I thought that myself when I read that article that talked about a Data Scientist(tm)
A little more on what looks to me like one of the major issues with the run defense: the two MLBs not reading plays quickly enough. This was one of the videos featured in the UFR, FWIW.
Michigan State has first and ten at the beginning of the third quarter and will run an inside zone from an ace formation; Michigan is in their standard 4-3 under with Kovacs rolled down:
On the snap State starts to develop the run action and the linebackers start creeping forward:
A moment later the handoff point is almost reached and the two LBs are still three and four yards off the LOS.
Contrast this with the MSU defense on Michigan's first and five on their first drive:
Both are two yards closer to the LOS and rapidly approaching. This was a consistent theme: MSU linebackers, even when not blitzing, were screaming at the LOS.
At the handoff there is one blocker for two guys because Heininger was doubled on the backside. Martin is driving his single block into the backfield and Van Bergen is cutting off the outside. Kovacs is still hanging around for backside bounceouts.
The above is not a good setup for an offense.
But Demens does not get outside his block.. and Hawthorne starts moving up into a hypothetical gap that the RB is not headed to. Even if he wants to cut backside the Martin penetration means it will take absolutely forever. Still, he starts moving straight upfield instead of flowing to the hole:
By the time Baker manages to squeeze through the gap left by the DL, Hawthorne is hardly closer to him than when he was three yards behind the LOS and Demens is still two yards downfield, not funneling the play back to help.
Baker pops outside. Countess fills quickly, but can't make the tackle…
…and neither can Ryan.
Items of Interest
The DL cannot do much more than this. They got a two-for-one on the double that leaves a free hitter. On the frontside they drive into the backfield such that the tailback has one realistic option. Short of throwing offensive linemen into the RB, they have done all they can.
The linebackers are uncertain of what they are doing. This has been a theme all year: me complaining about guys pulling in front of the LB's face only for that LB move directly upfield instead of scraping over to the POA. Sometimes poor DL play has washed them out, but often it's just derp.
Both linebackers screw it up here. Demens has to get into his blocker further upfield; failing that he needs to pop outside of him to funnel back to help. He does neither. Hawthorne can't see that his assigned gap is not an option because of the penetration and slows up for what turns out to be no reason. Either could have made this play themselves; it takes both of them screwing up to send it to the second level.
I'm sure they're more concerned about play action than Michigan State was because of the quarterbacks in question, but they get blocked way too often for my tastes. Hawthorne had already given way to Morgan for a series or two in the first half; IIRC this would be one of his last drives before Morgan re-entered for the remainder.
Ed Baker is hard to tackle and fast. I wish he was on the football team I liked instead of one I do not.
Countess does a great job here. I know he misses the tackle but a cornerback impacting a tailback just outside the hash four yards downfield is quality run support. If the linebackers hadn't compounded their Keystone Kops impression by banging into each other and falling over Baker is gang tackled after a moderate gain; as it is only Ryan is there to tackle and he is run through.
[Ed: commenter wile_e8 makes a great suggestion: check out the earlier ND Check Yo' Self Picture Page for everything Michigan wasn't doing against MSU.]
One of the main issues with Michigan's offense was an inability to adjust to Michigan State's constant double-A-gap blitzing. BWS has an example where it ate up a Smith run; this post has two more focused on the precise timing MSU used to shoot into the backfield untouched on multiple plays.
Two plays in this one. The first is actually a 25-yard run on Michigan's first drive on which Vincent Smith breaks a tackle when the WLB gets too far upfield. It would be a disturbing omen.
It's second and one; Michigan is in a three-wide shotgun set and MSU in the 4-3 they'd run all day. Don't bother screaming that the bubble is open.
All right, so Molk starts to put his head down; when it comes back up he snaps immediately.
Molk's head starts down…
And by the time it's completely down Allen is nearing the LOS.
Bullough is next; the blitz seems like it is designed to have Allen pick off Molk while Bullough gets a free run:
But Molk snaps the thing so quickly that he doesn't even get his head up before the play. Instead of blocking Allen he goes to double the playside DT. He does not see the blitz at all:
Allen is through untouched.
Schofield actually does a nice job to adjust and kick out Bullough, giving Smith a crease when he breaks the tackle.
So that's a problem. Michigan endures another half-dozen of these throughout the game, gets the ball back down seven with under five minutes left, and comes out empty.
Molk head down, Molk head up…
…instant snap with two LBs running straight up the middle of the field. This time Molk does block Allen; Schofield does not slide over to get Bullough, which would put someone else through but someone else not running up the middle at the snap.
Denard throws a slant; Smith runs a hitch. Ballgame.
Video of that:
The timing of the snap is the same, the result different.
So what's going on here?
While some of the timing issues may have been playclock related, neither of these are. Michigan snaps the ball with around ten seconds left on the first play and while there is no playclock listed on the second it was the first play of a drive and I don't remember being upset about getting the play in. This is just… like… voluntary.
Once or twice Michigan did go to longer counts and got the opponent to jump, but one of those was a hard count from under center. The fact that they could get the jumps meant MSU was timing the snap; the fact they could continue into the fourth quarter meant Michigan was using the long counts too infrequently. Michigan
- consistently tipped their snap count
- never motioned for the snap to reveal what the defense planned
- didn't even bother to pause after Molk got his head up so he could evaluate the guys coming hell-bent up the middle of the field
- did not check out of plays
- did not execute what looks like a hot read here
This is not a toughness issue. Air cannot block people even if you're the Clint Eastwood State Fightin' John Waynes. It's an inability for Michigan to deal with a simple, grandiosely unsound defense that leaves simple throws in the middle of the field wide open*.
All of this is coaching at some level, but we can separate out getting execution out of your players from strategy. On the interception Michigan had an answer that they did not execute, which can reasonably be chalked up to transition/mindflub/one of those things. Michigan QBs passing up wide open guys on that second quarter drive is execution, not strategy. Those are costs of installing a new system, especially one with a lot of post-snap reads for the WRs, something I don't think Rodriguez ever did. On some level that's understandable.
However, they failed to adjust their strategy to help the offensive line out. MSU is running full speed at the line on the snap; varying the count would make those well-timed blitzes poorly timed, allowing Michigan to slide the protection and letting Denard know what he's in for pre-snap… or forcing MSU out of the play. Michigan State timing these snaps so precisely puts immediate pressure on Robinson, robbing him of a half-second he needs to maybe see Koger on the other side of the field or the actual route Smith is running. It gives Smith more time to read the play and understand his hot route. Even if you want the double LB blitz on the INT because you think you have it beat, waiting that beat lets everyone on the offense know it's there without letting MSU check. At the very least make your standard count long enough for Molk to look at the situation in front of him before he doubles on a guy who's going outside because of a blitz.
I find this incredibly frustrating. This was an inexplicable Rodriguez-era problem canning him was supposed to solve. Instead it got worse. Hoke tried to explain away the snap issues…
Did you notice that they were jumping your snap count? “I think everyone has an idea of snap counts from guns, because there’s a mechanic that every team has. We have a silent count, and we have a double silent count. I don’t think that’s all the way correct.”
…but clearly there is something there that is bloody obvious to the opposition that has destroyed Michigan's offense against MSU on their last two trips to East Lansing. (Michigan moved the ball fairly well in last year's matchup only to be undone by turnovers.) The next time Michigan visits they'll presumably be in more of a MANBALL offense with Gardner better equipped to go under center and a line that probably reads Lewan-Bryant-Miller-Kalis-Magnuson, so we may have seen the last of this.
*[I was just reading that Smart Football post he linked about matching short passes with runs, which would have been perfect here. A-gap blitz? Immediate toss to slot/TE. Still need to block up the middle to get the QB some time.]
We complete our Pulling Mike Schofield trilogy with a reason he may have been so eager to get outside on the play action Tacopants interception: he spent the week getting coached up on getting outside.
You might remember last week's complaining about rollouts that only seemed to get Denard pressured. Borges attributed some of this to "protection." Those issues were on the pulling guard, who twice saw two opponents coming up at Denard, slowed up in a moment of indecision, and blocked no one. Once this was Schofield; once it was Omameh. Schofield's moment of indecision:
Denard had to pull up; he threw off his back foot and the ball sailed over Koger's head.
One week later Michigan went easy on the rollouts so those edge guys aren't so eager to get outside. Their first of the game came deep into Michigan's second drive. Hopkins picked up a questionable holding call on first down and Gallon got eight of that back on a throwback screen, setting up second and twelve from the Northwestern 44.
Michigan comes out with two TEs again, this time in a balanced formation. Northwestern undershifts their line and leaves a strongside linebacker to the outside.
Michigan runs a play action with a mesh point and Schofield pulling across the formation.
By the time the mesh passes, ludicrously aggressive Northwestern linebackers are ludicrously aggressive. The MLB threatens to shoot up in the gap to the outside as Denard pulls.
This time instead of being hesitant, Schofield deepens his pull:
He's moving away from the LOS, and he's not worried about anyone except the backside DE. As Denard sprints out to the corner he makes contact…
And seals the guy. Linebackers can blitz all they want now.
They aren't catching Denard.
He would pick up eight yards before running out of bounds; three plays later Schofield zoomed outside on the Tacopants INT.
This is a bit easier for Schofield. On the Minnesota play there is a second guy scraping to the outside who he has to deal with. Here the outside guy is the outside guy and that is that.
This is still an obvious coaching point for the week. Schofield is headed hell or high water for the corner, not paying attention to anything that may show inside of him and focused on getting that outside guy sealed. As soon as that guy declares himself the contain man Schofield is directed to lock in and get outside of him.
I would not be surprised if the obvious coaching point was in some way responsible for Schofield getting too far outside on the Tacopants INT. It's a play action pass on which he is pulling a week after having the thing that happened against Minnesota happened—negative reinforcement—and three plays after the above play happened—positive reinforcement.
Schofield's in a tough spot as the designated backup at every position except center. He should be the top backup tackle, full stop, and only pull rarely (to date tackles only pull on some outside power plays and the sprint counter). Instead he's starting at guard and having a lot of very complicated things to do on a regular basis.
It's interesting that Michigan chose to pull him over and over despite his relative inexperience. Have they lost faith in Omameh's pulling to the point where they won't run behind Lewan, the most grinding of their OL? Is it because Denard is more of a threat rolling out to his right? Something in Northwestern's defense? I don't know.
This is a pretty impressive drop from Northwestern's #51. On the video it's clear he's steaming towards the LOS, reads what's going on, and drops into Koger's route. Without that Robinson can toss it to him for what looks like a certain first down.
When Michigan can manage this seal good things will happen. I'm not sure they'll be able to consistently. When they can Denard on the edge with receivers is going to equal big gains. The problem with the rollouts against Minnesota was they never got that edge.
That seems like something a defender can force by getting upfield far enough that he won't be sealed. Shooting the gap between a guy maintaining leverage and pressure from the inside is a potential sack. So you have to have something in your arsenal that makes the guy think twice. Not sure what that is; maybe Borges does.
Last time we saw Michael Schofield run by a blitzer coming up an interior gap. That combined with a panicked back-foot throw from Denard to result in an interception on a play that had otherwise opened one of two receivers up for an easy touchdown.
This time we're going to get an almost identical play from the offense, except instead of play action is it QB power. This is the fourth and one Michigan converted en route to the endzone.
The setup is the same: shotgun with twin TEs and twin WRs. Northwestern lines up in an even 4-3 with one of the linebackers over the slot and a safety rolled into the box. For fourth and one this is fairly conservative:
With Denard running the ball Michigan has a blocker for every opponent.
On the snap, Schofield pulls…
…and the SLB blitzes, hell-bent for the gap between the playside DE and DT, both of whom are doubled:
Faced with a similar situation on the last play, Schofield ran by the linebacker:
This time not so much.
With both linebackers gone—the other one ran into the line on the backside—and a double on the playside DE, once Smith kicks out the corner it's an easy conversion.
Items of Interest
Being the pulling guard seems a lot more complicated than you'd think. A lot of power blocking is derp simple: block down on this guy. By contrast, everyone who runs a zone system talks up the need for their linemen to be intelligent because to run the zone you have to make a lot of split second decisions about who to block and when to release.
On these two plays we've seen what happens when a pulling guard gets challenged from a gap he doesn't expect to be threatened. He can miss it, at which point rivers of baby blood, or he can adjust, at which point your unsound defense has put the QB one on one with a safety for bonus bucks. He's got to have the vision and agility to pull that off. That's tough.
This seems like one of the major problems with the pulling scheme: the guards are crappier at it than the defenses are at defending it. Last year when they pulled out power blocking, defenses were trying to defend the zone and often got caught off guard. This year Michigan does not have that luxury. As a result we've seen a lot of plays on which the pulling guard gets caught up in some wash or just takes a bad angle to the hole.
"Adjustments." Is this an adjustment, or is it just telling the guard what he did wrong and not to do it again? In my view, an adjustment is changing your scheme to combat something the other team is doing—like throwing Ryan out on the slot to prevent argh bubble death. Telling your players how to stop screwing up is coaching, but it's not adjusting. What I was trying to say in the game column was that because of the nature of the offense they didn't have to do much adjusting, they just had to stop screwing up, at which point points fall from the sky.
This is not black and white. Borges did bring out some actual adjustments, like using Shaw to get the edge on theses aggressive linebackers, but I think the second-half turnaround was less figuring out what Northwestern was doing and stopping it than having a few specific players fix things the scheme is already telling them to do.
Short yardage numerical advantage. Not running Denard on short yardage is a goofy idea. Here you'd have to be nuts to not run the guy. He gives you the ability to double the playside DE and still block everyone except a safety rolled up. He has to be cautious because if he misses it's six points.
Handing it off, even on a zone read that should occupy some defenders, runs the risk of the defense selling out and Denard missing a read. Going under center takes away one of those doubles and turns the read into a call-and-hope situation.
I can see running conventional stuff in a low-leverage situation like first and goal from the one, sure. Keep the wear and tear down. When it really matters, this is the way to go.
Perfect mirror. This is a perfect mirror of the play that Denard got intercepted on, which is why the latter suckered Northwestern so badly and would have likely resulted in an easy TD if Denard can buy some time or Schofield makes the adjustment.
So… it wasn't necessarily as crazy as it appeared when he threw it. Is this good news? Maybe. It seems that Denard had one major problem in the Northwestern game, which was throwing off his back foot.
- Inaccurate but complete TD to Watson
- Interception #1
- Interception #2
Robinson had time to step into the some of the above throws throw but did not. Other times he didn't read the play fast enough and got pressure because of indecision. When not throwing off the back foot he was his zippy 2010 self; when he did it was armpunts away.
Sometimes you have to throw it off the back foot. These times are when there is a guy in your face and you have a really wide open receiver. None of the above are events that fit that profile. On the first he does have a guy really wide open but also has time to step into the throw. On the second he also has time to step into the throw. On the third he doesn't, and that's what this post is about.
Interception #2 exposed some of Robinson's flaws as a passer but it still should have been a touchdown. Michigan has a second and six on the Northwestern 16 after Devin Gardner's tricky rollout of the Denard jet action turned into a scramble. They come out in a common set for them, shotgun with twin TEs:
On the snap Denard moves towards the LOS and Schofield pulls. This will turn into QB Oh Noes.
As Denard withdraws into a passing position Koger releases downfield; Smith will head out on a wheel route. Both of NW's linebackers are headed upfield:
At this point you have two guys trying to cover two Michigan players, One of them is Koger, who will run a post. The other is the flat-footed corner on the LOS.
This is the key frame. Smith is gone past the blocker. The safety is similarly flat-footed against Koger, and Schofield has run past the blitzing SLB to double a defensive end:
This is all kinds of touchdown except for Schofield running past the gap in Michigan's line:
Without this linebacker getting in Denard's face the safety faces a choice between leaving either Koger or Smith wide open for six points.
But linebacker is in Denard's face, forcing an early throw off the back foot…
…that does not end well.
I think there was a bust in the Wildcat secondary, possibly by this safety, because Koger is open for an easy TD and the pressure cannot be anticipated. If the safety is going with Koger this is still incomplete. Denard overthrew it by five yards because he chucked it off his back foot.
Items of interest
This is definitely a protection the pulling guard is expected to make. On fourth and one later in this half Schofield will pull and correctly read this gap, then fill it, opening up the first down.
When Denard throws off his back foot, rivers of baby blood flow from my eyes. This was a thing that Michigan evidently got fixed in the second half when Denard was 8/9 for many many yards, but it threatens to pop up whenever the opponent gets a little QB pressure. The Watson one is the worst: no one is even in position to hit you after the throw.
This is not actually an insane read. I think his assumption was that the S, being the only guy on that side of the field near Koger, would go with him and this would leave the wheel open. The key moment:
He's not staring Smith down. He's looking at Koger and naturally assumes the only guy with a shot to cover him will take the hint. This was wrong in the same way it can be difficult to play poker against someone who doesn't really know what they're doing—they do something very very bad that turns out well because you didn't expect them to have a pea-sized brain.
Again, because of the back foot stuff this was five yards long and would have been incomplete in a best-case scenario. Robinson should probably just take off when things like this happen instead of doing this.
Needs moar play action. The super aggressive Northwestern defense was super aggressive, as you can see here. When Michigan went to QB play action it invariably got dudes vastly wide open, and while Michigan didn't have much luck getting these things completed, the passes are easy (seam to Koger is too high) or the problems easy to fix (block that guy, Schofield). A good chunk of the issues running the ball were on these aggressive linebackers—Michigan doesn't seem to make them hesitant. Maybe right after scoring 42 points while turning the ball over three times isn't the best time to bring this complaint up.
Last time on Picture Pages we looked at a 35-yard iso on Michigan's first drive against Minnesota. A 35 yard iso means someone did something terrible on defense or your tailback did something ridiculous; Michigan was playing Minnesota so it was the former.
So Michigan scores a touchdown and gets the ball back and opens up with the same set. Minnesota again has both safeties rolled up.
On the snap the line pass blocks but the backfield executes a historical reenactment of The Battle Of Minnesota Sucks At Isos.
There's a gap in the line that Hopkins is thundering towards again and by the time it's clear Denard has the ball the three LB type substances have started moving towards the LOS:
When Hopkins hits the LOS the two guys who could hypothetically cover him are four yards from the LOS and stationary.
This is what it sounds like when doves cry.
Items of Interest
Constraint theory right here. Minnesota just got hit with a big iso and got chewed out on the sideline about it. They are hyped up to stop it, so when Michigan shows it again the LB and S suck way up and leave Hopkins open for a big gain over the top. This works not only because Minnesota overreacts to it but because of the omnipresent Denard threat posed in the shotgun. That means the Gophers are operating with essentially zero deep safeties.
This is what happens when you can force the defense to cheat. An actual opponent playing this way probably would have given up five, not 35, on the first iso, but that's enough to force them to cheat to it, whereupon bang.
This isn't unique or new. Literally every program in the country except Ohio State* tries to do this. Michigan's old-school waggle is an example. The hope with Borges is that he'll use them frequently to score lots of points instead of occasionally as part of a bler offense run by 70s thinking. Denard is a hell of a thing to try to stop without cheating, much more threatening than Michigan's four-yards-and-a-cloud-of-dust running game was in the late 90s and early aughts.
This is the stuff you get in the shotgun even when Denard is not running because the zone read demands attention at all times. Under center Denard's legs mean waggle or Incredibly Surprising QB Draw, neither of which forces safeties into the box.
We saw a bucket of constraint against the Gophers. This is in part because the Gophers are so bad they have to cheat every play in the hopes of stopping the opposition.
I think it's also in part because Borges is finding his legs in this strange environment where your quarterback is one of the most threatening rushers in the country. Michigan tried the waggle a bunch earlier this year (and in spring) and got little out of it; against the Gophers that was ditched in favor of plays that use fear of Denard's legs as a major component of their success.
Hey man, nice touch. Denard floated a nice catchable ball to Hopkins about 18 yards downfield. He could do that because there were no safeties, so I've got no problem with the throw.
Let's see if Shaw can block this. Because this is a 75 yard touchdown if a fullback isn't running it. Also Hopkins almost dropped this.