The secret of steel has always carried with it a mystery. You must learn its riddle, Conan. You must learn its discipline. For no one - no one in this world can you trust. Not men, not women, not beasts. [Points to sword]. This you can trust.
-Conan’s Father, Conan the Barbarian
I loves me some prognosticationatin. Sure, Its imperfect and makes you look like a daggum moran when you’re wrong which is the majority of the time. Even when you’re right, folks will just ask you to do it again, and then again before they even consider listening to you next time. Heads: you’re wrong; Tails: you’re lucky. And its nowhere near a fair coin. Bloody infidels. You’re left with the choice of making peace with being considered either a fool or a crackpot. Its almost all downside. The prudent thing to do is just wait for things to play out and slap hands when things go well.
Maybe so, left brain, maybe so… But [screw] that. This is the internet for Pete’s sake. If you cant be wrong here, then what is it good for??? And everyone thinks we’re a crackpot already. Oh noes, another jerk I’ve never met thinks I’m an idiot. Woe is us. I pray my fragile psyche will some day recover from the indignation that is internet shame.
Aren’t we here to have fun? Isn’t it more fun to invest internet cache in a prophecy and see if you’re right. Its a simple social contract pick a side and to victors go the spoils of watching the losers eat crow. Even those who aren’t as in-your-face about it all…they know. They know they were right and the others were wrong. There is nothing sweeter than being able to say “I told you so” even when you choose not to say it. Its about that which is best in life: to crush your enemies, to see them driven before you, and to hear the lamentations of their bloggers. Indeed.
Last week my man Ron Utah had the onions to put out a diary that set the stage for Devin Gardner to the best QB in history next season. Damn near guaranteed it. Relax Ronnie; I keed, I keed. Devin put up a 160 passer rating against solid competition over a 5 game stretch. I think its safe to assume he could up good or better performances against Air Force, UMass, Purdue, and Illinois. Five games of data isn’t exactly nothing but it’s not a whole lot either.
This is a topic that is near and dear to my heart. Prior to the 2010 season I spent a lot of time studying Quarterback statistics and progression in an effort to get some insight into what would be reasonable and maybe not so reasonable to expect to see that coming season. Over the course of those writings I made a bunch of prognostications on various QBs relevant to Michigan that year. I’ll suppress the desire to assess each statement player by player in this space but I’m pretty proud of those pretty specific predictions. I was wrong about Ricky Stanzi, but I claim everyone else as a prediction of reasonable to uncanny accuracy or uncallable. I’m claiming a 10 right, 1 wrong, and 5 no calls. Word. /appeal for authority
White Rainbow Reprised
Anyway, I’ll get back to Devin in a minute but I need to lay some groundwork down first. Gots to hit ‘em with a…
Chart. Actually, four charts. These are the components of Passer Rating. The data is compiled from 951 QB seasons from 2003 to 2012. The point is to show the relationship between each factor and the overall rating. The interception rate chart is interesting because the slope hooks up when you start dropping below 100. That makes sense because sometimes when you suck, you really really suck and your foes all go hyuk ‘cause they hate you…or something like that. Just remember to switch lines down there.
The the high quality fits present here are no surprise: passer rating is a linear mathematical equation. But, these fits are useful because we can use them to evaluate how “legitimate” a given player’s performance was. For example, Denard Robinson’s 2010 Passer Rating was 149. Typical components for that rating are shown in the second line of the table below, his actual values are shown in the first line. Denard’s Completion percentage was inline with his final rating but everything else was either slightly out of whack (TD%) or WAAAY out of whack (YPA, INT%).
|Denard Robinson 2010||Cmp %||YPA||TD%||INT%|
|Single Factor Rating||146.09||160.00||141.72||105.37|
A reasonable cynic would say: so what; it counts. Indeed and we should make no apologies. But understanding these gaps helps us understand if there are any areas of improvement or if a guy is overly dependant on big plays and such. This was exactly the case with Denard in 2010. Remember all those “Oh, wide open” seam throws he made to Roundtree that year? Those big plays were generated by the offensive scheme and the threat he posed as a runner, not by the threat he posed as a passer. So, when the scheme went away, so did his performance. When an opponent was able to neuter his running threat, the offense sputtered. Capice?
What’s Your Rainbow Got To Do With Me?
Finally we get to Devin. Devin posted an overall passer rating of 161.7 in the 5 games he played QB in last year. His stats illustrate the “White Rainbow” Effect of the Passer Rating formula beautifully. None of the components were typical of that passer rating, the great cancels out the terrible and we end up with a number that doesn’t quite mean anything in and of itself.
|Devin Gardner 2012||Cmp %||YPA||TD%||INT%|
|Single Factor Rating||132.84||176.28||176.97||98.33|
Devin’s completion percentage was fine, but nothing special really. His INT rate was flat out terrible, better than Denard’s but that's not saying a whole lot. This isn't a surprise given the fact the Devin had never seen significant PT at QB and then spent spring, summer, and damn near two thirds of the season practicing at WR. Those numbers are inline with reasonable expectations given his circumstances. Actually, he probably out performed reasonable expectations given those circumstances and the inferable information that comes along with a position switch.
On the other hand the TD rate and YPA numbers are both ridiculous. In fact, they’re so ridiculous that given the context of the other two components, they reek of unrepeatability. AhOOOOga! Dive! Dive! Dive! [Everything starts shaking.] WHAT THE [HECK] IS HAPPENING!?!?
Relax, that’s just the mean reversion alarm. I’m reprising a chart I’ve previously discussed in this space which shows the mean reversion effect in a forecast able format. This chart suggests that we should look for Devin to register a legit 145 in passer rating next year. Uh, that’s a damn fine rating, folks. Here’s a stat line about your QB that you just might like:
|Devin Gardner, 2013||360||225||2900||23||10|
Sure, that’s not the 3600 yds and 33 TDs Ronnie promised but its not 15 INTs either. That level of performance is typical of a seasoned veteran QB and given Devin’s performance last season it’s totally reasonable to expect him to perform at this level. If he were to underperform that expectation the world would not end but I’d be disappointed. If he over performs then gravy. This is the reasonable and prudent outlook for Devin Gardner’s performance for this coming season.
Reasonable. Prudent. Pfft, those words suck. They aren’t sexy at all. They aren’t fun. They’re words Brady Hoke would use to describe Notre Dame. A seasoned veteran QB will do in a pinch but I want to plan on hearing the lamentations of all of those that oppose us. No, to lay those plans we need a monster. We need an expectation that we’ll have a performer at QB that will dominate. We need what Ron Utah has promised us. Surely there must be a way to construct a plausible expectation for domination given the data available. Indeed, but we need some abstract thinking. Shamalong to the right side of knowledge.
What if those numbers for YPA and TD rate weren’t an accident and are actually evidence of elite playmaking ability. Numbers like those are characteristic of system, support, or flat out skillz; usually a combination of the three.
System: We’ve seen what system can do for a player. We need look no further than Denard Robinson. In RichRod’s system Denard rated out at about 150, in a more typical system he rated out 20 points lower. In Devin’s case, he’ll be playing in a familiar system with more time to learn and refine his understanding. Borges will be able to game plan around a playing style he’s more experienced with and I see no threat to Devin’s performance being affected because of system factors next year.
Support: A quarterback can benefit from the players around him. I submit freshman Chad Henne. Having Braylon Edwards to throw to and Mike Hart around certainly helped his numbers. That O-line wasn’t too shabby either. So the question is: did Devin’s numbers get a boost from the players around him that will suddenly not be there anymore? The receiving corps returns mostly intact and as much as I enjoyed Roy Roundtree as a player, I don't think he was a transcendent talent that cant be replaced by Chesson and/or Darboh. Who is to say that they wont be better?
As to support from the running game, again I don't think there was anything there that made opposing defenses think twice. Michigan was aiight running the ball last season. Fitz’s production seemed to pick up when Devin stepped in for Denard but then he got injured so its tough to say if that's a trend or not. Here again, support from the running game should at the very least stay where it was and maybe it picks up a bit.
Finally, the offensive line. Michigan loses some solid players but retains its best player. Also, some high powered recruits from Hoke first few recruiting classes may finally be ready to contribute even if its just to add depth. It’s possible that the o-line could be touch weaker with the addition of new blood but I have no idea how to gauge that. I’m not going to worry about it.
In terms of support, Devin should have everything he needs and maybe even have an overall better receiving corps and stronger running game.
Skillz: Everyone around these parts know Devin Gardner was a high high level recruit. Once he got to campus and we saw him in Spring Games it was like, well, not that great and when he got moved to wide receiver last year I, for one, kind of convinced myself that maybe he was just another talented athlete with poorly developed QB skills. But, the numbers he put up when given extended playing time, while rusty, against not to be scoffed at defenses…jeepers. Maybe Scout was onto something when they gave him that 5th Star.
Its not crazy to think that Devin Gardner is that good. He might just be a monster. The only thing that prevents anyone from saying that confidently is sample size. So, as is my way, I try to look into the future by looking into the past. What does Devin Gardner the Monster look like?
|Devin Gardner, 2013||360||235||3200||27||8|
Yeah, baby. Lather up. Dennis Dixon. Troy Smith. Vince Young. What’s up with all those guys wearing #10 anyway? What’s up with all those guys murdering my team? Whatevs, Devin Gardner the monster looks like Michigan’s Revenge.
See, the only thing that makes this a difficult thing to believe is that posting a 160 passer rating is a difficult thing to do. But, its possible and what fundamental reason is there to say that Devin lacks the wherewithal to post that type of performance in the coming season? Depending on which set of assumptions hold, Devin will be either a really good QB or a devastating one.
If Al Borges says Devin plays like RGIII, I believe him.