Sunday, January 29, 2012

Post-Bowl Rankings

Better late than never, if I don't get these posted here by the end of January I think I would completely forget about them. Anyway, here we are. I've never been a fan of running the post-bowl rankings. I believe if you look back through my track record of when I usually get around to posting them, my typically long delay represents this lack of enthusiasm in most cases.

This year, I had a gut feeling how it was going to work out even before I ran the calculation. Put simply, our algorithm treats the BCS Championship Game just like any other football game. There is no special indication putting more weight on this game nor is there even any input into the ranking that this game occurred after LSU's earlier victory over Alabama. To top it off, we do not include margin of victory either. So the direct information used about LSU-Alabama is restricted to the fact that they split their two games. The net result is that Alabama isn't selected as the #1 team at any of the bias parameter values, as seen in the plot below.

Is this a clear sign of a flaw? I think that depends on your philosophy in both creating and using such rankings. Once we've all agreed that there is a "championship game" then the outcome is decided on the field, not what some formula says after the fact. Could we modify the rankings to include the date of game or margin of victory or phase of the moon? Sure, but that would be counter to our stated philosophy in making the ranking system simple. Alabama won it all on the field and that's what matters.

Our rankings (at p=0.75) for the full FBS and FCS are available on Kenneth Massey's College Football Ranking Comparison.

Labels: , , ,