1. Does Michigan have a decided schematic advantage and should I sue you for making me ask that question?
Last year's edition of this post had a question about how Harbaugh's ultra-manball ways fit in an increasingly spread world. Despite my long history of spread zealotry I was pretty sanguine. Harbaugh had a fantastic track record and when I went over some of his Stanford-era games and a couple things stood out. One was yes, this:
The other was that in certain ways the Harbaughffense and spread stuff ended up at the same place despite taking diametrically opposite paths to get there:
Harbaugh's offenses put mental pressure on the opposition in a way that previous manball offenses at Michigan did not. This came up constantly during the Al Borges's tenure; I said that having to dodge a safety near the line of scrimmage sucked while Borges's defenders said they'd take it all day and twice on Saturday. It's clear that Harbaugh is in the former category. Like spread offenses, Harbaugh loves to screw with opposition safeties.
Rich Rodriguez did that by playing 11-on-11 on the ground with Denard Robinson. Harbaugh does it by whiplashing the opposition between jumbo sets and four-wide, by flooding the field with big guys safeties have to get around, by constantly screwing with their keys, and by adding new stuff on the regular.
Last year's UFRs were a ton of fun to do because every week Michigan would come out with a new package of plays I hadn't seen before. The sheer diversity of Michigan's ground game fairly boggled the mind, and I say this as a person who has broken down six or seven seasons of pro style offense. Hell, Harbaugh changed offenses across the league. Michigan started facing down trap blocking at a far higher rate than they ever had before.
UFRs pretty quickly picked up a section about the "Stanfordization" of the offense that detailed the tweaks and new packages I picked up on weekly. After Maryland:
The most obvious new wrinkle was the T set, which Michigan used a couple different ways. A counter iso play was successful when Isaac was not fumbling:
The two ILBs went to entirely the wrong hole, buying Michigan a free blocker, and if the Kalis block had gone a little better Isaac is one on one with the last guy for six points. Harbaugh's ability to buy back the extra defender you have to deal with when you aren't running the spread is a consistent theme so far this year; this is yet another example.
Northwestern had been super successful with aggressive linebackers when they rolled into Ann Arbor, and Michigan had a number of plays that made them pay for it:
I did appreciate the Kerridge fullback dive. Here is the play just before it. Watch the linebackers.
Here is the dive. Linebackers again.
That play exploited the blitz-ball mentality to spring a big gain. It also gave us a brilliant still shot demonstrating how weird this offense is to players born and raised on the spread:
"Wait… he's got the ball? Can they even DO that?"
And I loved Harbaugh's ability to see what his team lacked and game plan around it. This was especially validating against Penn State. A few years ago the infamous 27 for 27 game featured snap after snap on which Michigan ignored the fact that Penn State was playing ten yards off Jeremy Gallon; in Harbaugh's first year he felt his OL was a bad matchup:
So this game was… okay.
It was. The default thing that happened seemed to be this:
And, like, I'll take it.
There weren't actually as many of those as it felt like there were. Michigan threw 8. Those picked up 48 yards, a solid 6 per attempt. One was called back on a ridiculous block below the waist call. One could easily have drawn a block in the back call on Perry.
Those eight screens had a minimum gain of three, that on second and four, and picked up four first downs. They also opened up a couple of actual runs when PSU had to get serious about putting their linebackers over slot receivers. They were successful and easy. PSU's defense wanted to give those yards up, and Michigan took them.
I love that Harbaugh is clear eyed enough to work around the limitations of his team—also a major theme against OSU. He doesn't think "the expectation is for the position," he thinks "we're going to get overrun, let's do something about it."
Michigan's offense was a rock paper scissors winner under Harbaugh. I had the UFR RPS metric positive in 11 games with slight negatives against Northwestern and Indiana, and that hasn't happened in a while. They've been pounded over and over in that metric (and everywhere else) by MSU; that was a slight win. And this is just the first, most screwup-prone version of the Harbauffense.
There's a reason he built Stanford into rushing powerhouse with a bunch of three star guys. Not only is Harbaugh a smart and creative football mind, but he surrounds himself with other guys like that. How many offensive coordinators does Michigan have? Three, maybe four. Harbaugh is one. Drevno is one. Jedd Fisch is one. Nick Baumgardner had an excellent article last year describing the way this works in practice:
"It's unique (compared) to what I've done before," Fisch said Wednesday. "But it's something I would always do from now on."
Instead of designating one person to serve as the team's chief offensive play caller, or limit the discussion to himself and one other coach, Harbaugh keeps an open dialogue going with his entire offensive staff from snap-to-snap on the sidelines during game days.
That is terrific.
So yes, Michigan can expect to win coaching battles now. Not every last one, but most of them. Lloyd Carr didn't even try to do this—congratulations to Mike Debord on narrowly escaping his nemesis last night by scoring 13 regulation points—and Brady Hoke was incapable of it. (RichRod was pretty good at it but let's not open that can of worms again.) They have a decided schematic advantage.
[After THE JUMP: QB theme fight, Smith sustainability, OL panic]
2. Who wins between "Harbaugh QBs are always good" and "Borges recruits are always terrible"?
The main concern about the latter is a callback to Harbaugh's first two years at Stanford, when he had absolutely nothing to work with. There is a version of this quarterback battle that looks like nothing to work with: three-star Borges recruit with scanty outside interest versus one of the worst QBs in the country in 2014. Morris going bust and Michigan opting for nobody the year before him thin out the reasonable options to just the two, and you could squint at the ghosts of TJ Ostweiler and Tavita Prichard if you're so inclined.
However, the bet here is on the former. This is not a quarterback competition like last year where Jake Rudock was the guy by default despite the fact that he'd just arrived in fall. John O'Korn is legit competition who got beat out. Spring ball isn't much to go on, but it's all we have, and both those guys looked light years better than the Morris/Malzone competition in 2015.
Speight in particular has a command about him now, both on the field and when put in front of a microphone. I admit this is 100% feelingsball that anyone with a newspaper column could fart out, but when we have little data you have to make big: in both of Speight's live-fire appearances against Minnesota and Ohio State he looked terrrrible for about two drives. Then he settled down, got in a rhythm, and drove Michigan down the field.
At both Ford Field and the spring game Speight felt like the guy who'd settled down. When the media asked him about how he'd developed he said he was now that guy:
"Because it's me, I can remember back to the same plays we ran last year. I watch film and I can remember a rep or a play and I remember thinking last year 'this is so fast, the game is so fast.' It looked like a blur," Speight says. "This year in camp, and in the spring, I drop back and the game is so much slower. I can see guys come open before they're open now, that's the biggest thing.
"Everything's slowed down now."
The scant evidence we've gathered since the end of last season supports that. Speight should be fine.
3. Can Smith be Florida Smith?
The skill positions are mostly settled, both in who will man them and how good they will be. Chesson and Darboh might have a little something left to prove; I expect they end up an excellent pairing. Jake Butt is Jake Butt. Grant Perry is going to be Grant Perry.
That leaves Smith as the biggest wildcard. His upward trajectory at the end of last year hit an exponential bit of the curve in the bowl game. If he is that guy for good, Michigan's offense could suddenly be a juggernaut. This is not an exaggeration. Smith's output in that game was best-back-in-the-conference level, and that would have been worth half a yard or even a yard of YPC over the course of the year. But it was just one game.
I cannot answer this question without handwaving. We've got the bowl game and late uptick versus the rest of his career. We've got the lock starter treatment from spring, and reports from fall that Smith is still the #1 guy over a surging Ty Isaac.
While I don't think Florida Smith is going to be the baseline from now on, the bet here is that Smith makes a significant move towards damn good this year. People will reference him along with Chris Perry when talking about running backs who broke out in their last go-around.
4. Will the offensive line ruin everything?
I caught a few rather mournful reactions to the offensive line post here and elsewhere that I thought were odd. I said "eh, they're okay," not "run for the hills." I've said "run for the hills" before. This is not that.
It is a bit of a drag that it doesn't seem like any of the seniors are going to get a ton better and that they seem set to start a true freshman left tackle. But we've seen a reasonable worst case scenario for that: Mason Cole 2014. It was okay. Bredeson projects as better than that; he's higher rated and has the endorsement of neutral BTN folks. He's also going up against this defensive line and catching the eye. There are many okay Big Ten OL who would have looked like cannon fodder to the BTN crew. Bredeson will be less than ideal; he won't sink the season.
And we know what to expect from the rest of the crew. They'll get better. They got better over the course of the season. Okay left tackle and better all around should mean Michigan treads water at worst, and last year's line was… fine. This year's line will be fine or better.
While I've projected fine, there is case for better. This Kalis quote jumps out at me:
"The big thing for me is having coach Drev as my coach. I can't stress (it enough). Coach Drev has helped me so much," he says. "I credit it all to him. The way he's coaching me ... I wish I could have had him longer. I'll say that.
"I wish I could have had him longer."
These guys are who they are physically, with the exception of the left tackle. Familiarity with the offense and the Drevno effect are where major wins come from, and that looks a lot like Kyle Kalis having the light go on. There is a reasonable chance the year two leap I talked about in The Story is a big enough one for me to revise the "eh, they're okay" projection upwards.
Unless there's a major dropoff at quarterback the question is now if they will get better but by how much. Bill Connelly's S&P+ stats had Michigan 43rd on the ground and 8th in the air a year ago, which… uh, wow. That passing number is going to be hard to top. (That feels like a number on which the advanced stats get confused: a lot of Big Ten pass defenses look good because Big Ten QBs were terrible last year, so Rudock gets more credit than he probably should.)
Still: I'd expect every conventional stat to improve except sacks allowed, which was already very low a year ago. Harbaugh's second year at Stanford saw them go from 3(!) yards a carry to 4.9; Michigan was already at 4.2. Something around 5 YPC is feasible, especially since DLs at MSU, PSU, and OSU are in flux.
Passing YPA should at least remain steady but expect things to go a lot more smoothly early with senior versions of the receiving corps and a bunch more experience for whoever the starting QB is.
Overall, this offense will feel more like the Citrus Bowl than the rest of the year.
- Fully weaponized Jehu Chesson >> Stonum version of Chesson
- De'Veon Smith with glasses >> Fred Jackson version of Smith
- Returning Kalis, Braden, Magnuson > Those guys in an new offense
- Bunting, Wheatley, Asiasi > AJ Williams
- Grant Perry > freshman-ass Grant Perry
- Jabrill Peppers > less involved Peppers
- Year two of Harbauffense >> year one
- Functional Isaac, Chris Evans, Drake Johnson > Just Johnson
- Wilton Speight == Jake Rudock
- Amara Darboh == Amara Darboh
- Jake Butt == Jake Butt
- Mason Cole, center == Graham Glasgow
- New Left Tackle < Mason Cole, tackle
- Poggi and Hill < Kerridge and Houma
Last Year's Stupid Predictions
Rudock starts the whole year and turns in a season like last year at Iowa except more efficient: 60% completions, 8 YPA, excellent TD/INT.
Bizarre how we got there, but on point.
Glasgow and Kalis perform excellently, with Glasgow getting drafted in the middle rounds.
Glasgow yes, Kalis no. Half point.
Both Braden and Dawson play early in the season, with Braden eventually winning the battle.
Dawson didn't get a run but I think this is accurate in spirit. Braden started off suuuuper shaky and became solid, or at least mostly so, over the course of the season.
Jake Butt leads the team in receptions and wins the Mackey award.
Butt had 51 catches, seven behind Darboh and one ahead of Chesson. He didn't win the Mackey, but this was a prediction about a tight end coming off a 21-catch season. Also the Mackey people are dumb. This is close enough I am giving myself a point.
De'Veon Smith and Drake Johnson emerge into the two main backs, with Smith getting a plurality of carries because he's healthy to start.
Ty Isaac functions as a weird H-back hybrid guy and third down back and makes 30-40 catches.
Nope. Isaac fell out of the offense due to fumbles and many other issues.
The run game moves up to good-ish, as Michigan approaches five yards a carry but doesn't get there.
4.2 YPC and static line yards, so no. Michigan didn't get the Harbaugh effect in year one.
This isn't a point, but it's a point. I mean. For real.
This Year's Stupid Predictions
- Quarterback production is very close to Rudock's a year ago but edges up slightly thanks to this ridiculosu WR corps: 62% completions, 8.4 YPA, 2:1 TD:INT
- Smith has five yards a carry and loses just one fumble.
- Bredeson is the left tackle starter for the duration and outperforms Cole's freshman year by a good bit.
- Cole is ranked as one of the top centers available for the draft.
- Butt wins the Mackey, is first TE drafted.
- Chesson has just under 1000 yards receiving and Darboh stays about static; both guys get drafted, Chesson early and Darboh late.
- Peppers scores 8 touchdowns as an all-purpose back, including one 40+ yard catch.