Previously: Podcast 15.0A, 15.0B, 15.0C. The Story. Quarterback. Running Back. Wide Receiver. Tight End. Interior OL. Defensive Interior. Edge. Linebacker. Cornerback. Safety. Special Teams.

Last year this question was "does losing Gattis matter?" The answer was "absolutely not":

I loved the gameplans down the stretch, but if Michigan put up two of the worst ten scores in UFR history early in the year, and things got a lot better late, and then Gattis decamped for 7-5-ass Miami after making the Cofopoff, what is the most likely explanation? Gattis got sidelined.

This was accurate. Gattis trashed Miami's offense, got fired after one year, and is now back with Locksley at Maryland. Insert Matt Weiss. Michigan's offense dunked on OSU again, moved up four slots in SP+, and looked exactly like it had when Gattis was around: meat, meat, meat, and meat.

Then Weiss took being the Madden Kid thing too far and got booted for still-mysterious computer access crimes. Kirk Campbell was elevated to QB coach and Sherrone Moore became the full-time offensive coordinator. In one area this has already been a massive win, as Campbell appears to be one NIL package away from landing #1 overall 2025 prospect Bryce Underwood. Ok, but what changes on the field?

[After THE JUMP: McCarthy items]

The answer here is also "approximately nothing." Look at this team. It is not just a Harbaugh team, it is the Most Harbaugh Team. I do have some optimism because Weiss was reportedly the playcaller on third downs and in the redzone, and for the most part I did not like what Michigan was doing in the redzone last year. Things came to a head in Michigan's slow-motion beatdown of MSU, when Michigan ran zone reads in an area of the field where zero coverage is a given and got thumped…

…then called a third-and-goal play from the eight that saw one receiver actually run a route into the endzone:

I suggested that Michigan just run Corum a bunch:

So what about this is actually redzone instead of just regular failure to execute offense?

Well, one: running your quarterback laterally inside the the ten is not a great idea. Both of the McCarthy keepers are solid or better gains anywhere else on the field because the MSU safety will be 13 yards downfield, not five. Two: Michigan is very opposed to throwing in goal to go situations. This is generally fine.

So…?

I don't think this has to be any more complicated than running buckets and throwing in the occasional RPO relief. TDs ensued when Michigan kept it in Corum's hands. This doesn't necessarily mean running the same thing every time; Michigan's first and goal saw them pull a tackle and then insert him on the other side of the line.

…and then they ran the same thing five times in a row against Rutgers:

Speaking of systemic problems, what is the dang deal with runs inside the two?

Well, for one, Rutgers was very well prepared and Michigan decided to Make A Statement. The last five plays of Michigan's first drive were all dive plays, and Rutgers just got lower and cut Michigan's OL to the ground on most of them. When that's happened this year Corum has leapt over the pile to score, but Rutgers had a linebacker tasked with defeating this strategy:

Rutgers DB #22 mirroring Corum presnap

…Michigan managed to get in on fourth and goal on both of these drives. So it was eventually the correct strategy. Getting there was agonizing.

Things did get better from there. The issues against Illinois were more about failing to get to the redzone than anything that happened once there, and the one failure to punch it in was an execution error on a great playcall (as McDonough says, Loveland is wide open):

Against Ohio State the slow-developing zone read was replaced by a quick QB stretch that punched it in from the three. But the thing I wonder about is the thing that I wondered about with Gattis: if something was pretty bad at first and then got better, is that Harbaugh putting his hand on the scale?

One thing Weiss was not was an actual QB coach; our most recent WTKA roundtable has a discussion in the commercial break about what his specialty was, and that was running the ball. So that leads into our next question…

2. Is Michigan really going to balance their offense?

[Fuller]

That's been all the talk. With Harbaugh it's very much a "believe it when I see it" situation, but if not now, when? Michigan's offense was considerably less efficient against TCU than 45 points implies because they had a whopping 18 possessions to score those points in. Much of the first half of that game was running into stacked fronts over and over again with little to show for it. Meanwhile the bulk of the first half against Ohio State was an agonizing exercise in doing nothing because Jim Knowles was shooting safeties down the throat of the offense.

Now, neither of those teams had to deal with Blake Corum but I do think those games exposed some limitations in Michigan's approach. I would prefer to come out hot, and for Michigan to not bash its head against +1 in the box repeatedly. Webb has some intel from within the program that indicates Michigan at least wants to be more diverse this fall:

“(McCarthy) has the freedom to change protections now, change plays, change routes… the full menu.”

Michigan is in essence showing a willingness to be more quarterback dependent. At least, it has so far in practice. And McCarthy has been razor-sharp in executing his expanded duties.

Last year the call was the call for the most part, and if the defense ended up in an advantageous position those were the breaks. In practice this didn't matter against most of the schedule because teams played Michigan inexplicably soft for much of the year. When they ran up against teams that were aggressive things tended to bog down unless Corum was able to conjure some magic. Take all-CB-blitzes-all-the-time Indiana. When not doing this on the first drive…

…Michigan running backs combined for 89 yards on 32 carries, 2.8 YPC. Similar head-bashings occurred in the first half against Ohio State and TCU.

There are two philosophies here. One is that Corum's going to do that so just give him the ball even if he has to dodge a guy—four guys—at the line of scrimmage. You land a bunch of body blows. Then you get guys crossing the line of scrimmage on play action and you get to lead the country in PA YPA by a huge margin, as Michigan did.

The other philosophy is that the only thing better than a linebacker who's wrong about run or pass is a linebacker who's not moving at all. See Mr. Penn State Backside Linebacker:

Last year Michigan was emphatically the former approach. Seth has repeatedly banged the table about this: Michigan averaged twenty five yards per play action pass attempt against TCU and ran play action six times. They started the season by running the ball literally 100% of the time when they were lined up in the pistol. When people yammering about it on the internet became annoying enough they threw out of pistol on the first play against PSU… and they didn't even bother to run play action.

This should change. Michigan cannot be bottom quartile in play action passes attempted again, not with this running back room and this offensive line. Games are shorter now so galumphing into halftime tied despite a huge lead in total yards will be more dangerous. Michigan needs to let JJ cook from the drop, to establish him as a real threat by the time the opponents get dangerous.

3. Aren't you excessively optimistic about the WR corps? 

At some point last week I was going over the draft of the WRs post and realized that poring over PFF drill-down stats had caused me to repeatedly slander Ronnie Bell. Bell was a captain, an NFL draft pick, and recently made the 49ers 53-man roster with room to spare, and I am out here implying that Michigan will be more efficient without him.

Well, time to double down. Here's targets and catch rate for deep balls and everything under 20 yards between Bell and the three guys most likely to suck up Bell targets (other than Tyler Morris, for whom data is close to nonexistent):

Player	Deep T	Deep %	Short-int T	Short-int %	Contested T	Contested %
Bell	18	44%	72	65%	18	17%
Johnson	7	57%	36	72%	7	43%
Wilson	11	18%	19	89%	4	75%
Edwards	5	60%	15	87%	1	100%
Total non-Bell	23	39%	70	80%	12	58%

Some of the gap here is because McCarthy had a tendency to lock on to Bell, and later Colston Loveland. The case I want to make here is two-fold.

Fold the first: Ronnie Bell is a bad guy to lock onto. This does not mean he is a bad player. He's a consistent route runner, has good hands, etc. He does not have electric speed or jitter that makes him all-the-time open and he does not have the size and strength to be good in contested catch situations.

Fold the second: ideally your quarterback will not be locking on to anyone. One reason for optimism about the WR corps is that McCarthy should make them look better by not pre-determining a throw before the snap, or defaulting to a safety blanket. Targeted WRs will be more open than they were a year ago because McCarthy will see the field better. And, yes, I expect the guys getting contested opportunities to be better at bringing them in.

4. Are they going to run McCarthy?

[Barron]

Michigan handled JJ McCarthy with kid gloves much of the year, which is a tradeoff. McCarthy did not get hit on 4-5 called first half runs. He also had to play deep into games when a more aggressive offensive approach could have led to fourth quarters on the bench sipping a mai tai. Michigan still ran the same plays, just without a live read, leading to slogs against teams (Rutgers) that knew McCarthy would not be carrying the ball. Then when Michigan got to the two most important games of the year McCarthy did not have a designed run in the first half against OSU and TCU.

Anyone who's read this blog for longer than thirty seconds knows I'm going to say this is suboptimal and Michigan needs to incorporate McCarthy's legs at full flight, or close enough, if they're going to scrape their ceiling as an offense. This does not mean 20 carries a game, or even ten. Let's aim for five. Just enough to get linebackers and cornerbacks checking him, a la Penn State.

That increases the efficiency of your offense, lets the whole starting lineup take a couple drives off at the end of games, and incorporates the rhythm of QB run calls into your base offense. Downsides are minimal. While I'm ready to believe that going Full Denard can mean your quarterback gets injured more often, that's not what we're talking about here. We're talking about an offense that looks like an NFL offense with a quarterback who can move. Repeated NFL studies cannot find a correlation between QB run rate and injury susceptibility.

Let's move the QB run meter from "extremely cautious" to "sort of cautious."

5. Well?

Two things are likely to hold Michigan back, at least in terms of overall efficiency. One is that they will continue to pull in their horns against lesser teams and end up in rock fights when they don't believe the opposition is a threat. At this point you have to assume this is ingrained in Harbaugh and will not change.

The second is that they don't project to have an offensive tackle who can delete elite edge rushers without a thought.

Everything else, though? Good lord. You've got a quarterback projected to go in the first round, a shot at the two top backs in the NFL draft, two senior WRs who will get drafted, Colston Loveland, the best guard pairing in college football, and a deranged Stanford transfer at center. There is sufficient depth that any non-QB injury can be met with a relative shrug. Michigan isn't bulletproof, but dang if it doesn't feel like that.

This should be the best offense of the Harbaugh era by some distance.

BETTER

• Junior JJ >> Sophomore JJ
• Fully Actualized, In Possession Of Hand Edwards >> One-Handed Pupating Edwards
• Older Zinter and Hopefully Healthy Keegan > Younger Zinter and Banged-up Keegan
• Colston Loveland >> Injured All/Freshman Loveland
• Absurd OL Depth > Pretty Good OL Depth
• Fullback Bredeson > Not Fullback Bredeson
• #1 Roman Wilson > Not #1 Roman Wilson
• Cornelius Johnson With Targets > Cornelius Johnson Without Targets

PUSH

• Blake Corum == Blake Corum
• Hayes, Jones, Barnhart == Henderson, Jones, Barnhart, Hinton

WORSE

• Drake Nugent < Olu Oluwatimi
• AJ Barner < Luke Schoonmaker
• Tyler Morris < Ronnie Bell

LAST YEAR'S STUPID PREDICTIONS

I do not want to make a QB prediction.

One point.

Leave me alone.

One point.

Fine: McNamara keeps his nose in front because he's less prone to mistakes and he improves his ability to see and hit shots downfield; with the talent around him he does not need to be the all-singing, all-dancing beating heart of the offense.

Aw, hamburgers. No points. 

No wide receiver is the clear #1. Johnson, Bell, and Anthony all are in a dead heat for catches and yards, with Wilson, All, and Edwards in a secondary group some distance behind.

Bell was a clear #1 with Wilson, Schoonmaker and Johnson in a group behind. All and Edwards got hurt and didn't get the targets as a result. In my defense, no receiver should have been a clear #1, IMO.

Two Michigan skill players are picked on Days 1-2 of the NFL draft.

This would have happened if Corum had entered, but the other guy on day two was not the one I expected going into the season: it was Luke Schoonmaker. Still, I am receiving one point for this and no one can stop me.

Zak Zinter blows up into a PFF fave-rave and is on NFL lips going into 2023 if Michigan's able to hang onto him.

One point.

Trente Jones at least replicates Andrew Stueber's 2021 grading; he's better in pass pro and worse as a run blocker.

More or less on point aside from some mental errors, but then Jones got injured and Barnhart replaced him. Incomplete.

The offense balances out almost 50/50 between runs and passes.

Lol no.

Michigan's season RPS grade is better in 2022 than it was in 2021.

One point based on the Penn State game alone. For the record this was positive in every game until TCU just went bonkers with the aggression, and Michigan was +24 against Manny Diaz.

Michigan finishes 4th in SP+ offense.

15th. I forgot that Michigan will gear down against Rutgers and the like.

THIS YEAR'S STUPID PREDICTIONS

• Corum and Edwards both crest 1,000 yards from scrimmage; half of Edwards's total is receiving.
• Starting tackles by midseason are LaDarius Henderson and Karsen Barnhart. Barnhart takes a leap forward and gets drafted.
• Roman Wilson is Michigan's leading receiver and catches half of his 20+ yard targets.
• Loveland mosses defensive backs on three redzone fades.
• McCarthy locks in the deep ball and lives up to first round projections.
• Ben Hall gets enough run to make everyone believe he is RB1 in 2024.
• Drake Nugent is very little dropoff, if any, from Olu Oluwatimi.
• Harbaugh plays 11 different OL over the course of the season; Sherrone Moore gives in and runs the 10 OL play with Alex Orji at QB.
• Michigan finishes 4th in SP+ offense.