alternate headline: man does job
You're going to have to bear with me on the offensive UFRs this year. The last time I saw a traditional gap-blocked, regular-ol-QB offense for anything more than a one-game debacle was ten years ago. That was the first year I did UFR and most running plays of that sort were deemed "another wad of bodies" because I didn't know what I was looking at. Since then:
- Two years of Debord running almost nothing but outside zone
- Three years of Rodriguez running inside and outside zone with a little power frippery
- Two years of Hoke trying to shoehorn Denard Robinson into a pro-style offense, giving up, and running a low-rent spread offense
- Al Borges's Cheesecake Factory offense that ran everything terribly
- Doug Nussmeier's inside zone-based offense.
I've seen plenty of power plays. Most of them were constraints that could be run simply and still succeed because the offense's backbone was something else. The rest were so miserably executed that they offered no knowledge about what power is actually supposed to look like. I watched a bunch of Stanford but not in the kind of detail I get down to with the UFRs.
One thing that I am pretty sure I think is that the popular conception of power as a decision-free zone in which moving guys off the ball gets you yards is incomplete. Defenses will show you a front pre-snap. You will make blocking decisions based on that front. Then the defense will blitz and slant to foul your decisions and remove the gap you want to hit. If you do not adjust to what is happening in front of you then you run into bodies and everyone is sad.
What Stanford was great at was running power that was executed so consistently well that it was largely impervious to all the games defenses played. This requires linemen who are downblocking to think on their feet, maintain their balance, and stay attached to guys who may be going in directions they were not expected to. It requires everyone off the line of scrimmage (tailback, fullback, pulling G) to see what's in front of them and adjust accordingly.
Michigan did a bad job of this against Utah. They also got blown backwards too much, complicating decisions for the backfield. The latter was not a problem against a much weaker Oregon State outfit. The former was much better, and that's the most encouraging thing to take from this game.
Here's an example. It's a six yard run in the first quarter on which Oregon State sends a blitz that Michigan recognizes and thwarts. There's no puller on this play; I think it was intended to be a weakside iso that ends up looking not very much like iso because Michigan adjusts post-snap.
M comes out in an I-Form twins formation; Oregon state is in a 4-3 that shifts away from the run strength of Michigan's formation. They are also walking a DB to the line of scrimmage:
By the time Michigan snaps the ball this DB is hanging out in a zone with no eligible receiver while both WRs get guys who look to be in man coverage. This is not disguised well unless the highlighted player is Jabrill Peppers and can teleport places after the snap:
He's going to blitz and the DL is going to slant to the run strength of the line. Michigan will pick this up, and I wonder if they IDed the likelihood of this pre-snap. No way to tell, obviously.
On the snap both the FB and RB start to the weak side of the formation; you can see Erik Magnuson start to set up as if he is going to execute a kickout block on the defensive end:
With the blitz and slant from the Oregon DL that's not going to happen. Each Oregon State DL has popped into a gap. Kerridge is taking a flight path to the gap that would normally open between Magnuson and Kalis, the right guard, on a play without this blitz. Without the blitz the DE would be the force player tasked with keeping the play inside of him; Magnuson would have a relatively easy job as he and the DE mutually agreed on where he should go.
Here the DE threatens the play's intended gap. Magnuson can't do anything about that. The D mostly gets to choose what gap they go in, and it's up to the offense to roll with the punches.
Michigan does this:
A moment later Magnuson has changed his tack from attempted kickout to an attempt to laterally displace the DE using his own momentum. Kerridge has abandoned the idea of hitting the weakside B gap and is flaring out for the blitzer.
Now, this could be successful for Oregon State still. The slant got five Michigan OL to block four guys. Nobody got downfield; the slant got a 2 for 1. But their MLB has stood stock still for much of this play, and Magnuson ends up shoving his dude past the hash mark—+1, sir. This is a ton of space to shut down, and De'Veon Smith is the kind of back that can plow through you for YAC.
Smith fends off the linebacker with a stiffarm and starts gaining yardage outside; it could be a good deal more but Chesson misses a cut* and the DB forces it back, creating a big ol' pile:
Second and four sounds a lot better than second and eleven.
*[Drake Harris will later pick up a 15-yard penalty for a similar, but more successful cut block; the genesis of that flag is probably Gary Andersen doing some screaming at the official after this play.]
Items of interest
You don't get to pick the gap even if it's gap blocking. Defenses slant constantly, and often in a specific effort to foul the intended hole and pop the back out into a place where an unblocked guy can hit. Post-snap adaptation is a must for a well-oiled power running game.
Slants win if they suck away an extra blocker. I would be peeved at the MLB if I was Oregon State UFR guy. While Michigan adapts to the slant well enough to provide a crease for Smith, the blitz means Michigan had to spend a blocker on the defensive back and the MLB is a free hitter. He should be moving to this more quickly than he does.
Slants also tend to open up giant running gaps. Adjustments like the above will often lead to a defender running in one direction suddenly getting unwanted help from an OL. If the OL can redirect and latch on just about everyone is going for a ride here. Once Magnuson locks on and Kerridge targets the DB these are two blocks that are easy to win and Smith is going to have a truck lane.
Given how much space Smith has even a linebacker playing this aggressively who shows up in the gap might lose or get his tackle run through; Michigan's getting yards here, whether it's three or six or more if Chesson gets a good block.
In the past this site has seen arguments about whether meeting an unblocked safety at or near the line of scrimmage is a win for the offense or the defense. I have largely come down on the side of "that absolutely sucks," but when the hole is so big that the defender is attempting to make an open-field tackle it's a lot more appealing.
Michigan WRs need to be more careful with the cut blocks. You can cut a guy from the "front," by which the NCAA means the area from 10 to 2 on a clock. (Seriously, that's the way it's defined in the rulebook.) This was very close to a flag, and Michigan got one later.
I wish Michigan was running pop passes, as those are good ways to get defensive backs hesitant about running hard after plays like this. Maybe in a bit.
Despite some post-burial kicking at the ceiling, Jake Rudock's pick six was the final nail in Michigan's coffin against Utah. It came on a route that I've called a "circle" for a bit now. The idea is that you run a slant, then abort that halfway through into an out route. Corner jumps the slant, you get some nice separation and hooray beer. Or you run an out, corner jumps the out, etc.
The general idea is that it is a horizontal double move. I've called it "circle" probably because NCAA football did back in the day; you can see that on a successful one the WR does tend to run in a little circle after his first break:
Both Utah and Michigan tried to run these routes on Thursday, with different results. Here are those plays… AT THE SAME TIME.
On the left will be a Utah throw on their first touchdown drive. It's second and six; Michigan is in the nickel they ran the whole day, showing press coverage on the outside.
On the right, Michigan attempts to convert a third and three halfway through the fourth quarter while down a touchdown.
As far as we're concerned these plays are completely identical to start: we are looking at the slot receiver to the bottom of the screen with a corner who is locked up in man coverage three yards off the line of scrimmage.
A couple moments after the snap both WRs have crossed the LOS; the only difference in the corners is that the Utah guy has taken a step forward, perhaps anticipating this route.
[After the JUMP: everything goes fine because HARBAUGH? Probably!]
Michigan's run game started out with a thud, with a series of short gains and even the occasional Dread Pirate TFL rearing its ugly head. As with the Notre Dame game, the problems due to were a mélange of errors from lots of people. And as you might expect against Miami, most of them were mental issues.
People have asserted that Miami was dropping an eighth guy in the box and that guy was blowing up the Michigan run game. That's simplistic; these days spread-oriented offenses are looking at one or zero deep safeties on every play. The eight man box is something you have to deal with as a coach, and anyway when you're playing Miami it shouldn't matter.
Michigan's issues were largely assignment-based, with the occasional bad block thrown in; the tailbacks were better but still had issues. The nice thing is that as the game went along we saw Michigan correct some of those those problems and start moving forward. Mason Cole in particular was evolving right on the field, hampering two plays with errors and then executing in near-identical situations just a few minutes later. One was mostly executing a block; this one was about IDing the guy he needs to address.
Which Guy Needs Help? Not That One.
This is actually the first play of the game. It comes from the 50 after the least interesting successful Dennis Norfleet kick return ever (run to the right until the kicker stops you), and Michigan comes out in an ace set. Miami has a 4-3 under on the field (sort of; their SAM is 190-pound Lo Wood) and will roll a safety down for guy #8.
Michigan's going to run inside zone and things are going to go pretty well all over the field with the exception of Mason Cole and AJ Williams trying to handle the backside DE.
This is your presnap setup:
There's a one-technique NT and a five-tech SDE. The SDE is splitting Cole and Williams down the middle, and the play is going to the top of the screen. It is very hard for backside blockers to do anything with a guy who is 1) lined up playside of him when 2) they get no help. This is about to happen to Williams.
I'm not entirely sure why this guy is free to fly down the line and blow this play up. The other DL are handled by Michigan's OL driving guys lined up a half-step to the playside of them. It seems like he figures that the OLB is going to be there to clean up anything that breaks behind him, so gap integrity is for suckers. (Michigan will get a bunch of waggles off this tendency, as Miami isn't using that OLB to contain hard upfield.)
On the snap Magnuson hops over a half-gap to get the nose; Cole goes to his zone a gap over without touching anyone and then starts helping on Magnuson's block. Williams is going for this DE:
Williams does not get the DE even a little bit, and with Miami DL set up to the outside on the frontside of the play Green is correctly going right up the gut, something that looks promising as Michigan gets movement on the DL.
Williams was put in a tough spot here; even so this feels substandard. Annoying the guy, pushing him so that he's on the correct side of the LOS, maybe cutting him: all of these things are better than escorting him to the RB.
A yard later it's clear that Michigan has cleaned out the DL; one linebacker shot a gap and is going to help tackle here but without the space constriction provided by the Williams block-type substance shooting that gap is a dangerous game to play that is 50/50 to put the tailback one on one with the safety for all the yards.
Green falls forward for two as Magnuson finishes pancaking the NT. Cole ended up not really doing much of anything on the play.
And the slow version:
[After THE JUMP: something that goes better]
Ryan under the microscope [Eric Upchurch]
Hello. As per usual, a game against a tomato can causes me to dig up something negative because I figure that the bad things that happen against weak teams are more likely to recur than the good ones. I'm not being negative, I'm being useful!
After this opening paragraph it may not surprise you that I didn't think Ryan had a particularly good game as Michigan's MLB. There were a couple of opportunities to contrast him with Desmond Morgan on similar plays that didn't come out well for Ryan. To the stillmobile!
Taking on blockers
App State had one drive of any consequence before Michigan started throwing third stringers on the field. That was a 75-yard march on which they ran an old Rodriguez staple, the "belly," repeatedly for good yardage.
Belly is designed to attack the soft underbelly of the backside of a defense facing inside zone. The end gets optioned off and then the goal of the defense is to use the backside DT's natural desire to shoot the gap to the playside against him. This usually sees the backside tackle get a free release on a linebacker on a quick-hitting play. (A quick google search indicates that this is Rodriguez-exclusive terminology, so your local guru's verbiage will vary.)
This was tough for Michigan to defend as aligned because the backside DT saw zone action and went GRRAAAH at it, driving himself way out of the play because he's Willie Henry and he is 1) strong and 2) not yet super disciplined. This put linebackers in bad spots, facing free OL while trying to shut down a ton of space.
Here's Morgan in that situation:
It feels like Michigan is a little misaligned here, with the linebacker shaded to one side against a formation that has no TE.
On the snap Beyer is let go and must respect the keep, so he flows upfield. Henry will get his own momentum used against him and get way out of the play, which I have designated by putting a frown at the end of his line. Morgan has an OT coming at him and a problem.
Beyer plays the mesh point well, inducing a give but forming up near the LOS so he can respond to a handoff. Henry is about to leave.
Here is the the key thing for Morgan on this play: he takes the contact. He in fact initiates the contact despite not having much forward momentum (which it is hard to get on a quick hitting play like belly). He impacts the OL and rocks him back:
Note that the guy next to him is Henry, who is trying to fight back to the play by giving ground. Also note that if Henry was anywhere near where the line would like him to be, Beyer is tackling as people wall up.
The back actually bounces off the OL…
And then a bunch of guys tackle him after six yards.
This is not a good result and I think Morgan's original alignment had something to do with that. He ends up taking the block to the inside instead of square and that gives the back room to the outside when otherwise this could have been a third down coming up. But: tough job in a lot of space. I gave him a half point for slowing down what could otherwise have been bad.
[After the JUMP: Jake Ryan tries his hand.]
Michigan couldn't get yard one with the veer against Nebraska, and most of them ended up with an unblocked Nebraska player blowing up Gardner. It is time to look at them. For some reason. Why didn't I start a blog about 1980s hairstyles? 1980s hairstyles never make you want to rub your face in gravel.
I digress. The first one comes on Michigan's first drive. A late blitz has just seen a power O slanted to and blown up for a one yard loss; it's second and eleven on the 24.
Michigan comes out with an H-back and two tailbacks in a twins formation, which necessarily means that the slot receiver is not an eligible receiver. Nebraska responds with 7.5 in the box, with the gray area defender just about splitting the difference between Funchess and the tackle.
On the snap Bosch pulls and the gray area guy sits and stares the backfield down.
Michigan shows veer action with Kerridge leading Toussaint to the outside; Michigan blocks the playside end, which would mean they're expecting to option the slot defender except 1) Kerridge is out there, so they're using one of their blockers on him anyway and 2) Gardner does not appear to be reading him but something further inside, if he's in fact reading anything. Gardner's awareness of this slot defender seems to start after the mesh point.
You can see that Gardner's helmet is not pointed at the slot defender as he starts making his decision:
What's he reading? Is he reading anything? I don't know. it doesn't seem like it. Watch the video in real time to get a feel for it. Toussaint does react like a guy who might get the ball, juking the blitzer, so I guess they're reading something. What is unclear.
Meanwhile, Kerridge is expecting the slot guy to contain upfield; instead he shoots upfield inside of him hard, too hard for him to adjust to.
Gardner pulls and seems to sense a disturbance in the force now; he goes straight upfield.
Toussaint dodges the blitzer, running into Gardner; Kerridge is prone, Gardner starts stumbling, and his momentum is taking him into the chest of an unblocked LB.
It's now third and nine, and Gardner's soul is now worn 1% more.
Items Of Interest
Optioning no one. We're back here, in year three. Michigan has a rudimentary read option game on which their QB doesn't know what to do too often and gets plays blown up, but here we're back to last year's Alabama game, where the defense made it so that Michigan's option plays didn't actually option off a defender, with similar results. No matter what happens on the edge here, the play still spends Kerridge and Toussaint on one defender and leaves an unblocked guy.
It would be one thing if I'd ever seen this fullback on the edge thing work. I have not. At best it's wasted him as he blocks a guy shooting up on the edge who is trying to contain Toussaint; at worst:
I'm about to get some comments about how this is Gardner's issue or Kerridge's issue and that Borges can't be held responsible for the results of this play. Sure. Any one play can be traced back to some execution error by the offense.
These posts are an effort to explain trends I'm seeing in the offense with particular plays, though, and this kind of half-ass option is par for the course. Michigan cannot get the fullback to be useful on these read option plays, and hasn't made him useful for three solid years.
This is the kind of stuff Denard papered over by being Denard. Even when Michigan was eviscerating Ohio State two years ago, they weren't really optioning anyone and it was left to Denard to make the magic happen against an unblocked dude at the LOS:
Michigan was fortunate that was a freshman Ryan Shazier on one leg. When you don't have Denard and you've turned your quarterback's ribs into a fine paste already, you no longer get 41 yard touchdowns and instead your unblocked dude gets a tackle for minimal gain, or more likely a loss.
They've had Kerridge for three years now and Gardner that long and Toussaint that long and they still can't get them to execute a real option. Either they're not trying or they're not coaching. And either way…
How is this supposed to work? It seems like the idea here is for the slot guy to run himself upfield outside of Kerridge to maintain a force back inside and then for Gardner to hit the gap between him and the rest of the defense. Nebraska beats that idea by using the slot guy super-aggressively.
How do you make this play work? Nebraska understood that Michigan's formation meant Funchess was not eligible; the gray area defender had no thought of a pass and ended up blowing up the play. But you can still make this work since Nebraska is sitting so deep with the safeties. Michigan has two options here: shooting Kerridge at the LOS, leaving Toussaint to his own devices, or using Kerridge to attack the slot defender and put Toussaint on the edge into acres of space.
This is the kind of thing you could come back to later with a tweak and bust a big gain. Clearly there were no big gains on this day. This design isn't necessarily bad; the inability to see what Nebraska is doing and get rock to their scissors at some point is. I mean, if you get this again and block the dude the defense has no force player, which means you get a lot of yards. This move by Nebraska violates a cardinal tenet of sound defense and works because they win on RPS, and if you probe at what they're doing here you can beat that. Instead Nebraska just kept chewing up Michigan's offense.
Hooray covered slot receiver. Hooray. I will never understand the point of that. If Michigan had some package where the ability of the H-back to get to the backside of the play meant something, okay. Instead you get nothing and if the D recognizes it, as they seem to here, you're playing 10 on 11. Temporary voluntary red card.
Again, maybe this is some sort of genius but since I've never seen it do anything productive it just seems dumb.
So yeah, I concluded yesterday that the quick fire throw to Gallon in the second quarter was a presnap read, not a true packaged play, and then about two plays into the remainder of the UFR, Michigan runs the same thing with the cornerback showing blitz presnap and M runs it after he backs out. Timing: I do not have it.
To the screenshots!
It's first and ten; Michigan's trying to respond to Indiana drawing to within one and has first and ten on their own 23. Same setup as the last play: 3 wide, Gallon alone to the boundary, IU in their nickel-ish package. This time the corner is indicating blitz.
On the snap, though, he backs out. Gardner's checking on Gallon, seeing if he's got the hitch.
When it become clear that the CB is not coming, he changes his plan. Taylor Lewan has the same idea, as his initial reaction was to pick up the corner. He's flared out to do so, and now has to frantically try to get back into the play and block someone who's trying to defend what's actually occurring.
Which is a handoff.
Now: Lewan's flare has borked a couple of things. See 98 below? He is being doubled and will end up three yards downfield, where it is ideal for him to go in the eyes of the offense.
The other DT, though, is being single blocked by Glasgow and if he chooses can decide to go upfield of him to the outside, which he does. Glasgow locks him out and pushes him past the play; Toussaint has one lane straight up the middle. Bosch deals with a DE well, but there's no one for an ILB.
That's unfortunate, since a guy dealing with him is a big gain with the other LB on a pass drop he's convinced to take by Funchess releasing into the slot LB. Even if he doesn't take this the backwards direction of the other DT would make it difficult for him to get to the hole.
Lewan tries to recover; can't quite; Toussaint makes the guy miss, which gets him a decent gain before the shuffling backside end comes down from behind to tackle.
Items of Interest
Nope, totally packaged. Gardner's first option is the hitch in the event of a CB blitz, and he decides that it's not there, so he hands off. Post snap read determining run or pass is the Smart Football-approved definition of a packaged play. Packaged.
An old bugaboo. This harkens back to some oddities Michigan had in their plays like this a couple years ago. When they ran the inverted veer in 2011, Michigan would often block the guy they were supposedly optioning with the pulling guard, leaving Denard to beat a guy if Michigan was going to pick up anything. This happened not infrequently, but it seemed pretty weird that you'd run an option and not option anyone.
This is a version of that old problem: Lewan flares out to block the corner when Michigan has a plan to deal with that. They're optioning him and they still block him, or would if he didn't back out into pass coverage. That leaves one of the ILBs free.
The rest of the line blocks it like they should if the corner blitz was coming; seems like someone on the OL made an adjustment to the blitz IU ostentatiously showed and backed out of.
A minor place. I don't want to make any grand conclusions from these two plays. A number of programs from the Okie State/WVU Hologorsen tree will build large chunks of their offense around packaged plays; Michigan has only dabbled in this department under Borges. They're still dabbling. The OL blocks this goofy because they are not on the same page as the play concept. If the guy making the line check understands that the corner is never a threat, this is a nice gain that doesn't require Toussaint to get his Hart on.
This isn't the first time they've tried these things—I remember pointing out a package to Smart Football a couple years ago. That didn't work, and it seemed like it got put on the shelf because the coaches weren't enthused with Denard's ability to read post-snap. Devin executed both of these; could they have been a test run for Michigan State?
Probably not, but here's hoping they've got something up their sleeve.
Bosch check. This was about par for Bosch's day. He got decent to good movement on his guys, probably better than Magnuson on average. Hopes were consistently tempered by the guy next to him, as when Lewan latched onto a dude he all but threw the dude into his teammates, ninjas-attacking-hero style. Indiana's terrible. He'll get a trial by fire next Saturday.
An accidental RPS. The other ILB's pass drop here is an interesting offshoot. He's reading Funchess and sees him release, and so goes to cover, as Funchess blocks (or attempts to block) IU's hybrid space player. That leaves the OL five on five in the box, which should be a profitable situation.
Every time a DE shuffles like this I want Gardner to pull, which is probably irrational. I don't think he should, but I have this visceral thing where it's like GO GIT EM, because is anyone in that position really keeping up with Gardner on the corner? I say no, especially when you've got Funchess bothering the slot LB. All DEs are shuffling and Gardner has beat them all around the edge. Sometimes there's help out there; that is the only thing that keeps these plays down.
Triple option? On this play it's asking a lot of him to read the corner and then come across the field to read the end, so the pull here is almost certainly not something that he has in the toolbox anyway. So, no, just a single option I think. The read option nature of the play does let you sort of option off two guys, though, except "optioning off" the corner is really just nerfing the corner blitz.