"There are three stages in an actor's career: Who is John Amos? Get me John Amos. Get me a young John Amos." –John Amos | Photo: John T. Greilick, Detroit News 

He's listed at 6'4" and 205 lbs. In his career so far – about 50 snaps – he's thrown 24 passes and completed 14 (58.3%) for 186 yards and 1 TD. He's also rushed for 56 yards (96 forward, 40 backwards) and 2 scores. He kinda  looks like a young version of that guy who played Admiral Fitzwallace on the West Wing.*

For the last three games Hoke has been working Devin Gardner into the offense more often, either with the "Fritz" formation, the Denard Jet, or in some practiced Gardner-specific packages. How much of a 'rotation' this is can be overstated. Before Marcus Rush's flagrant roughing the passer knocked out Denard on Michigan's last drive I counted 67 snaps, of which 5 were Gardner's alone, 6 used both QBs, and 56 were just Denard. That's about a 90-10 split.

This is an attempt to discuss some prevailing theories as to why Devin is taking away a tenth of snaps from a Davey O'Brien award semifinalist.

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

* DYK: John Amos played ball for Colorado State back in the day.

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

THEORY 1: RUNNING QUARTERBACKS ARE FOR SOCIALISTS

The inevitable result of Denard's 2011 regression has been leagues of old blues who see Devin Gardner's hype and strapping physique and decide that anyone who looks more like a Scot Loeffler acolyte must be able to complete more than 30 percent of his passes in a trash tornado. They would be correct on the last bit; 3 for 7 is 43 percent. He also had a play so freshman-y against MSU it got through the entire first stanza of Yakety Sax:

The old men say things like "he's a better fit for Borges' offense," because the columnists they read break down all offenses into Manballicans and Spreadocrats, Borges being of the former (never mind that he's run a different offense at every stop). Their ranks are swelled by the same "Put in Henson" effect that has made 2nd string QBs and goalies so beloved wherever starters are struggling.

But there is evidence that Gardner is a more accurate passer than Robinson, not from the games but because his coaches say that. Hoke at the post-MSU presser:

Can you talk about philosophy of alternating Devin and Denard? “Yeah, we thought we may do some of that, and part of what pushed it over a little more was that it was a windy day, and I think Devin at times can throw the ball a little more accurately.”

So if we take the politics out and pare this theory down to "Gardner is more accurate," that accounts for two plays last week. One was when Denard overthrew Hemingway down the sideline in the 2nd quarter and was yanked for Gardner. Gardner immediately missed a wide open Hopkins for a 50-yard score. Later on they brought in Devin for an important 3rd and 11 early in the 4th quarter; Gardner fumbled the snap.

THEORY 2: THE MOST TALENTED GUYS ON THE FIELD

By recruiting metrics, Gardner is the most talented offensive player on Michigan's team. He was the No. 1 ranked dual-threat QB in 2010 to everybody but ESPN (Bolden), a 5-star to Scout, and the 5th ranked QB to pretty much everyone. Everyone who has a believable opinion about scouting says Devin has every type of it. After a year and change in the program, is there any way we can get some of that it on the field without giving up the rainbows and love and liquid happy of Denards?

Borges after debuting the Fritz:

We talked about it a little bit and after spring football it became apparent Devin was a skilled kid, we just tried to figure out a way we could use him. It’s hard, without literally giving quarterbacks series, and I know they did that here in the past, I’m just not a fan of breaking the starting quarterback’s rhythm by taking him out for another guy.

Game theorists and bloggers love this explanation. Technically they're replacing a receiver so the talent tradeoff is Roundtree for Devin. But if you're already overreacting to Denard and then you have to deal with Devin's legs on the other side, and Devin's arm, and oh yeah there's a running back who can go up the middle or option or (Gotchya!) pass it, you can see how opposing DC's can end up with disorders.

Last week this was very effective. The Denard Jet play netted two first downs that were both a shoestring tackle away from breaking big, and two well-defended fake jets that Gardner dumped off for 4 yards on 1st down, and a ran for 3 yards that should have been 8 if Devin hadn't missed his hole.

THEORY 3: THE LAW OF DIMINISHING DENARDS

Molk(+.5) and Schofield block down, Robinson(+1) finds a small hole and squirts into the secondary for 7 yards, where two linebackers and a safety converge. Denard gets up holding his hand.

The thinking goes, the more you play Denard, the more banged up he gets and thus the worse his performance. Talk during the offseason was all about limiting Robinson's carries. One way to reduce wear is to have him pass more, but his passing this year—scheme is part of it, inaccuracy is most of it—makes that untenable as his 2009-y performance against MSU shows:

This is rushing and passing together, with baby seals and EMUs excised. I was looking for some evidence that Denard's effectiveness goes down the deeper into a year you get but it doesn't show that. What it does show is Denard is less effective against better defenses (duh moment) and that he was very much Freshman Denard against MSU. It was also his passing-iest game yet.

Another way to keep miles off the Robinsonmobile is to drive the Gardner more. The tradeoff is that the best part of this offense is Denard's legs and that defenses have to overreact to that whenever Denard's legs are in the game.

Thus the Fritz and Denard Jet packages, which so far have gotten Denard hit about a third of the time but have Denard's legs as a threat 100% of the time.

THEORY 4: THE TATE FORCIER EFFECT

 
Ariel Bond|Daily / Lon Horwedel|AnnArbor.com

One of the reasons advanced stats loved Michigan's offense last year is we put up way above average yards against Illinois and Iowa, two statistically good defenses. Those also happen to be Tate Forcier's two extended appearances not in garbage time.

Tate had 597 yards on 84 passing attempts last year, all but 13 of those passes against real Big Ten defenses when filling in for a dinged Denard. With him gone and most of the Big Ten season ahead, this job falls to Devin Gardner. It might be a good idea to come up with a few plays he can do well and get those snaps logged. We've got that. In the Monday presser Hoke elaborated a little bit on the just-Devin package:

When you put Devin in the game, is there a purpose to that? “There’s three plays that we like Devin to run. One was the touchdown that he had against Northwestern -- that boot. And there’s another boot in there that we really wanted him to be a part of besides the combination of them both in there.”

More of this after NW from Borges:

Does it help to have Devin play a couple snaps when Denard was out from a coaching standpoint and from his standpoint? “Oh yeah, absolutely. Because Devin’s a highly skilled kid. He is. When he can go in the game and score a touchdown, really not make any tactical or technical errors, he gains confidence, and we gain confidence in his ability to do it.

Experience is needed. We had a preview of Devin in relief at the end of the MSU game and Gardner twice tried to throw the ball past the L.O.S. (the one above and the TD pass called back). That is an experience problem, and highlights the main reason, in my opinion, for getting Devin snaps: having a backup ready if/when Denard gets his dings. The experience thing is a double edged sword since it seems Robinson needs the experience snaps in this offense just as much as his backup.

In a perfect world Michigan would have an extra 15 snaps per game per quarterback we can throw away to learning. As it is that opportunity has come once this year, against Minnesota. The rest of the way, I figure the coaches will be looking for opportunities to give Devin a few more looks here and there. If he's replacing Denard more than 1 play in 10, it's because the dings have already come.