Monday, November 7, 2011

What it will take for Stanford to pass Alabama in the BCS

It was a bit surprising when the latest BCS rankings came out yesterday and Alabama was ahead of not only Boise State but also Stanford.  I'd warned about how this could happen for Boise and Stanford last week and when Alabama was #4 in the Coaches poll, the ball was rolling and the BCS rankings confirmed it had happened.

Now, in Stanford's case they are pretty close to Alabama, just under 0.01, so they have hope.  Boise State is nearly 0.04 back and would need a lot of help.  So let's look at how Stanford could pass Alabama and we'll save looking at a Boise State scenario for when a few of the teams lose they'd need to lose and it is within the realm of possibility.

Obviously, Stanford needs to win out.  They are fortunate to be playing Oregon this weekend, a team ranked #7 in the BCS and a #8 average in the BCS computers.  This will help their strength of schedule, which frankly has not been that impressive thus far.  Should they win, some voters may move them up but they are more likely to see bigger gains in the computers where they have more room for improvement.  Alabama averages #3 in the computers while Stanford is #7.

Also, this weekend Alabama plays a decent Mississippi State.  If they win handily, voters will not be inclined to drop them much.  So for the sake of argument, let's assume the polls stay basically the same but Stanford improves in the computers.

Stanford currently ranks 3, 5, 6, 8, 8, and 10, and average of 7 in the BCS computers.  My BCS compatible algorithm has them #10, similar to the lower end, so it can give us an indication of how much those 3 computers might move them up.  Simulating a Stanford and Alabama win, and also a Boise State win since they are around these teams and playing TCU, this weekend shows Stanford moves up to #8 and Alabama stays #4.

Let's thus assume that Stanford's computer rankings end up being 3, 5, 5, 6, 6, and 8 and they average 5 and their computer percentage moves from 0.770 to 0.830.  As that is 1/3 of the BCS they would move up 0.02 which would be enough to move ahead of Alabama.

All good for Stanford right?  Not so fast my friend.  The season is not over at this point and the computers can still change.

Stanford finishes with with 5-4 Cal (my BCS #60), 6-3 Notre Dame (#33), and the Pac-12 title game (likely ASU or UCLA).  Alabama on the other hand finishes with 8-1 Georgia Southern (non-FBS but ranked equivalent of #63) and 6-3 Auburn (#15).  So Alabama's regular season finishing schedule is slightly harder giving them an opportunity to gain a little back in the computers.  It could come down to who Stanford plays in the Pac-12 title game.

If the season ended today, I believe they'd play 5-4 UCLA (#48) as they are in a 3-way tie with ASU and USC but USC isn't eligible and they beat ASU.  If that is the case, that won't help their strength of schedule and Alabama could gain a fraction more in the computers.  However, if 6-3 ASU (#32) wins out and UCLA falters, playing a 9-3 ASU would likely be enough to keep them ahead in the computers.

I expect it will be very close and it will depend on who Stanford plays in the Pac-12 title game.  Of course, both teams have to win out (not a given) and this assumes Oklahoma State loses so the Alabama/Stanford decision becomes meaningful, that also not a given.

Last, I have assumed the polls stay about the same.  It is possible if not likely that Stanford gets a bump after beating Oregon, but Alabama could get a bump back after beating Auburn, especially if Stanford had a close game with Cal for Notre Dame.  But if the voters decide they don't want an LSU/Alabama rematch, they could change their votes to make sure Stanford moves/stays ahead.