Well. Here is the formula and explanation:
http://en.wikipedia.org/wiki/Rating_Percentage_Index
As I understand, the rpi is only one of several things examined by the selection committee. But, it is probably a lot more accurate than polls. The explanation above suggests that some don't like the rpi because they think it favors major conferences. I think that if there is a bias, it is on winning. George Washington, for example, still has a high rpi, even after losing to St. Louis. Yet, they really have played nobody other than Maryland and Dayton, losing to Maryland and Gulf Coast. They get credit for beating up on people with winning records who have played nobody to justify those records, like Dayton. Dayton is rated #15 in the rpi despite having beaten nobody that would ever be getting votes for the top twenty-five.
But, if all you look at is records, the teams that play in weak conferences have a distinct advantage. Teams that play nobody have an advantage.
Oklahoma is 16-8 on the season.
Had we played OSU's schedule, we would almost certainly be no worse than 10-1 in non-conference play, meaning that we would be 20-4. If OU were 20-4, we would probably be getting a lot of top ten votes. So, we are being punished for playing a decent schedule.
