Monday, September 17, 2007

Polling irregularities

Though it pains me to admit it, the final score from Lexington, Kentucky, on Saturday night reflected football reality: Kentucky 40, Louisville 34. Kentucky was the superior football team in Commonwealth Stadium. The Cardinals will simply have to regroup and run the table for the rest of its schedule.

The anomaly is this: Louisville and Kentucky are both ranked in the AP and USA Today college football polls. Both polls' voters gave Louisville the nod over Kentucky (18-21 in the AP poll, 19-23 in the USA Today survey). But Kentucky won head-to-head.

What explains this sort of polling phenomenon? Presumably the voters were able to absorb the box score, view a few minutes of video, and evaluate the relative strengths of the Louisville and Kentucky teams. Perhaps they believed that Louisville would beat Kentucky on a neutral field, certainly at home. Perhaps they try not to put too much stock in a single game.

Or perhaps, just perhaps, voters try not to downgrade a highly ranked team after an upset because dropping the team hard undermines the entire system of polling.

I still love my Louisville Cardinals, and I hope they will rebound after this bitter loss. In my capacity as an objective evaluator of football, however, I must confess to being somewhat puzzled by the polls.

1 Comments:

Blogger Geoffrey Rapp said...

The college football polls are notoriously “anchored” to questionable pre-season evaluations. I don’t think Louisville’s position rel. KY this week can be explained away by some sort of home victory discounting. It has more to do with the fact that no voters ranked Kentucky in the preseason, yet most had Louisville among their top ten teams. Voters resist re-ordering the teams (perhaps fearing the cognitive dissonance that would follow). This may also explain why Louisville hasn’t fallen more – even though it wasn’t this loss that first revealed the team’s inability to play defense.

Given the strong anchoring that goes on here, it’s not surprising that only a truly shocking defeat (like Michigan’s loss to App. State) can jar voters away from their pre-season preferences.

This also explains why teams that have had great starts and quality wins – Michigan State, Air Force, Cincinnati – aren’t receiving the votes they are due.

Doesn’t US News voting exhibit the same biases?

9/17/2007 2:52 PM  

Post a Comment

Links to this post:

Create a Link

<< Home