Wednesday, August 29, 2007

What makes a Heisman? (Part 2)

First off, I'd like to point out that I did not consult with BC Offensive Coordinator Steve Logan before (a) making my comments yesterday and (b) devising this formula. You see, yesterday in a post on the NYTimes Blog, Logan had some comments which look like they were a part of my rant yesterday:

“That’s a looney-tune deal. You’ve got to have a great defense. You’ve got to have some help. The one thing about quarterback play, I said it on my radio show every week, the quarterback, unlike any other position in any sport, cannot and never does operate in a vacuum. If you want a Heisman Trophy-winning quarterback, he’d better have a pretty salty running back with him. He’d better have a wide receiver that can make a play. He’d better have an offensive line that gives him time."

He followed it with something that I loved nearly as much:
“The quarterback out in Hawaii, he’ll throw for 80 quadrillion yards because of the system. If he does and they win every game he’ll come in second to someone from Oklahoma, Texas or Southern Cal.”

Ah, agreeing unintentionally with an assumed eccentric genius. I love it.

I need to stop patting myself on the back here, for a second, because I guess that I should keep talking about more of the results of this little number. The thing that was most interesting to me is that trying to determine an offensive MVP to the best teams in the country was pretty much a self contrived formula.

What's the saying, there's lies, damn lies and statistics? I could have played with them anyway I want, and, for the most part, at the end it came down to being fairly consistent with what the voters thought. I mean, in 2004, Reggie Bush made it as a finalist but not too many were concerned with whether he would beat out Leinart. It was a battle of QBs for the trophy that year, and it was defending champ White versus the USC slinger.

USC actually played a statistically weaker schedule that year than Oklahoma, but where the stat line differs between Leinart and White was thanks to Adrian Peterson. The Southpaw's South California TD/INT ratio was 50 percent better than his Sooner counterpart. For one, Leinart was significantly more efficient, but also because when the endzone got close, instead of throwing into traffic, White handed off to Peterson. The OIF for both was extremely close and only slightly favored Leinart.

Leinart, QB, USC, OIF=2.436, Score=9.899
White, QB, OK, OIF=2.424, Score=8.622

This year was an example of how those overall production stats on two teams that are both very similar and make up will create the difference between the trophy winner and the runners-up.

2005 played out as many thought, with Bush clearly dominating. Whereas he was merely a role player the year before, he did more with fewer carries/receptions than anyone else. That assumption is what should be driving the "Most Valuable Offensive Player." Efficiency and impact on a winning team.

Pressed to keep going back to one of the year's of the most controversy, I was able to tap into the 2003 season archives to see if Larry Fitzgerald got screwed.

It was Jason White and a whole bunch of "really? they're a finalist?" type players that year. Fitzgerald earned his spot in NY, but White's objective numbers may have been the least impressive of any Heisman winner in recent memory. The same formula which had been pretty accurate over the last several years found a massive hiccup in a WR from Pittsburgh.

White's line in 2003: 40 tds, 10 picks, almost 4000 yards on nearly 500 attempts. I haven't discussed this before, but sacks count as an attempt in my formula, mainly because it is a designed passing play and how the QB reacts should factor into it.
Fitzgerald's line: 22 TDs on 92 catches. Even with Pitt's lackluster record and schedule, that massive efficiency computed the following.

White, QB, OK, OIF=2.387, Score=8.474
Fitzgerald, WR, Pitt, OIF=3.998, Score=8.928

Fitzgerald didn't get anywhere close to a BCS bowl, either, and the Panthers didn't win the conference. White got some significant bonuses from that even in my formula. It looks like the voters gave him even more. Why did he compute so high? Because he was really the only part of the offense. He factored in to almost half of Pitt's total touchdowns on the season. That's impressive to do, to begin with, but to do it on only 92 touches, that's even more ridiculous.

What does all of this Heisman number mumbo jumbo mean? For one, it shows what Logan was talking about (even though I was talking about it before he was). The player who the offense would be only good, not great, matters more than a leader or role player on a winning team. And it will be interesting to follow all year against the prognosticators.

Here's my current candidates who I'll be tracking all year:
Colt Brennan, Hawaii, QB
Brian Brohm, Louisville, QB
Matt Ryan, BC, QB
John David Booty, USC, QB
Colt McCoy, Texas, QB
Chad Henne, Michigan, QB
Tim Tebow, Florida, QB
Ray Rice, Rutgers, RB
Darren McFadden, Arkansas, RB
Steve Slaton, WVU, RB
DeSean Jackson, Cal, WR
Percy Harvin, Florida, WR/RB

Watch that clock to the right, it's almost time for football.

No comments:

Post a Comment