Yesterday, I linked to a couple of preseason top 25 polls. I've linked to some others this off season, and plenty more will be coming with preview magazine season and fall practice ahead. There is a veritable cottage industry of preseason polls these days.
So with that said, what kind of preseason poll do you like? There is more than one way to do one, after all.
Projection Poll
This is the most common kind, I think, and certainly every poll regardless of intent gets judged as if it was one of these. Basically, the person putting the poll together takes a stab at predicting where teams will end up at the end of the year. The top two teams are the pollster's prediction for who will play in the national championship game, with the top team the projected national champ.
Power Poll
This kind is the next most common. It's a listing of teams based on how good the pollster thinks they are regardless of where he or she thinks they'll end up. In other words, it's like a projection poll but without allowances for schedule.
The best example I can give for the difference between power polls and projection polls is the Great Preseason Poll Debates of 2006-07. Projection poll makers tended to have West Virginia in the top two or three because they thought the team could run the table against Big East competition. Power poll makers had the Mountaineers lower because they didn't think WVU was truly one of the best two or three teams in the country. Notre Dame has the potential to be a similar kind of team this year.
Snapshot Poll
These are less common, though creators of the previous kinds like to say they actually made this kind when they whiffed on something. Basically, the pollster creates a power poll based on expected quality in Week 1. If they think a team will be shaky at first but come together later, then that team would be lower in this kind of poll than the others.
Think about Auburn of last season. The Auburn team that squeaked out wins over Mississippi State, Clemson, and South Carolina in September wasn't nearly as good as the team that won the most lopsided SEC Championship Game ever. A snapshot pollster would have been justified to have AU ranked in the 20s initially (as most did), whereas projection and power poll makers that had the Tigers that low just missed on them.
Nonexistent
There's a small but growing contingent of people who are against preseason polling of any kind. The rationale is that by setting up a preseason poll, you create poll inertia that is difficult to overcome as the season goes along. These folks usually contend that the first polls shouldn't come out until some time in October. The most high profile adherent of this train of thought is the Harris Poll, the BCS component that doesn't do a poll until into October each year.