Unpacking why there seems to be a lot of unread space. [Bryan Fuller]

Neck Sharpies: Why Don't You Read? Comment Count

Seth April 19th, 2022 at 9:00 AM

So yeah, I'm UFRing the Spring Game. Did you want recruiting roundups instead? Yeah, that's what I thought. Anyway you're not supposed to read much into spring stuff except I noticed they were still doing a thing that annoyed me a lot last year.

If you missed all that, there was a running question through these exercises whether Cade McNamara's keep reads were hot or if they were determining what their post-snap mesh results were going to be beforehand. This was true for RPOs as well as zone reads. And it made life hard because I couldn't tell for certain if the coaches were telling McNamara to give no matter what, or if that was just how he was reading things, or maybe he was reading something else.

With the benefit of an offseason and Sherrone Moore's recent coaching clinic, where he echoed Josh Gattis's claim that they put reads on all this stuff, I think I can say it was mostly on McNamara.

However there were a few instances where his coaches gave him a read that wasn't the thing that turned out to be wide open. And I wanted to call attention to that, not because Michigan's missing out on some easy yards, but because I think they're making a conscious decision about this, and…you're going to laugh…it kind of makes sense.

The Wide Open Bubble

Let's begin with this play from the spring game.

Maize has Darrius Clemons in the backfield (the one on top) as a split-back flanking McNamara with Donovan Edwards. Clemons takes an orbit motion to the opposite flat, but nobody goes with him. The defense is blitzing the slot safety, Caden Kolesar, behind a crash from the OLB, Jaylen Harrell. That means there isn't anybody for Clemons until the deep safety.

image

McNamara is clearly looking at Kolesar. Kolesar is clearly the slot defender in the best position to be running outside with Clemons. Kolesar is clearly blitzing. And yet McNamara gives. Why?

We're going to ignore the play's result (it's a risky but successful demolition of Counter GT by Kris Jenkins) to focus on what McNamara does after the handoff. First thing he does is look at Clemons. And probably feels kinda bad.

image

But it's what Cade does next that interests me. This is a trot, but he's clearly been practicing this as a rollout/QB run.

image

The look to Clemons suggests that's an option. But given his reaction to Kolesar, I think I can say with some confidence that McNamara wasn't reading Kolesar on an RPO bubble to Clemons; he was reading Kolesar on a QB keep with a pitch option to Clemons.

[After THE JUMP: There's also something every familiar about all this.]

-------------------

The Wide Open Keeper

Let's go back to another snowy day (really, winter, enough!). This was in late November, just before Thanksgiving, and Michigan was on its way to beating Ohio State 42-27. For this example we are removing the Cade factor, since J.J. McCarthy is in at quarterback.

It's a nice memory, Blake Corum getting down to the 6 yard line two plays before another touchdown (OSU contested that one, but not the next one). But you also might have wondered at the start if it was the right read.

Conceptually this is the same play as the one from the spring game, except the read is reversed: Corum's bash run is effectively the same play as Darrius Clemsons's bubble screen, while McCarthy keeping would have been the same result as the spring game, with JJ serving for Edwards behind Counter GT.

image

Ignoring the CORUM WHEEE!!!! result and just looking at the disposition of these defenders, the keep seems to be the right option here, right? The free safety made his intentions known when he scooted across with Corum, and the cornerback to the Counter GT side was dropping back. Unless the crashing backside end can catch up to McCarthy that's a touchdown.

But McCarthy isn't reading this like an RPO. He doesn't see the safety at all. He's reading this the same as McNamara was reading the one above.

image

Simple zone read. That's what you'd expect here, since there's a good chance the crashing end could take out the ballcarrier before the slow-developing blocks of Counter GT are made. There's no check to see if the defense has abandoned all hope of covering the run side. It's just a simple read. The same read that McNamara is making. A zone read.

So What?

Sherrone Moore recently put out an online clinic (accessible from Throw Deep Publishing) and this Bash from the OSU game was one of the plays he went over. He even noted, after JJ was running into nowhere, that there was a lot of space out there, but that wasn't the read. Duh, right?

The question isn't whether JJ made the correct read, but why these huge imbalances are being generated. The answer goes back to the construction of Michigan's offense, which has simple concepts for the players, and a lot of frippery to make defending it hard. Defenses started responding to this by turning the tables, forcing the read they wanted, and overreacting to the result. The frippery created imbalances, but the offense wasn't actually threatening all the things they were showing.

Technically both quarterbacks last year were freshmen (Cade had a redshirt in 2019 and 2020 didn't count against anyone's eligibility). The coaches weren't going to be putting too much on their plates. McNamara had a complicated passing offense to get down, and JJ was a true freshman. Something had to give, and that turned out to be zone reads, which had very simple rules and different sliders per QB.

As of spring, that was still something that you could mess them up with. Michigan's Blue team showed you how a defense can overload the playside and force a give right into it. Ohio State showed you how a defense can overload the backside and force an outside run into it. Only Michigan's blocking and Corum's ability to use it to take out the linebacker pursuit turned that into a positive play.

But that was last year, and they're not going to show you their next step in spring ball. Since they've got their own and their rivals' coordinators willing to abandon entire sides of the field to force reads into their traps, the logical next step would be for Weiss to start teaching his quarterbacks to recognize the traps.

What that looks like, hypothetically, is training the QBs (or since it's offseason now, having the QBs train on their own) to recognize when there's a reaction to that play's frippery and take what's there. In JJ vs Ohio State, that would be if you see the safety bailing on the mesh, KEEP IT AND BOOK IT to the other side to escape the crashing DE, with the RB taking that guy out if possible. As for McNamara in the spring game, if you're getting a double blitz to the side of the bubble, cancel the read and throw the bubble.

But adding reads is a big coaching decision. Added complications are multipliers for mistakes, which is why many coaches keep hard and fast to simple rules. Michigan will likely opt instead to build in plays that attack those spaces automatically. If you see they're flipping the safety and corner like that, run the back out and have JJ run the Counter GT. If you see they're crash/replacing on the option edge, throw a bubble.

At best, I think what they'll start adding this year are pre-snap reads for this sort of thing. RB motions can tell you if they're going to have one of these overreactions, and McNamara is very good at picking up the subtle clues that signify a hot read. Now that defensive coordinators have had a season and a winter to figure out Weiss's additions to the offense and how to counter them, they're going to have to do something. The yards are too good to pass up.

Comments

QVIST

April 19th, 2022 at 9:20 AM ^

Joe Moorhead talks a lot about QB bias to give/keep. It is obvious that Cade's bias is heavily toward give, which is fine so long as the defender isn't blowing up the intention of the play. I know with the team I coach for, we have 1 QB who we tell "give it 'til you can't" and two others we tell "pull it if you see space."

Good point on the presnap reads, too. It's the easiest form of RPO there is when you factor in a QB run play. Flare the back away from the run direction presnap. No movement? Throw it. Do they rotate? Execute the Q run the other way.

ex dx dy

April 20th, 2022 at 8:14 AM ^

I actually go back and forth on this, and played my comment mostly for the laughs. I think my main point was that the suggestion that Michigan's coaches don't know how to implement a QB post-snap read seems pretty implausible given the simplicity of the concept and its ubiquity in college football, and so the discovery that there may have been a method to the apparent madness is unsurprising.

outsidethebox

April 19th, 2022 at 9:31 AM ^

The good news here is that the offense showed, by the end of the year, that virtually all the options were good...there may have been better or worse ones but, nevertheless, there was goodness everywhere. For instance, on the opening TD against OSU with the fake flair to Donovan and give to Henning on the reverse-I'm pretty sure Edwards scores too if McNamara had thrown him the pitch. When the OL is playing at such a high level everything comes up roses. 

But here, you have to be brain-dead not to see the significant (additional) threat JJ adds in these situations. 

dragonchild

April 19th, 2022 at 9:33 AM ^

Seth, where would you place the slider for pre-snap motion to show frippery (i.e, coerce the defense to move to open up the play), vs. giving McNamara pre-snap intel (i.e., observe the defense move to take what they give)? Understanding that they’re not necessarily mutually exclusive.

UMich2016

April 19th, 2022 at 9:40 AM ^

This is great content.  Taught me something.  I feel like they are setting up this offense more for JJ's skillset, or at least constructing it so that it has the opportunity to run plays designed for a fast QB that can run and sling the ball on the run.

My belief is that JJ would thrive in a complex, fast paced spread scheme where read options lead to bubbles etc.  

Cade seems best in a traditional pro style offense with bootlegs and throwing from the pocket (not on the run).  

Weiss and Moore are likely designing an offense accordingly.  Hybrid.

MNWolverine2

April 19th, 2022 at 10:08 AM ^

Agree with this.  Last year, the defense was so solid, that it didn't make sense to make things too complex.  Turnovers are what could have blown games and as long as the offense was putting the defense in a tough spot, they could grind out games if they had too.

This year, they are going to need more explosivness with a leakier defense.  

Carpetbagger

April 19th, 2022 at 1:09 PM ^

Show me one thing that proves JJ would operate well in a "complex" offense. McNamara appears to be much more football cerebral at this stage of their careers.

In fact, I would argue that's the chief reason McCarthy hasn't supplanted McNamara. McCarthy is obviously superior in almost every athletic respect.

RAH

April 19th, 2022 at 6:21 PM ^

I think that JJ could thrive in such a complex scheme but I'm not sure he's ready for that level right now. After all, he is just in his second year and has very limited game experience. I think the coaches will gradually introduce the complexity and monitor the speed at which it is absorbed.

BluePhins

April 19th, 2022 at 10:11 AM ^

The beauty of this example is that it looks to be built for JJ but it actually increases Cade's options while maintaining all of the teams core strengths. JJ can 1. Give 2. Keep and flare 3. Keep. Cade can 1. Give 2. Keep and flare to wr 3. Keep (probably uneffective). You're essentially kicking the normal 2nd read (qb keep) to the 3rd. If you follow order of operations this play can be run with expected success by either qb, or even a rb or wr.

jdraman

April 19th, 2022 at 10:22 AM ^

Seth, is the play you examined from the Spring game not just essentially a triple option like what Oregon used to run in the Chip Kelly days? An inside zone read to one side of the LOS and a backside speed option?

If it truly is a "keeper + pitch" option on the backside, that's a little weird no? Neither of the QBs Michigan will be playing have true speed or agility to bust those plays for big gains. I guess that keeper is just for taking the easy small yardage chunks. But, that still leaves an uneasy feeling for me about why the bubble is not the main backside read in your estimation. We'll just have to wait and see. 

jdraman

April 19th, 2022 at 12:28 PM ^

I mean I think I agree, but Seth's diagnosis is the complete opposite: 

The look to Clemons suggests that's an option. But given his reaction to Kolesar, I think I can say with some confidence that McNamara wasn't reading Kolesar on an RPO bubble to Clemons; he was reading Kolesar on a QB keep with a pitch option to Clemons.

The issue you've raised of distance between the two players is not insignificant.

I *sincerely* hope that this is an RPO with a bubble screen on the backside and that Seth is reading too far into this "trot" that Cade does. That's why I said would feel uneasy about this play being a backside speed option akin to what those Oregon teams used to run. But, with how downhill run-heavy this team seems to want to be, I think we just have to wait and see what they do this fall during games. 

jdraman

April 19th, 2022 at 1:47 PM ^

I’ll take your word for it, but I’ve never seen that done. Does that also suppose that the QB can throw adequately off of a non-stationary platform in either horizontal (field) direction? I don’t know if we’ve seen much of either Cade or JJ to suggest they’re competent at that certain throw. 

Pumafb

April 20th, 2022 at 11:00 AM ^

It's done quite a bit. It's a 2 part read. The first read is the stand-up 9 technique. You either give the Counter GT or, if he squeezes, pull the ball. If the QB pulls, there is now a 2nd level read. In this case it's the down safety. If he widens with the bubble, the QB can take off. If he comes, throw the bubble. Teams do this all the time by attaching a bubble to the typical zone read to put the force defender in conflict. In this case it can be done with counter instead of zone. 

Space Coyote

April 20th, 2022 at 2:16 PM ^

Yup, you and Seth are correct. Both PSU and Nebraska ran it quite a bit last year. It’s a form of triple option, though the pitch portion of the play comes much earlier (usually) than the old fashion triple option play.

But you have to read the End first. He dictates if the Counter (or Dive, or whatever) should have numbers to work or not. If he crashes, it doesn’t (the End will cancel gaps behind the LOS) so you pull. If he attacks the QB or stays, you run the Counter (or dive or whatever).
 

Doesn’t matter if the pitch is open if the dive is open first, because it’s a progression

Gohokego

April 19th, 2022 at 10:30 AM ^

I don't mind if JJ gives more. He needs to stay healthy. Hopefully the shoulder is feeling good and the strength and conditioning program has him stronger and more durable.  He's a super talented kid but playing d-1 ball the hits you take running add up. 

NeverPunt

April 19th, 2022 at 10:49 AM ^

I don't think he needs to run that much to be honest which should limit injury potential. The biggest asset here is the threat of the run. If JJ is a legitimate threat to pull it and go and has the athleticism to punish you for ignoring him, now you've given a player responsibility to spy him who can no longer attack the play effectively. With his arm you are now at a major disadvantage.

If you know for sure that Cade is going to give the ball every time, that's not a serious threat, and even if he did pull it, he's not likely to take it to the house. Cade has other skills that make him a threat, this just isn't one, and an area where JJ has a leg up on him.

I like that they've included these other options as well. Give JJ a couple runs a game when he's in at QB and watch what it does to a defense. He doesn't need to be Denard and act like a RB at the position to be effecive.

mGrowOld

April 19th, 2022 at 10:58 AM ^

Great work as always Seth but, like most things in life, people can look at the same data and come up with very different conclusions.  I look at the play in the spring game (and the OSU game which we won 42-27 BTW) as evidence that the coaches remain firm in their pre-snap decisions and that the QB has precious little options to deviate from that decision.

And candidly, I think that's ok.  I get that we look at the plays where we basically refuse open yards but what we dont look at, because we cant, are all the screwed-up plays that never happned becuase the coaches employ the KISS (keep it simple, stupid) method.  I still believe that their operating philosophy is execute the correct play based on down & distrance to perfection and good results will follow most times regardless of the defense.  And it's the OCs job to call it and the QBs job to execute it.  I still believe that they want the QB focused on execution of the play, not focused on a multiple decision tree based on what they think they are seeing pre-snap.

jesse.knowsfoo…

April 19th, 2022 at 10:59 AM ^

This is why Cade should start. JH and company build offenses like an onion just keep adding more layers. With Cade at QB the coaches have the ability to open up the playbook even more because of how long Cade has been working in the system you can add more and more layers. JJ starts and they have to simplify the offense and bring him on slowing.  Cade's been working in the system much longer. 

outsidethebox

April 19th, 2022 at 12:33 PM ^

This sounds reasonable but it isn't how it actually works. Yes, execution is critical however, today, for Michigan to win against the more elite programs they have got to play smarter. Georgia made it abundantly clear that you had better be able to have more than a pre-snap read in your arsenal if you are going to make them pay. Even Saban, with all that talent, has abandoned the game-manager QB...it's not your ticket to the mountain top. Cade is an excellent QB, and he has his place, but he has got to bring more to the table than he showed last year.