Colley's Bias Free College Football Rankings (last updated October 27, 2004)

Frequently Asked Questions

1. How does your ranking system work?
2. If the computer rankings closely match the press rankings, what's the point of the computer rankings? If the computer rankings differ much from the press rankings, they must be wrong, so what's the point of the computer rankings?
3. Why just wins and losses?
4. How can you possibly think [team A] is better than [team B]?

1. How does your ranking system work?
   Please see this link for an intuitive description, and this link for a complete mathematical description.
   
   
   
2. If the computer rankings closely match the press rankings, what's the point of the computer rankings? If the computer rankings differ much from the press rankings, they must be wrong, so what's the point of the computer rankings?
   In short, there is no reason to believe that the press rankings are more accurate than the computer rankings. I could just as easily say, "If the press rankings closely match the computer rankings, what's the point of the press rankings? If the press rankings differ much from the computer rankings, they must be wrong, so what's the point of the press rankings?"
   
   We have a case where no one knows the "truth," though many think they do. The computers and press are doing their best to estimate the truth. Both have their quirks, but neither has demonstrated greater reliability than the other to date.
   
   
   
3. Why just wins and losses?
   There are several different reasons here:
   • The object of the game is to win, nothing more, nothing less.
   • The NFL, NBA, NHL and Majors use only wins and losses, and do so with very little objection. The big difference is that in the professional leagues there are comparatively few teams and many games. College football features a large number of teams, but a fairly small number of games. As such strength of schedule must be factored in, but I still prefer to stick as closely as possible to the model of the professional leagues: wins and losses should be the ultimate discriminant.
   • Anomalously large margins of victory are difficult to interpret.
   • The home advantage of one team vs. another is very difficult to assess (e.g., Neyland Stadium vs. Wallace Wade Stadium, or FSU at Michigan Stadium in November vs. Wisconsin at the Swamp in August).
   
   
   
4. How can you possibly think [team A] is better than [team B]?
   What I think about a team has absolutely nothing to do with my rankings. My rankings are based strictly upon a unique mathematical prescription, which is applied directly to wins and losses this season (only). Namely, my method has no available adjustments, tweaks or tunable parameters. As such, my method was applied identically in Week 7 of 2004 as it was in Week 3 of 1999. In fact, for my rankings to be influenced by any subjective factor, I would have to possess a magic wand that changed the properties of arithmetic such that 1 + 1 = 2.3. Short of that, my thoughts on teams simply cannot affect the rankings.