Sorry if there is already a post on this, but I don't see one. (And no, this is not just another post about the Barn. Well, not exclusively.)
I'm watching the BCS Countdown show on ESPN right now and listening to the talk about how the "computers don't like Alabama". I don't know about the other computer rankings, but at least within the Sagarin ratings there is an interesting discrepancy.
Sagarin's "ELO-Chess" ratings, which get used by the BCS, have us ranked way down at #17. So that follows the trend of the other computers in not giving us much credit. However, his "Predictor" ratings have us ranked #3, behind only Oregon and TCU.
The difference in the two ratings is that the ELO-Chess does not use margin of victory. The BCS forced Sagarin to switch to this system. But the Predictor is the one that Sagarin actually puts credence in. So the formula that he trusts says that we are the #3 team in the country right now. Basically it boils down to the fact that our strength of schedule so far has not been as strong as what we would've expected, but with the exception of Arkansas and South Carolina we've won all of our games comfortably.
(Personally I think it was a mistake to remove MOV from the BCS rankings. In my opinion, it's the main factor that makes the computer rankings spit out rankings that sometimes defy common sense. Garbage in, garbage out. If you don't let the computers consider all of the info, then don't be surprised when they spit out odd results. Obviously the voters consider the MOV, so why shouldn't the computers? I'd be fine with putting a cap at say 20 points, so that teams don't have an incentive to run up the score on helpless victims. But I think a 1 point win is clearly less impressive and less confidence-inspiring than a 15 point win.)
The other twist is in AU's rankings. The ELO-Chess (Sagarin BCS) rating has AU as the #1 team in the country. But his Predictor ratings have them way down at #15. They've won all of their games against a decent schedule, but they've had to scratch out close wins in several nail biters. I'm not clear on why they're being billed as a dominant team when they haven't really dominated that many teams on the field.
Looking ahead at the rest of our schedule, LSU is #22 in the Predictor, and Mississippi State is #35. (FWIW, Georgia State is #185.) So at least by Sagarin, we should be favored to win the rest of our games. But of course that only gets us so far, since we were also favored to win all of our games before the season started.
Disclaimer: If the season ended today, AU would be more deserving of being in the BCS championship game than UA. So the Predictor ratings that have them as #15 shouldn't be taken as gospel in determing the CG participants. But then that's why the BCS has the voters too.