Ballot Bias

It's election season again, college football voters, so let me make a plea to your reasonable sides. When you fill out your ballots, don't just look for those familiar names. Take a few minutes and think about what each candidate has done to deserve your vote. It's not time to worry about red states and blue states, but rather to be color blind and make the best decision possible. Barack Obama and John McCain? No way — I'm talking about the BCS polls.

When the BCS was first charged with determining the two teams deserving of national championship berths, it was comprised of many components. However, as controversy was blindly pinned to the system like a tail on a donkey, those components were gradually stripped away. Today's system is a glorified lowest-common-denominator consensus of the "computer" polls and the "human" polls. Gone are the quality win points, the penalties for losses, and most of the other objective measures. And while many superficial critics of the BCS harp on the vagaries of the "computer" element, I think it's time to take a tougher look at the rarely questioned Coaches' and Harris Polls.

First, it's a joke to pretend that either group can take a large-scale snapshot of the college football landscape. There are 119 Division 1-A (yeah, yeah, FBS) teams this season, and Western Kentucky will make it an even 10 dozen next season. That's not a landscape; that's a latitude line. Am I supposed to believe individual media members, let alone very busy coaches, are able to keep tabs on that many teams?

But let's just pretend, in the name of keeping this column going, that each coach or Harris Poll panelist does his due diligence and reads box scores, calls local media members, and splurges on the extended satellite TV package. Let's compare some of the pollsters' work from their latest submission. I've used the November 2nd Coaches' rankings for two mystery teams below, extending their opponents' rankings into the unofficial "others receiving votes."

Team A

Record: 9-0

Games Against Teams Getting Votes in Coaches' Poll:
7-point victory over Coaches' No. 12 team
31-point victory over Coaches' No. 31 team
14-point victory over Coaches' No. 33 team

Team B

Record: 9-0

Games Against Teams Getting Votes in Coaches' Poll:
6-point victory over Coaches' No. 7 team
42-point victory over Coaches' No. 26 team

Admittedly, this analysis focuses solely on the big games these two squads played. However, given that each escaped mostly unchallenged from the rest of its schedule, it seems reasonable to say that Team B has posted a slightly more impressive resume. And the computer polls agree with this assessment; in Sunday's latest BCS rankings, Team B was tied for the top ranking in the computers, while Team A was fourth. However, both sets of human pollsters did not see it that way. They had Team B third in each poll, with Team A second. Sure, the resumes are similar, but what set Team A apart for the human voters?

And now the big reveal: as I'm sure most of you divined by this point, Team A is Penn State and Team B is Texas Tech. This is the major bias that creates discord between the computer element and the human polls. To the pollsters, Penn State was slotted behind Alabama and well ahead of Texas Tech going into their bye on Saturday. To them, why should they be bypassed without playing?

Not convinced? Let's try another. Here are two more mystery teams and their resumes, as judged by the Harris Poll this time.

Team C

Record: 7-1

Games Against Teams Getting Votes in Harris Poll:
45-point victory over HP No. 39
32-point victory over HP No. 11
6-point loss to HP No. 38
34-point victory over HP No. 32
7-point victory over HP No. 37

Team D

Record: 8-1

Games Against Teams Getting Votes in Harris Poll:
10-point victory over HP No. 5
25-point victory over HP No. 13
4-point victory over HP No. 8
6-point loss to HP No. 3

While Team C could point to its somewhat more impressive margins of victory in these games, Team D clearly proved more by winning against higher ranked teams. Team C's schedule has had one more "big" game than Team D's, but that difference seems more-than-eliminated considering that only one of those games was against a team currently in the Harris Poll top 25.

Again, these are fairly similar teams. However, I wouldn't hesitate to say Team D has proven more, based on these results, than Team C. And, once again, the computer polls agree with that opinion, placing Team D third in their rankings, while Team C is 10th. However, not only did the human voters narrow that gap, but the coaches placed Team C sixth, immediately in front of seventh-ranked Team D (in the Harris Poll, Team D is sixth, and Team C is seventh).

The two mystery teams: Team C is USC and Team D is Texas. On a side note, it's interesting to see that USC's supposedly horrendously easy schedule has actually produced more games against teams receiving votes than Texas' perceived meat grinder. But once again the time element came into play, this time in that the Trojans' loss to Oregon State, despite being to a far inferior conqueror than Texas' loss to Texas Tech, came earlier in the year. The human polls resemble an annual game of musical chairs more than any kind of objective ranking system.

So where do we go from here? At the risk of being labeled a college football anarchist, maybe it's the human element that put noise into the BCS system and not the maligned computers. After all, in many states, the high school playoff teams are determined and seeded purely by a computer points system. After all, how are coaches or writers supposed to fairly judge dozens of teams spread across a state?

I know, college football purists, that's heresy. We need our pollsters to maintain order in the universe, so that the machines don't go Agent Smith on us and use our passion for college football to power their Matrix world, or even worse — allow a mid-major into the title game.

So, dear voters, please keep these lessons in mind as you punch those ballots. Hanging chads and long waits at the polls are mere inconveniences compared to getting the wrong teams to Miami this year. Now that would be a national crisis.

Leave a Comment

Featured Site