Technical Flyover: What About the Quarterback? Comment Count

Ian Boyd September 15th, 2021 at 10:27 AM

One key detail seems to have Michigan fans concerned coming off their mostly positive, 31-10 drubbing of visiting Washington. Sure Washington was coming off a defeat against Montana, an FCS program, but Michigan ran for 343 yards and took control of the game after converting a fake punt on 4th-and-1 from their own 30-yard line.

Pretty exciting to have a coach approach the game aggressively and see the team respond by immediately imposing their will on an opponent.

Still, there was something missing from the performance...

To say Cade McNamara didn’t light the world on fire would be an understatement. His 7-15 day passing for 44 yards at 2.9 ypa was very low, to say the least. He added three yards on three carries, none of which were sacks.

He pulled the ball once on a tight zone play when the nickel blitzed and went around the open edge for nine yards, this was certainly a positive. Then he also pulled the ball on a GT counter-read play when the outside linebacker wasn’t really committed to the running back and was chased down and tackled for a loss of three yards.

pic.twitter.com/h7DMbH2VSn

— Asst to the Minister of Culture (@Ian_A_Boyd) September 15, 2021

It was good for the Wolverines to pick up a win in a game many would have penciled in as at least a very plausible loss before the season. What’s next though? Will this offense be able to maintain success against Big 10 defenses if the quarterback is producing 47 yards of offense? Probably not. How did it work so well in this game?

[After THE JUMP: opposing imposing wills.]

The Washington defense

The Huskies were not well equipped to handle this Michigan offense. For one, they were routinely baited by offensive coordinator Josh Gattis into playing their nickel personnel when the Wolverines would play in 11 personnel.

The Huskies’ nickel defense is typically a base 2-4-5 scheme comparable to Michigan’s own base set. Their “Daxton Hill” is Brendan Radley-Hiles, a 5-foot-9, 185 pound transfer from Oklahoma. Radley-Hiles was notorious at Oklahoma for being good in coverage, but often a no-show in the run game and prone to committing cheap penalties at the most inopportune times.

Now let me back up and explain why this matters. One of the principal threats to a spread offense is a dynamic nickel who can work off the edge unblocked and blow up your run game. Anyone this far north remember when Oklahoma won the National Championship in 2000? They used Roy Williams in this fashion and it was devastating. How about the 2011 LSU Tigers with Tyrann "Honey Badger" Mathieu?

Eventually teams begin to focus on RPOs to protect their runs and prevent good nickel players from disguising a late edge pressure which could inflict negative plays.

Michigan’s use of RPOs in this game, or quarterback run reads, was sporadic and not always well executed. Put more simply, they actually ran a number of RPOs in which no read seemed to take place. A team with a run-stopping nickel might have exploited these issues, but Washington wasn’t well equipped to make the most of it.

This particularly mattered on two of Michigan’s biggest runs of the game.

The first one, Blake Corum’s 67-yard run following the fake field conversion.

It’s a counter run scheme, pulling the center instead of a guard to kick out the edge player and then pulling the off-ball tight end over to lead block. The edge player was taken aback to find the center and was unable to spill the ball, yet the weakside linebacker headed to the edge expecting the ball to be spilled only to then be caught outside by the pulling tight end.

The big issue for Washington here is they got beat at the point of attack and had no safety support because they were rotating to the other side of the field. Corum outruns the deep safety and it's 10-0 Michigan. Fake punt! Harbaugh is a genius! 

Later in the game on 3rd-and-2.

They flex the tight end out and run GT counter-read, a play where they pull the guard to block the edge and the tackle lead blocks behind him. The backside edge is left unblocked for the quarterback to read. Again, Washington is rotating to the three receiver side to keep Radley-Hiles in man coverage outside, they contain with their edge play, and once again they lose the point of attack and their safety can’t clean up until Hasaan Haskins has picked up 20 yards.

The Huskies were outmuscled by Michigan on a variety of sweeps, counters, power runs, and straight ahead zone plays. Zac Zinter got the start at right guard and was terrific as a puller on counter, regularly trapping Washington’s edge players inside when they tried to spill runs. But the Huskies also had some really poor answers for trying to make the most of Michigan’s limited passing game.

McNamara’s 47 total yards were not an inaccurate reflection of his impact on the game. He didn’t make consistently correct reads in their various spread-option schemes and he can only do so much damage pulling the ball down himself to run the ball. Nor does Michigan want to run him regularly given his smaller frame and injury history.

The RPO and option game

The main reason a team like Michigan would be interested in RPOs is to protect their running game. However, you play a dangerous game when you turn to RPOs as a solution to protecting your spread run game from defenses sending numbers late into the box.

If you rely on the quarterback reading defenders for direction on whether to throw or hand-off, you’re essentially surrendering agency. The defense can either choose for you where the ball should go, in this case perhaps loading the box and daring Michigan to make perimeter throws perhaps against man coverage, or they’ll muddy the reads and complicate matters.

Today’s defenses include a lot of teams using “replacement” blitzes. They bring four pass-rushers and play base defense behind it, but the four pass-rushers include two to three defensive linemen and then one to two blitzers. Who’s playing where? It’s hard to be sure and it’s easy for the quarterback or offensive line to let guys free who can blow up a play. Michigan has this in the arsenal themselves in Mike MacDonald’s own 2-4-5 defense. Here's how it can look when done correctly:

https://twitter.com/Ian_A_Boyd/status/1437767325168685064?s=20

The edge player drops back to replace the Mike linebacker, the defensive tackle stunts wide to replace the edge player, and the Mike linebacker blitzes to replace the defensive tackle. Arkansas loses the plot, doesn't block the blitz, and a normally lethal quarterback GT-counter run is a negative play on their own goal line. This is what you face against higher level teams when you want to be all-in on the RPO game. Things get complicated.

It doesn't appear Michigan is really invested in getting complicated with their own RPO game. When they ran apparent RPOs against Washington the decision-making often seemed predetermined.

Washington is staying home on the swing screen here opposite another GT counter run. Is it really a run/pass option or were they trying to scheme an easy throw to Blake Corum in space? pic.twitter.com/yB6CQEAsRp

— Asst to the Minister of Culture (@Ian_A_Boyd) September 15, 2021

RPOs also don’t solve for short-yardage or the red zone. Everything tightens up and you can’t create space or run/pass conflicts very easily, so you need different answers for pounding the ball on the ground.

Michigan has the hard part down, they can run the ball from multiple formations and do it from spread sets or bigger formations with extra blockers on the line. They don't need RPOs to make the run game easy. Against Washington, Gattis also revealed a knack for scheming open opportunities for the run game even when the RPO or quarterback option threats weren’t serious.

Michigan’s challenge in future games is protecting the run game as teams wise up to the situation.

The obvious solution was to throw the ball to Ronnie Bell, which was also a useful solution for passing downs, on RPOs or play-action. Obviously this is no longer an option.

The rest of the Michigan wide receiver corps is mostly players recruited to be “speed in space” threats for the RPO game. Currently they’re pretty young and redshirt sophomore Cornelius Johnson is the elder statesman in the room. Additionally, it’s not clear Cade MnNamara thrives in the RPO game and again, he’s certainly not the guy for getting regular work on quarterback run options.

What made McNamara a good fit as the starter was his ability to execute Harbaugh’s passing down schemes in the dropback passing game and to hit Ronnie Bell. Now? His best utilization against Washington was to present the possibility of a backside keeper when Michigan ran GT counter-read from spread sets.

So from here, what does Michigan do? Future opponents are going to load the box against bigger formations while playing man coverage outside and will send late pressure against spread sets and dare McNamara to beat them in the read game (running or throwing).

The easiest way to maximize this team would be with a quarterback who can present a real threat in the quarterback run game, but Joe Milton is gone. Alan Bowman executed an RPO system at Texas Tech, but he’s still learning the Michigan offense and it’s not clear whether the Wolverines have been practicing or emphasizing RPOs in the offense anyways.

Then there’s J.J. McCarthy, who is a potential upgrade over McNamara in throwing vertical routes against man coverage outside because of arm talent and in pulling the ball around the edge due to some solid quickness. Is he an upgrade though in managing the offense? Helping to check out of bad plays? Protecting the football with good decision-making? Less clear. I'll venture a guess and say when he proves in practice he can handle 3rd-and-long without giving the game away he'll get his chance.

One thing is clear, from here on out the story of the Michigan season will be how they scheme ways from week to week to make the most of a legitimate power run game while figuring out how to set up their quarterback to punctuate drives.

Comments

DonBrownStache

September 15th, 2021 at 10:35 AM ^

I'm not convinced that Cade 'can't' make the necessary reads/throws when 3rd and long approaches in a tight ballgame. I will wait to worry about that when the situation presents itself, maybe against Rutgers.

Blake Forum

September 15th, 2021 at 11:02 AM ^

Yeah it’s not clear at this point what Cade can or can’t consistently do. He’s certainly hit some nice deeper throws in his brief career. I worry more about Michigan putting him in tough situations (like these clunky, predictable run/run/pass sequences Harbaugh teams too often fall into) more than I worry about his physical abilities per se

atticusb

September 15th, 2021 at 11:58 AM ^

I think the point of the comment is that relying on Michigan out-scheming opponents every week in order to win is a dicey bet.  Ian's write-up is excellent, and highlights the tight line Michigan must walk.  A dominant power run game on it's own is no longer enough to beat good teams... something must be added to that, and it's not clear what Michigan can or will add to, in Ian's words, "protect" the run game.  Frankly, I agree with Ian's take that we don't want McNamara to be a run-option workhorse... so if we're not going to risk it by taking the training wheels off in the RPO game, then I say we better mix in some traditional play-action/drop-back pass stuff.  I know the "we have to be modern"-ists will have a cow, but I'm not sure running McNamara or betting the farm on good RPO reads (against good defenses) are either good options.

gbdub

September 15th, 2021 at 4:09 PM ^

Ok but Cube is just griping. If we simply overpower overmatched opponents, we're a dinosaur that will get exposed as soon as we get a "real" opponent (a common complaint about this game). If we outscheme "lesser" opponents, well shoot now that's a bad thing too because it means we can't impose our will. 

atticusb

September 15th, 2021 at 6:03 PM ^

Yeah, I hear you... For me, at least, I'm happy to see a few of things:  (1) any Michigan coach formulate, implement, and then stick to a scheme clearly tailored to an opponent and *that works*, (2) that we, at least, have an O-line that can overpower a power-5 line with reasonable DTs, and (3) the RBs look talented and well-coached.  Beyond that, we've pasted two opponents who will be nowhere near top-25 at the end of the year.  Good!  This Saturday I have every expectation of pasting another sub-par team.  Good!  I see some ways that this team could be competitive with quality programs (read: Ohio State), and know that there are still plenty of unknowables that will ultimately determine outcomes of those games.... still Good.... but, as the UFR's have it:  results based judging, so we'll see. Of course, I can't wrap up a summary of my feelings without noting that a Suckeyes loss makes everything better!... imagine the bitching and moaning on their end!  :-)

mikegros

September 15th, 2021 at 10:54 AM ^

I agree that McNamera did not do a lot to inspire confidence, but I also think a lot of that was Washington didn't give him much option and, as Seth put it yesterday, the offense took what was there. 

You mention the bubble read to Henning looking like a pre-determined read in your last Twitter embed, but I'm not so sure. The end crashes in and Henning has numbers to the outside. The LB is far enough in that this should be at least 4 or 5 yards and potential for much more if Henning makes that LB miss. 

It ends up 0 because Wilson's block is very passive and his DB blows up the block and makes the tackle. If Wilson is more aggressive, he and the DB make contact upfield 2 or 3 yards. Henning has space to pick a side and try to make a move on the DB and trailing LB. 

I feel like this was not the only perimeter play where the issue wasn't the read, but that one of the blockers just got beat. This is another area where Bell is missed, he was a great blocker as well, but one that can be cleaned up going forward.

Ian Boyd

September 15th, 2021 at 11:17 AM ^

The linebacker is widening to the screen, which means his chances of having a positive impact against the GT counter play are slim to none. Presumably what you want in this offense right now is to run the ball whenever you have favorable angles and numbers. They have those here, so why no run? Hasn't the swing screen option already done its work?

The 1-on-1 block on Radley-Hiles, whom I noted isn't known for being a physical force on the perimeter, does go poorly. Even if it only went okay though, so long as Radley-Hiles kept the ball inside the linebacker was going to get there in a hurry.

I could be wrong but I think this was predetermined.

Pumafb

September 15th, 2021 at 11:39 AM ^

Spot on here. The RPO read is clearly that backside LB. He widens for sure. We have numbers with the GT Counter so this is a clear give read. There is no chance he can add his hat to the counter away from him. The block on the perimeter was horrendous, no doubt, However,e ven if that block is executed, they still have an unblocked defender there since the LB flew to the screen. It was either pre-determined and we are just faking RPOs (a really dumb idea) or it's a very bad read by Cade.

nscharer57

September 15th, 2021 at 12:26 PM ^

My opinion: This is a pre-snap read only. If you watch the beginning of the play closely, there really is no mesh point. Corum is past Cade before he has control of the ball and Cade's eyes do not appear to be looking at the LB or DE during that time. To me this is a read based on UW's pre-snap alignment. The UW safeties were playing so deep, this alignment basically puts Henning in a footrace with a trailing LB if Wilson's block is even just a stalemate.

If the Wilson's block is even a 50/50, the most likely outcome of this play is Henning being pushed OOB at the sticks or just short of them right by the outside WRs feet.

mikegros

September 16th, 2021 at 12:35 AM ^

Honestly, I think both reads are right here. The LB is in no man's land. He hasn't moved at the mesh point, just sunk his hips. With both safeties 16(!) yards downfield, there's a lot of space to maneuver.

The give to Corum does have a good angle, but the are only two pullers for 3 heads on that side, so Corum would have to beat someone to create yards. There's a good chance he does, but it's not a gimme: 

On the other hand, if either of those pullers misses their block, like Wilson does, chances are the run isn't going far either.

Speaking of Wilson, if he makes that block upfield a couple yards, Henning catches the ball with forward momentum and a couple steps to make a move. I assume the DB would take the edge and force back to the LB. But now you have Henning with momentum. Either the LB holds it to 4-5 yards by making a tough open field tackle on a quick WR, or Henning makes him miss and it's at least a first down, maybe more. 

If you imagine Wilson engaged with his man 2 yards further and forcing him to pick a side, this is exactly the sort of situation modern offenses love to get: your 4.4 jitterbug against their LB with space to make a play. 

Again, I do think either read was good assuming the blocks get picked up. They just didn't this time.

Blake Forum

September 15th, 2021 at 10:57 AM ^

Really interesting analysis, as always. It’s concerning that Michigan seems to be half-assing their RPO game. One thing I’ll disagree with is that the loss of Bell necessarily has to dramatically alter Michigan’s passing game. Baldwin and Johnson have shown enough to be credible threats in all kinds of situations, including against solid corners. And there’s young talent elsewhere, not to mention Erick All looking dangerous again. Michigan shouldn’t be afraid to let their receivers make a play. The unanswered question is what exactly Cade can reliably do  

Blake Forum

September 15th, 2021 at 12:14 PM ^

Baldwin not being at 100% is another valid datapoint, yeah. In this game, I think it’s mostly just that Washington couldn’t stop the run. It made sense to keep handing off. I think the situation with down-field receiving itself shouldn’t be too dire going forward, but that’s a projection rather than something I can prove 

zerocool

September 15th, 2021 at 11:16 AM ^

The writeup's from yesterday to today seem to be at odds with each other.  Seth's take from yesterday "There were a bunch of positive reads for McNamara in this game despite Michigan being as QB givey as ever, simply because Washington kept making the handoff the correct decision."  The take today "When they ran apparent RPOs against Washington the decision-making often seemed predetermined."  Hopefully the UFR will offer more detail to support one over the other (i.e. was McNamara making reads accurately or were the reads in fact predetermined).   

Ian Boyd

September 15th, 2021 at 11:24 AM ^

Seth definitely ain't wrong, I was trying to make a similar point but didn't do so in a particularly articulate fashion. Washington's nickel defense kept getting gashed by counter runs from spread sets because they wanted to play man coverage outside with Radley-Hiles and play contain with the edge defenders.

1VaBlue1

September 15th, 2021 at 11:16 AM ^

I'm not so concerned about Cade, but I need to see the WR group improve quickly.  I don't think they are very good at running routes yet, so they don't create a lot of separation.  Against probable NFL draft picks (UW's corners), they were shut down aside from a great back-shoulder throw to CJ (and his adj to it) and an underthrown ball to Wilson (5 more yards in the air and Wilson was gone).

Blocking on the bubble screens was also an issue, in that it was non-existent.  I think I could have done a better job than Wilson did!

Is there a chance that All might actually make a catch at some point this season?  I feel it's time to stop telling us how good his hands are.  His hands are butt (not that Butt).

Ian Boyd

September 15th, 2021 at 11:22 AM ^

The places where they line up All on third downs suggest they think he's one of their better receiving options on the team, so I'm sure we'll see it eventually.

Cade as a drop back passer executing some of Harbaugh's third down solutions is something I still have some faith in as well, it's just that A) they lost their best weapon there and B) the main emphasis needs to be supplementing the run game.

I wouldn't rule out Cade remaining the starter all year though and Michigan working out some solutions with him.

WormWould

September 16th, 2021 at 9:46 AM ^

All has already mae at least one reception - and his hands are most definitely not "butt". Yes, he had several bad drops last season, but that was psychological, not physical. Seems, so far, like he's got his head straight. Remains to be seen, of course, and maybe the yips return, but I'm holding out faith that he's on the upswing.

dragonchild

September 15th, 2021 at 11:30 AM ^

The rest of the Michigan wide receiver corps is mostly players recruited to be “speed in space” threats for the RPO game. Currently they’re pretty young and redshirt sophomore Cornelius Johnson is the elder statesman in the room. Additionally, it’s not clear Cade MnNamara thrives in the RPO game and again, he’s certainly not the guy for getting regular work on quarterback run options.

skatin@the_palace

September 15th, 2021 at 11:38 AM ^

Ian, curious to know if you have any thoughts on the groupings they ran out there on Saturday night? It would seem like you could cover a lot of bases if you could get All, Johnson, Henning, and Wilson on the field at the same time but putting Wilson in spaces where he needs to be the keystone block on a bubble didn't work on Saturday. Does that seem like a UW issue and how their DBs come down hard on perimeter defense (and play at a high level) or a true concern moving forward? 

Pumafb

September 15th, 2021 at 11:45 AM ^

I'm not Ian of course, but really it looks to be a matter of just not being physical enough. He caught the defender instead of initiating the engagement. His path was decent to start, but he needs to work a little vertical after the initial flat path to eat some grass and create a some momentum for himself before engaging . I also would prefer he work a bit harder to the outside shoulder and run the corner in or wash him upfield since the route behind him a is bubble and any help is coming form inside. Again, I don't think it matters here as the LB is flying inside out and is in a good position to end the play even if we get a serviceable block on the corner.

Ian Boyd

September 15th, 2021 at 12:16 PM ^

I think they liked the lighter, 11 personnel group with All on the field because they had a lot of good spread run game schemes and Washington was bad at defending it in nickel.

Against a team who's good at stopping it in nickel, maybe you see more 12 personnel or 22 personnel to try your luck against their "base" personnel on defense.

As for perimeter blocking, you could just have everyone run routes in quick game progressions rather than blocking for screens. This requires even more reads by the quarterback though and, like I said, it doesn't seem like he's even reading "hand-off or screen" plays.

My example on the GT counter to the field is a good one here. They run a slant/flat concept on the backside and the slant/sit is WIDE open but since they'd already distracted the backside of the defense, they didn't seem to care. Happy to run the ball to the field.

Those sorts of RPOs require an all-in commitment. You have to be practicing them all the time with the right footwork, quick reads, and then potentially multiple arm angles. Is Michigan that committed to it? I don't know, seems like they were with Milton, less clear with McNamara.

tkgoblue

September 15th, 2021 at 11:51 AM ^

This is a great article, but I do have one question. Why do we automatically assume Cade can’t execute the deep passing game and can’t thrive in an RPO offense. I understand everyone wants the 5 star eye candy freshman; however, I haven’t seen enough of Cade to completely count him out. The 4 deep balls I have actually seen him throw this year have all been on target and catchable. Now, this doesn’t mean he can do these things. I just don’t understand how we have enough information on him to say he can or can’t run this system. They obviously haven’t had to win by Cades arm yet. Let’s take a deep breath and see what he can do please ?.

tkgoblue

September 15th, 2021 at 2:45 PM ^

Definitely a solid point. I would say we saw him run the RPO offense to effect in the rutger game last year. The PSU game he got hurt on the second drive (he did move the ball until then FWIW).  Western he moved the ball well, but the RPO offense seems to have gone away. Washington, Ran the ball so well he was never really needed to make a throw more than 5 yards.  I just think this offense is completely different this year. More emphasis in the run. Less RPOs. Also, I would look to see if Michigan completely changes scheme every big game like they did in 2015. I remember other teams saying that what Harbaugh did in his first two years.

WormWould

September 16th, 2021 at 9:51 AM ^

No one seems to be mentioning the fact that we've only seen two games thus far, both of which have been in Michigan's control from early on, so they haven't had to open the playbook. Why show B1G opponents our RPOs with reads, when we can just predetermine a play and have success with it?

All of this concern over Cade and his potential inability to run an effective RPO is overwrought. Patience, wee ones. Patience.

charblue.

September 15th, 2021 at 12:46 PM ^

The point of playing the game is to perfect what you learn in film review, game plan study and practice. So, if Cade isn't the game manager as quarterback we've been all assured he is, or the best choice at his position at this time, then he will eventually be replaced.  Sometimes you just don't play as well as expected.

Listening to Devin Gardner review Cade's work on Saturday, he indicated that McNamara's performance against Washington may have been affected by a number of issues outside his own physical limitations. And those issues and slow decision-making apparently grounded the offense, literally and figuratively.

Whether he "handcuffed" Gattis play-calling because he was slow in making RPO decisions and was otherwise late in delivering the ball downfield, because of an off-night is clearly fixable going forward. Whether he's the man for the job can only be determined by his level of play going forward after critical review. Clearly, the coaching staff wants him to succeed because their other qb options have less time in the offensive system. But they don't have a choice either. Because either Cade does well and performs as expected or this team won't win as at the level we all want it to.

As of yet, this not a national championship caliber team. Michigan might win most of the games on its schedule. But we all know there is only one game that now impacts how its perceived nationally no matter how well it performs or plays in any other. And the general consensus coming out of Saturday, is that at least in the way the passing game performed, it won't get the job done in beating the Buckeyes unless it gets a lot better when The Game is played in late November.

 

 

Clarence Boddicker

September 15th, 2021 at 12:52 PM ^

Great work on this. The Forrest Gump gameplan was the result of a pretty rational assessment of the teams actual strengths and weaknesses, today. The line and RBs vs a talented but very young wr core and a game manager qb. Cade seems like a great kid, but I haven't seen evidence of great arm talent. He's good, but I'm thinking his ceiling is Rudock--which is fine! But this could well be what the offense looks like until McCarthy is ready.