Last week we took a look at two of the computersthat factor into the BCS formula - this week, we take a look at two more.
Richard Billingsley Rankings - The Billingsley Report is one of the most controversial computer rankings. Perhaps the factor most difficult to understand is that it begins each season using the final rankings from the previous season. Even though the BCS rankings don't officially start until about midway through the season, Billingsley's system does in fact give teams somewhat of a head start.
Still, Billingsley defends his sytem:
In my system, you carry the rankings from one season to the next, exactly from last season's final rankings. You must have a starting position. Both Sagarin and Massey use that philosophy, the only difference is I found a way to do it without MOV and still have a pretty accurate system. Some people think before the season, everybody should start out equal, but my answer to that is it looks fair on the surface to start everyone equal, but it's not logical because we know for a fact that all teams are not equal, so how can we ignore that? It's more fair if Idaho starts out the same as Southern Cal. It's fair to Idaho. But is it fair to USC? In my mind, (to start out teams equally) skews any hopes of an unbiased SOS. That's why my SOS is dramatically different from what other computers are showing you.
Personally, I don't like the idea of preseason rankings. A national champion that loses half its team to graduation shouldn't start out the season ranked No. 1 simply because that's where they ended the season. But in the end, most of these things come out in the wash anyway, so I guess it's not a huge deal.
Each week, the BCS tosses a team's best and worst rankings and Billingsley's ratings get thrown out quite a bit and has some fairly large variances between the other computers. A good example of this is Pitt's ranking this week. Billingsley has Pitt ranked No. 29, but Pitt comes in at No. 42 in the full standings of the BCS.
Colley Matrix - The Colley Matrix is run by Wes Colley, a PhD from Princeton. The Colley Matrix, unlike the Billingsley Report, starts all teams out equally.
It also does not include any sort of a human factor. Despite this fact, the computer formula has regularly named the same National Champion as the media and coaches. In the words of Colley, it's hard to argue with his computer's results:
So, here we have a scheme to rank college football teams that is absolutely free from human influence or opinion, accounts for schedule strength, ignores runaway scores, and yet produces common sense results, which at the end of the season compare favorably with the press poll rankings. What else do you want?
And as most systems, the Colley Matrix of course factors in strength of schedule. It also factors in top 25 wins and top 50 wins.
All in all, a pretty accurate system at least when it is compared to the human polls.