In response to my post yesterday about preseason polls, a commenter questioned whether postponing the initial polls until the sixth or seventh week of the season would have any significant effect. Logic would indicate so, unless we want to believe that in fact we have just as good a picture of the college football landscape in a given year before a single snap is taken as we would after five or six games have been played by the schools.
Fortunately, we can go to the Stassen website for more concrete data on this. Stassen tracks the most overrated and underrated college football teams since 1989 by comparing their initial rankings in the preseason poll to the final polling results.
The results to my mind aren’t that surprising. The underrated teams tend to be those that generally don’t show up on the national radar until the season is well underway and their records begin to catch people’s eyes: Washington State, Oregon, Boise State (watch its numbers begin to become less unbalanced now that it’s caught public attention from the Fiesta Bowl), Louisville, etc. The overrated teams tend to be college football’s royalty: Nebraska, FSU, Southern Cal, Texas, and, of course, Notre Dame.
What that suggests is that in the preseason pollsters are lazy and tend to go with familiarity more than anything. And in many cases, that’s a poor indicator of a school’s merits as the season plays out. Again, you have to stretch logic pretty far to believe that starting the polls midyear wouldn’t produce more accuracy in the initial balloting than what you see here.
Looking at 2006, there are ten teams on the list with deltas in double digits (meaning they went up or down in the polls at least 10 slots). Half of those started or finished the year in the top 10. That’s a pretty significant disparity.
And I would argue that the larger the playoff pool becomes, the more significant the effect of preseason polling on its makeup is. Remember that the BCS now provides slots for ten schools and that a “plus one” format, using that as an example, would select the four top seeds from that group. While it may be easy to identify three obvious teams in ’04 that were worthy of playing in the MNC game, now we’re faced with the relative merits of, say, numbers 9, 10 and 11 to get into the BCS and numbers 4, 5 and 6 to qualify for the playoff. That’s not nearly so clean a problem. Seeding could also be affected.
When you don’t have any objective standards for qualifying for whatever D-1 football considers its postseason, such as, for example, specifying that only conference champs are eligible to play in the BCS title game, it seems to me that there should be an effort made to see that the rest of the formula for selecting the participants is as fair as possible under the circumstances.
Waiting until a respectable portion of the season is under way before ranking the schools is a way to reduce the subjectivity of the selection process to some extent. That’s a good thing. The big boys already have a leg up in so many ways. There’s no need to game the system to their advantage even further.