Everyone who reads in these spaces know that I dislike everything about the BCS and prefer a playoff. The computers are the biggest issue with the BCS; these computers do not even meet the basic standards of for quantitative reliability nor are the held to any mathematical standard. Then there is one computer ranking ran by Richard Billingsley who says himself that he is not a mathematician.
Besides those obvious problems with the computer rankings, and the fact that they do not consider margin of victory, another flaw is that the computers can not compute is how good a team was when the team was at full strength. The main example is Oregon State's James Rodgers, he is an explosive playmaker and Oregon State's best player. In the Beavers game against Arizona Rodgers' tore up his knee as he was taken down in endzone on a touchdown:
The Oregon State team that TCU played and beat in week one is much different, and by different I mean better, then the one that just lost to an average to mediocre Washington team. The computers job is to take the information they are given which range from stats, location of the game and whatever else the secret computer formulas include. These six computer rankings build upon what information they are given as the season goes on, so a loss to Washington obviously downgrades Oregon State's ranking in the computer rankings and in turn effect every opponent they played prior in the season.
However, there is no way for the computer formulas to take into account how strong a team was when they played each other -- which is a problem -- but rather the information is cumulative and effects every team as the season goes on. A small adjustment that probably should be made would be to have some sort of consideration within these computer rankings for how good a team was when they played the game. The amount should be small, but the computers need more information to be more accurate so why not add a section for how good a team was at the time they played the game.
While the human voters should take into effect of how good a team was when they played, but are the coaches (SID's or interns) really going to spend valuable time looking at the injury list or go back and see how good a team was when they played.
Looking at Oregon State's schedule they had a good chance of winning eight or nine games with a tough schedule, but now without Rodgers Oregon State just lost to Washington and now have only Washington State that is a sure win the rest of the way. Oregon and Stanford are now almost for sure loses with Cal, UCLA, and USC toss up games.
With how close the top of the BCS standings are every little bit of those numbers can make a difference. Just look at the difference between Oregon, who is second with a BCS score of .9069, and Boise State who has a BCS number of .8846, so a bump in their computer score by .03 -- which equals only a one spot jump in the computer rankings -- which would put them in the two spot and play for the national title.
So, the loss the James Rodgers from Oregon State looks to have a larger effect on Boise State then TCU, at this time. This is much different then Utah playing Pitt on opening weekend who was ranked 15th when they played and just ended up being a bad team.
While TCU still has a lot of work to do to make the BCS title game and will need some help from other teams losing, the season ending injury to James Rodgers already have cost TCU computer points in the strength of schedule category.
Using computers do actually determine who goes to the title game makes me sick. College basketball does it right with using the RPI as a guide and not the ultimate factor for their post season play.