorry for the delay. I knew watching the Tigers game would result in a disastrous lack of sleep and delayed content but it was too compelling to turn off. Video is still uploading and will be added ASAP. Update: video up.

Personnel notes: same deal with the substitutions—very little in the back seven, regularly on the DL. It seemed like the backup DL got a lot less time. Floyd spent the entire game at CB; Williams was yanked very early (second drive) for Kovacs, who remained in for the rest of the game. When Woolfolk was dinged up early, Andy Van Slyke got a play or two.

Well, that ended with a thud.

Yeah, I don't know what that was all about. I've never seen a DL just take off upfield for no apparent reason on a stretch play.

Chart?

Chart.

Before we get to the chart: I did not minus the DTs on the QB scrambles above but after talking with Steve Sharik I decided to retroactively ding them. Each lost three points for various instances of getting out of their pass-rush lanes.

There's a considerable amount to be encouraged in here: Van Bergen is going from a non-entity to a guy who's making plays (when he's not vacating space for Kirk Cousins to run wild). Jordan Kovacs registered a +4.5 and is single-handedly responsible for about half of the + tackles Michigan saw yesterday. And Graham murdered all.

On a drive-by-drive basis, Michigan faced 11 drives in regulation and gave up 20 points, three of which were a gift from the fake punt. Along the way they forced three turnovers, one of which offset a long MSU drive that would have put up points; the other two set up a field goal drive and a short field that Michigan did not convert on because Darryl Stonum fumbled. That's a good performance. The Mathlete's numbers for the field position:

Even with the turnovers, Michigan failed to have an advantage in field position.  The regulation numbers for field position were 24 expected points for MSU and 22 for Michigan. The first down at the 25 for OT is worth just over 4 points, so for the game, MSU scored 1 less point than expected (PAT is assumed) after being -4 in regulation while Michigan was -2 in regulation but -6 on the game.

So Michigan actually held a team under their expected points given starting field position and added in two turnovers. (The Warren interception came after a long drive and basically neutralizes that drive in terms of field position; the other two were three-and-outs on which MSU lost the ability to punt.

The yardage numbers are less sunny, as MSU got to around 420. Subtract the standard 50 yards for a turnover gained, though, and MSU's sitting at 270 equivalent yards. You do take turnovers as predictors at your peril, but in this game Michigan created all three turnovers with QB pressure, the one thing that reliably causes opponents to cough it up. That was without question the defense's best performance of the year. It should have been enough to win.

But the downside?

Yes. The downside:

Aaaaargh argh argh argh. Michigan was playing in the parking lot on a lot of plays and just gave Michigan State huge chunks of yards. This was mostly a problem on the first drive, when three different bubble screens picked up 24 yards and a simple hitch went for 15. (The hitch was called back because of an illegal man downfield.) The next two drives each featured a play on which it was stupidly easy to pick up ten yards: Floyd was in the parking lot on a long handoff and they ran another successful bubble. And the bubble fear instilled by these plays set up what was basically the only type of MSU rushing play to go anywhere all day:

Michigan's inability to defend something they can't even throw anymore drove me crazy, as you can see above ("AAAARGH," or something to that effect).

Was there some structural reason for this?

I think Michigan is moving to something sort of like a 3-3-5 in one aspect: when Kovacs is in the game he lines up on the weakside a lot and blitzes a lot and almost never sits back in the position you'd expect a traditional safety to be in. This is what he does:

He's the bottom defender in the box. Michigan's actually in a three-man line here with the linebackers almost stacked over the DL: this is either a bonafide 3-3-5 or very close to it.

Now, Michigan usually didn't go this far. They usually had Kovacs in the box like so as an addendum to a four-man under front. The line and linebackers are obviously differently aligned, and the rush defense works differently. But the pass defense is going to be almost identical to the 3-3-5: one deep safety, corners that tend to play soft because of that, and tight ends/flats being defended by your two safety-type wing defenders—Stevie Brown is the other guy on the wing.

But when Kovacs is in and tight to the line and JT Floyd is one of your corners it's really hard to roll up tight when you've got twins to one side. I'm still pretty pissed off you guys about everything but it is understandable. Sort of.

Heroes?

Brandon Graham blasted his way into the Michigan State backfield time and again, enshrining himself in UFR lore until such time as Kovacs registers a +20 by making every tackle against Ohio State. I mean:

Damn. Also, yes, Kovacs provided hard-nosed run defense that makes me think he'll be a positive contributor going forward.

Goats?

Mouton, Ezeh, and the two interior DL opened up the rushing lanes that turned Michigan State drives into too many points for Forcier to moxie away, thought they also had their positive moments.

What does it mean for Iowa and the future?

I'm a little leery of declaring the run defense fixed after the performance against State, as State's run game to date has been poor… but it was definitely a step in the right direction. Kovacs will get another shot at a pounding downhill running team against Iowa and Michigan has to be feeling a lot better about its chances at stopping it than they were last week. Michigan crushing anyone's ground game is reason for optimism at the moment. Iowa's offensive line will provide another level of test, however.

In the pass defense it was more of the same. Floyd isn't great and will have to play conservatively all year; he will be victimized. The inside linebackers are not good in zone coverage, and Michigan will suffer on crossing patterns and seams and whatnot all year. Michigan's turned in a couple games in a row where the pass rush was good, though, and if they can maintain that against better offensive lines that will help out. They'll have to prove that. Obviously.