Partner Series

As college football fans prepare for the bowl season, there is no shortage of opinion on the system we all love to hate: the Bowl Championship Series (BCS). With no national playoffs, the selection of the two teams that play for the BCS national championship is based on a combination of three polls or rankings: the Harris Interactive Poll, the Coaches Poll and an average of the results produced by six computer-based ranking systems.

Of the three components, fans seem to accept the wisdom of the two human-selected polls, but the mysterious computer polls seem to draw universal ire, especially from those fans who feel their team was slighted by a circuit board. Here's a quick tour through the software algorithms. Don't worry if it makes you want to reboot: as you'll see, they don't really matter anyway.

Since its start in 1998, the BCS has included both human polls and computer rankings in its overall formula. The logic of adding automated computer polls was that humans, especially the coaches, would not provide an objective ranking of dozens of teams. Not only were there conflicts of interest of voting a rival's team up or down, there was also the amount of effort to evaluate many different teams without the time to watch or study game results.

The controversy with the computer polls is that their exact algorithms or methodologies are proprietary to their creators, so fans and coaches can only speculate on how their rankings are calculated.

Today's mix of six computer polls all use the same basic set of inputs; who played, who won and the date and location of the game. The BCS does not allow any of the polls to factor in the margin of victory, as they do not want to give coaches an incentive to run up the score on weaker opponents.

So that you can redirect your opinion of the computer polls at their human creators rather than at a pile of silicon chips, here's a quick introduction of who they are and the rough guidelines their formulas follow:

Jeff Anderson-Chris Hester
Publishing their poll in the Seattle Times since 1994, Anderson, a political science teacher, and Hester, a local sports broadcaster, are both University of Washington graduates. They claim to have the most accurate poll and focus on the strength of a team's opponents in calculating their rankings.

Richard Billingsley
The only poll creator who does not have a college degree, let alone an advanced degree, Billingsley wrote his formula more than 30 years ago. Without as much emphasis on game location and home-field advantage, his ranking most mirrors the human polls, giving increased emphasis to head-to-head matchups and preseason polls.

Wes Colley
With the Colley Matrix, a team is penalized for scheduling those early-season cream puffs, as he strongly factors in the win-loss record against strength of schedule. Game data and location have no effect.

Kenneth Massey
Massey is the bona fide numbers geek in the group as he is a professor of mathematics at Carson-Newman College in Tennessee. As proof of his smarts, his research involved Krylov subspaces in the field of numerical linear algebra. Schedule strength and home-field advantage play a part in his Massey Ratings.

Jeff Sagarin

The dean of the group, Sagarin has been ranking teams since 1972, two years after he graduated from MIT. At the BCS' request, he removed margin of victory from his famous formula and now focuses on the basics of home-field advantage, strength of schedule and win-loss record.

Peter Wolfe
Despite working as a physician and professor of infectious diseases at UCLA, Wolfe still finds time to rank both college football and basketball teams. His method is described as similar to Sagarin's, focusing on the basics.

Despite all of the blood, sweat and brainpower these men have put into tweaking their rankings, it has all been for naught. Since 2004, when the current configuration of the BCS poll was introduced, every team that was ranked either first or second in the two human polls has, indeed, played in the national championship game. In other words, the computers and their logic have not changed the selection of the final two teams over what the imperfect humans had decided.

Now, if only college football would join every other major sport and hold a national tournament playoff, we might be able to use our computers for what they were meant for, watching strange sports news conferences.

Dan Peterson writes about sports science at Sports Are 80 Percent Mental.