By Martin Kleinbard

It’s not often that people use the word “underdog” in the same sentence as either Melvin Gordon or Amari Cooper, two of the most explosive college football players in the past decade. Yet that’s exactly the role that the star Wisconsin running back and Alabama receiver will play in tomorrow night’s Heisman Trophy presentation, when they will attempt to wrest sports’ most prestigious individual honor from the presumptive favorite, Oregon quarterback Marcus Mariota.

Good luck, boys. Las Vegas has tabbed Mariota with shoo-in 1-10 odds, which would equate the chances of his not winning with those of you actually doing work right now.

For those of you who can’t wait for March Madness (or even the Oscars) for a good office pool, Mariota’s all-but-guaranteed coronation seems like bad news. Luckily, I’m here to whet your degenerate appetite with a pool that works for every Heisman crop, no matter the impending landslide: guess the winner’s share of the total votes.

Like with March Madness, this game has a low knowledge floor—anyone can write down a number between 0 and 100—and it takes a whole lot less time than filling out a bracket with 63 different winners. In other words: George from accounting who thinks that Urban Meyer is a brand of lunch meats is in.

If you’re in a rush and looking for one number to pick, go with 40%. I actually didn’t get that out of thin air; it was the result of a regression that predicted Mariota’s overall share of votes based on the performance of the last ten years of quarterback finalists. In case you were wondering, my regression was made up of only five input variables: number of losses, QB rating, rushing touchdowns, whether the player went to a non-BCS conference school, and whether he was playing in the upcoming national championship game (1).

Before you send me a scathing email on Saturday night after you’ve had too many eggnogs and 40% shockingly doesn’t win you those coveted Monday donuts promised to the winner, understand that there are a couple of mitigating factors. First, my 95% confidence interval—the range within which I can be 95% sure that Mariota’s total will fall—runs all the way from 22% to 58%. Second, my regression wasn’t able to accurately weigh the effect of the competition on the winner’s share. In 2006, for example, Ohio State’s Troy Smith had an excellent year, but his 50% share—the largest of the entire ten-year period—was more a product of a weak crop of contenders than his own greatness.

All of which is a nice way of saying: do your own homework. How much are voters going to factor Gordon’s record-breaking season? Will an East Coast/general Alabama bias work in Cooper’s favor? Will voters hold Mariota’s key lost fumble that led to Oregon’s only loss of the season against him? Do enough voters just hate those space-age Oregon uniforms? It’s all for you to decide. If you’re looking for historical Heisman data, you can find it here.

Lastly, if picking one number is too boring for you, there are plenty of ways to spice things up. Pick the order of the top five vote-getters (which will have to include two non-finalists). Pick the margin of victory over the second-leading vote-getter. Pick whether Mariota will cry after winning. Go nuts—it’s a long time until March.

I could go on, but given the fact that it’s already Friday, you’ve got to get a move on rounding up your colleagues to be in your pool. Happy picking!


Martin Kleinbard is the co-founder of The Bandwagn, a newsletter/website that allows non-sports diehards the chance to join the sports conversation. He can be reached at

1. Other variables like passing yards and touchdowns were encapsulated by QB rating, which is why you don’t see them here. Since this is the first year that the four-team playoff replaced the BCS nonsense, I gave Mariota the benefit of the doubt and marked him a “1” in the national title dummy variable.

You May Also Be Interested In These Stories
[display-posts category=”sports” wrapper=”ul” posts_per_page=”4″]