Interesting comparison of the predictive models for college football

This shows how various polls and ranking systems fare when prediiting the winners of post season Bowl games. Surprisingly, PRE-season polls do a better job than the polls at the end of the season. A lot of good information in this article - including some links to investigate.

I must say that looking at the convective metrics make me less optimistic than I would otherwise be about the game this Saturday. On the other hand, i’m buoyed by the realization that had we looked at our metrics this time last year, there is no way we would have projected the finish to the season that we had. To coin a phrase “that’s why they play the games”.