The tumultuous recent years of Michigan football have spawned some truly boggling "this hasn't happened since" statistics. Most prominent was the OSU streak, of course, but this year Michigan enters a year running the same defense it did a year ago since 2007. Next year they'll have guys with a third year in a single system for the first time since 2003—remember Jim Herrmann's one-year experiment with the 3-4 in 2004. I'm just wow, man.
Even more remarkable is that if Denard Robinson remains healthy they'll get a season's worth of starting from a senior quarterback* for the first time since John Navarre 2003 almost a decade ago. If I hadn't used the Grosse Point Blank "TEN YEARS!" joke for basketball's 2009 tourney bid, I'd deploy it now. I'll still use this:
Yeah, Piven, I feel you.
If you don't remember, senior quarterbacks are good to have. They're generally efficient, even when they aren't escapees from a top-secret government experiment attempting to breed a new race of Sonic The Hedgehog soldiers. They change plays at the line and don't throw interceptions and sometimes pilot the kind of offense that can sing your baby to sleep with its metronomic precision.
Is that happening? Uh… probably not. But the Sonic thing gives you a lot of room for error.
*[Chad Henne's '07 season does not qualify, as he missed games against ND, Penn State, Minnesota, and Wisconsin (basically—he threw five passes) and was wounded most of the rest of the year, most painfully when he was throwing up Sheridan-esque moonballs en route to 68 yards passing against OSU. That space station was not fully operational until the bittersweet Citrus win over Florida.]
|QB drawin'||Hopkins floater|
|juke in out in out in .||holds up Hemingway|
|drop snap deadly||ZIPPY ARM|
|bursts outside||no pressure == good|
|patience FTW||can zip it|
|damn Lavonte David||no pressure again|
|VEER ME BABY||no pressure rollout|
|gashing Purdue||UNLEASH THE DRAGON|
|woop woop||bombs it deep to Hemingway|
|zooooooom||chuckin' it up|
|shortie vs OSU||fires deep anyway|
|seeya Shazier||deep corner|
|SCRAMBLE MORE||back foot bomb|
|you never do this||bumpy bumps|
|hate you David||Odoms back of EZ|
|WTF READS||UNLEASH THE D'OHGON|
|looks open on the corner||this is more Smith's fault|
|again misses a keep read||PLAIN IMPRESSIVE|
|pitch the ball!||sets feet on roll|
|refuses to pull||nails Dileo|
The comedown was predictable, and a little sad. Denard Robinson's electrifying 2010 season saw him garner Heisman votes and All-American nods like they were rushing yards. He shoved a rainbow down Notre Dame's throat and made them pop out of your head. He shattered records, big records, NCAA-wide records for rushing quarterbacks, and if he hadn't been saddled with the Worst Defense Ever on the other side of the ball, he might have continued doing so ad infinitum.
But he was, so he didn't. Enter Al Borges, a guy who has about as much experience with running quarterbacks as Rich Rodriguez did with intricate West Coast passing games. Enter Brady Hoke, who declared that POWER would be run powerfully. Exit the Denard-iso-based offense that disguised for Robinson's many shortcomings as a passer by getting guys flabbergastingly wide open.
Down went many of the stats. Denard's rushing yards dropped from 1702 to 1176. His yards per carry went from 6.6 to 5.3. Passing yards dropped by about 400, yards per attempt went from 8.8 to 8.4, and Denard's already high interception rate ballooned from 3.7% to 5.8%.
There wasn't much compensation in terms of keeping Denard hale. His carries scarcely dipped (256 in 2010, 221 in 2011) and he got knocked out of games against Michigan State and Illinois. It just did not work as well.
That said, the offense didn't fall off too much. Buoyed by a tough schedule, Michigan's offense didn't slide much in advanced metrics (FEI went from 2nd to 9th) and had barely budged after the regular season. They put up more points against Ohio State than any Michigan team since Fritz friggin' Crisler. They return eight or nine starters depending on how you want to configure the offense and how you feel about tagging Ricky Barnum a starter. If they can refine things…
Denard just wasn't very good at reading defenses (or wasn't allowed to be) in any phase of the game. There's no reason he would be good at the passing stuff given the Rodriguez offense. He'd stare down guys, like when Kevin Koger ended up wide open in the flat against Purdue. If he got a little pressure he'd chuck balls off his back foot, like he did three times against Northwestern.
It didn't seem like the coaches had a whole lot of faith in Denard's decision-making on the ground either. While they showed various option looks, these were basically run plays on which opponents had to respect the RB. Denard pitched once, and that was a fumble. In the Nebraska game I became increasingly more perplexed at Denard's refusal to pull the ball, eventually giving him an epic negative score for not doing zone reads at all properly. In retrospect it seems clear that those weren't reads at all if they were being so consistently missed—they were called keeps or handoffs and if the defense did something unsound, oh well. Here's a paradigmatic screenshot:
That speed option was a keep for a loss of three, and Denard's not even looking at a potential pitch.
After arghing arghing I came to this conclusion…
The one time he did pull the backside tackle blocked the end inside and nobody scraped, which makes me wonder if I am putting all of this on his shoulders when Michigan has abandoned the zone read in favor of making it look like the zone read but not actually giving Denard the option.
…and after several months remove that still seems like the most likely explanation.
Hypothesis: you choose to rep one thing hard to be an expert. Previously, it was zone stuff that forced defenses to be wrong with certain players and get players wide open. Under Borges, it was West Coast passing. The read skills atrophied to the point where they were not reliable enough to use regularly, Denard was always coming from a thousand miles behind in the air, and the results were a step back all around.
Borges hasn't turned into someone else, so the way forward is obvious… but might not be achievable.
Denard's season trajectory told a story of real progress culminating in that brilliant Ohio State performance:
So… we have a pattern now. In the beginning of the year Denard had no idea what to do with this passing offense and his lack of comfort screwed up his mechanics. As he progressed and Borges adapted to his strengths the comfort level rose and he hit a plateau of totally acceptable performances before lighting up OSU. The progress is undeniable. He'll regress a bit against VT but if he nudges his DSR above 70% it's time to quietly hope he can have a ridiculous career capping year in 2012.
The best part of going 14/17 for ten YPA? Three QB draws for 10, 10, and 16 yards. Run and tell that, homeboy. If Denard is the QB he became after the trash tornado game, look out: 59% completions, 7-4 TD-INT, 8.4 YPA against Purdue/Iowa/Illinois/Nebraska/OSU translates into… I don't even know what.
Here's his UFR chart for the year:
[Hover over column headers for explanation of abbreviation. Screens are in parens.]
[I went back and did a passing chart for the Sugar Bowl FWIW.]
After the dismal Michigan State trash-tornado-and-double-A gap game, something clicked. Or he stopped being bothered by an abscess on his elbow (or "boo boo" in Hokespeak). Either way, the uptick was dramatic.
This is hard to see in the traditional stats for two reasons: Gary Gray and weak opponents early. Denard's YPA actually dipped by a yard (Sugar Bowl included). The closer look UFR provides shows progress, and a lot of it. After racking up 17 BRs in the first seven games, Denard had just one per game in the last five before the bowl. His interception rate plummeted from an insane 7.1% to a still-very-bad 4.2%. His downfield success rate leapt up into the same approximate range he spent his sophomore year in, and he even scrambled a little against Nebraska.
The Main Thing
This is not going to be news, but my God, the interceptions. Last year when Football Study Hall took the top 100 I-A QBs by passing yardage per game and ordered them by interception rate, Denard was 84th. His 2011 number (5.8%) would have been 99th(!) on the list. Worse than Jacory Harris, BJ Daniels, Stephen Garcia, and everyone else except Boo Jackson of OHIO.
Virginia Tech is of course the great raspberry in the narrative of progress here. Since I am the worst (seriously: I apologize profusely for not doing that UFR, I really am the worst) I'm more hand-wavy than usual about what went on but I did go back and chart all of Denard's throws. I found an MSU-like game in which he responded very poorly to pressure, and since VT pressures a lot and got Michigan behind the sticks all the time, he was often put in positions to fail.
I noticed a similar trend when I went back through my Denard clips from last year. Most of the throws filed under "zippy arm" are ones in which Denard can set up in the pocket and chuck it without having to re-set his feet. Por ejemplo:
It's when he has to move around and re-set that we get most of the erratic throws, and it didn't take much for him to revert to bad habits last year, as the first half of that Northwestern game showed. He backfooted a bunch of throws that he didn't have to:
Stepping into those gets them off accurately without getting you sacked.
It's no coincidence that Denard's by-far-worst outings of the year were against the two teams that got in his face over and over. All quarterbacks see their performance decline when they get pressure; few have as an abrupt a cliff as Denard did last year.
Rodriguez's solution to this problem—if he ever had to consider it—was to make the offense so heavily run-based that passes were rarely met with heavy pressure. Straight dropbacks were rare, and defenses were hesitant to blitz in case they got a constraint play in their face or blitzed up the wrong gap. Borges probably won't and probably can't assemble an offense where the parts move just so, and anyway Denard threw a bunch of interceptions as a sophomore.
There's only one thing that can fix this, and that's Denard not deviating from his mechanics as much and knowing where to go more. At the Glazier Clinic, Al Borges talked about the "backside cuts" that are built into Michigan's passing game. Those are deep routes that are supposed to be aborted to whenever the guy away from the main thrust of the play gets one on one coverage. Borges said Denard was "very aware" of these backside cuts, "very aware," and visions of double-covered WTF bombs danced in my head.
He should be more aware of the other guys running those routes (e.g., safeties) this time around, and have better timing on some of the underneath stuff that was an issue. He will get pressure, and I don't think his issues are the kind of thing you can fix in an offseason. There was a groaningly inaccurate pass at the open scrimmage in which Denard had to move around and he made a leaping heave across the middle that was yards behind a crossing route. That's just never going to go well.
Improvement should be expected, though. Denard was still super raw a year ago, he is entering year two, and you know he worked at it all offseason. How much will be the tale of the season.
A Couple Other Things
1. FOR THE EVER-LOVING SAKE OF SNEEZY JESUS WOULD YOU JUST TAKE OFF WITH THE BALL WHEN NO ONE IS OPEN?!?!?!?!?!?!?!?
Sorry. Sorry. I just don't get it, man. The ratio of ridiculous athleticism to passing skillz dictates that whenever Denard gets past read two he should be thinking about an exit strategy. But Denard all but refuses to run. He scrambled more than once in exactly one game last year, Nebraska.
It's almost as if he wants to prove he's a quarterback so badly that he refuses to use his legs when not required to. This combines with the feet-setting thing in a toxic cocktail. I'd prefer it if Denard either set up and threw without moving his feet or ran, with nothing in-between.
Survey says: unlikely.
2. Shorten the passing game.
The trend in the ND game—bombz—was one that lasted throughout the season. That trend:
Hoping for the fluke explanation, but there seems to be some merit to Door B. He's a breakdown of passes in last year's Notre Dame game:
- Hitch: 9
- Flat, seam, bubble: 6 (one waggle FB flat!)
- Deep curl, flare: 3
- Tunnel screen: 2
- Post, corner, fly: 1
- Throwaway: 1
- Run around like Tate: 2
This is a dedicated short passing game that ran a ton of curl/flat. Denard completes 60% for 1 TD and no INTs, averaging 6.1 YPA. This year we've got the eight downfield chucks, two throws behind the line (8% of attempts) instead of 11 (over 28%), and a total lack of free touchdowns in the seam or hitches to stationary targets that worked well last year when Roundtree wasn't dropping them.
Some sort of perimeter stretch would help Michigan a great deal. Those flats and quick hitches and bubbles are not only highly effective quasi-run plays but drag linebackers away from Denard in the middle of the field. Borges is still thinking like a guy who has a quarterback who happens to be the fastest kid in the country instead of a guy who has the fastest kid in the country at quarterback.
3. I-form snaps are inherently dumb with Denard.
Short yardage, whatever, fine, but any other I-form snap is burning money.
What to expect
I don't actually know, man. He's not going to do that Cade McNown thing. He should get a lot better. Extrapolate the back-half of his season out across a full year, add in a year of experience, and pray for health, and you've got a Heisman contender.
This is the worst thing I've ever said, but I don't think he'll get there. Thus the 4.5 above instead of last year's 5-plus-exclamation-point. There will be a game or two against elite defenses in which Denard's limitations are made plain, and that will keep him out of the running. His interception rate can end up halved and still be really bad. There's just too far to go in one offseason.
The projection in numbers: YPA remains static. Completion percentage jumps up a few points, scrambles are a little more common, and interceptions drop to around eight. Rushing is about the same as last year.
Devin Gardner spent last year making cameo appearances in the two-QB package Al Borges calls "deuce" but should by all rights be called "Fritz" and running the base offense when Denard was inevitably banged up. Neither of these things went that well except on that one pass in the Illinois game. Gardner rushed for 3.5 YPC even if you exclude the Michigan State game and its yakety sack, completed fewer than half his passes, missed a blitheringly wide open Hopkins against MSU and threw a fugly interception against Purdue.
Then spring practice hit and rumors leaked out that Gardner was not only playing wide receiver but playing it ridiculously well. Gardner has spent every waking moment since deflecting questions about his position; 42% of all sports content on the internet since has speculated about a potential position switch, its costs, and its benefits.
This site's been on Team WR from the beginning and became even moreso after an alarmingly poor performance in the spring game that caused me to survey the Gardner oeuvre with a suspicious eye:
In three consecutive spring games he's looked bad. You may remember Jake Ryan bursting onto the scene last year with a pick six thrown directly at his dome by Gardner. Yeah. … [The year before that] Gardner got safetied and intercepted on the same play and still probably had a better overall outing than he did yesterday.
When fall practice started up Hoke offered up the only piece of solid information he's provided in months by admitting that yes, Gardner was practicing at WR. His potential impact there will be covered in that position preview.
Gardner will have an opportunity to play at both spots. He's getting the same load of QB reps and moonlights at WR when other quarterbacks are taking snaps, and Borges pointedly defended Fritz from a reporter's question despite the thing seeming to run out of gas after the Denard end-around package was adequately scouted. He may not be the first guy off the bench if Denard needs to come off for a play, but any long-term issue will likely see Gardner ascend to the starting spot, where his performance is anyone's guess. He needs to get a lot better to be plausible; raw athletes going into their second year in the same system do that sometimes, but maybe not often when they're spending at least half their time at another position.
Redshirt freshman Russell Bellomy [recruiting profile] is the third(?) stringer and only other QB on the roster. He was clearly more effective than Gardner in the spring game, but had the luxury of going against backups and was a checkdown-heavy dink-and-dunker. He went six of nine, sure, but he averaged six yards a completion. He displays some athleticism, though not anything in the same stratosphere as either of the veterans. Tate Forcier was his YMRMFSPA; former Purdue quarterback Brandon Kirsch is also a decent comparable.
The coaches have been talking him up some. When Tom Dienhart hit up a practice he returned back with this news about Bellomy:
…the guy to watch is Russell Bellomy. Brady Hoke told me he is faster than you think. I also asked him if he’d be comfortable if Bellomy had to play, and Hoke said he would.
It's hard to tell whether Bellomy's development allows Gardner to play wide receiver or the crying need at wide receiver forces people to play up Bellomy's progress. Either way the downgrade from Denard to Bellomy would be severe. A few plays here and there for dings will be fine.