You can get all sorts of wacky stuff early in the season.

At least ECU beat two D-1 schools, which is more than most programs can say at this point in the season. Add in the fact that one was on the road, and they might as well be #1.

It’s not necessarily a great idea to avoid human bias entirely in ranking. Hinton’s method, which is logically and intuitively attractive, obviously introduces personal opinion into the process. But resume ranking teams using opponents’ records from the previous year, or anything along those lines, could avoid human-introduced bias entirely — ranking reduces to a calculation — at the expense of a huge increase in error likelihood, because the metric used might suck (such as the one I mentioned).

Bias isn’t bad if it leads to smaller error, but resume ranking this way in the first part of the season is a waste of time. ECU deserves zero first place votes. For me, its attractiveness has always been that there at least exists a possibility of avoiding these late-season, hand-wavy, neutral-field arguments which are guesses and inevitable with power polling.

Far more offensive than all of this, though, is the underlying assumption in ranking that an ordering of teams actually exists, when evidence clearly suggests it does not. I wouldn’t be bothered by skipping the polling and just giving some random team Burger King crowns at the end of the year.