Considerations when creating polls

[UofMLaxGoalie11]

In light of my first attempt to create a top 20, I was just wondering what most people use as criteria to help decide placement. Do you more heavily weigh the happenings since last poll? How heavily do you penalize for losing to unranked teams? How much do you reward for big upsets? Id just like to get a feel for what most people think make a good poll.

[Kyle Berggren]

I make my first poll based on teams last few games (going back to last year if I must), then move teams around as I see/hear of off season improvements. For instance, I moved my own team around because of our age (young) & lost players, knowing we'd either be the team to beat in the PNCLL North, or struggling to get to the playoffs. After that, its as teams play games, I make changes in the couple days before the poll, mark the date, then make changes again from that date forward.

The first poll is always tough, ranking the conferences is difficult at that point. I don't like to move teams down when they don't lose, but I seem to spend the first few polls re-ranking the conferences once OOC games are played. For instance, I had Montana at #1, now after a big loss to UVSC (I don't have all the details yet, but a 12-4 second half throws me off a bit), I'll have to move UVSC up, Montana down in my own poll, I don't care what the combined poll shows (although I will look at it to see if I am missing someone that everyone else is not). Good luck, its easy to spend a lot of time looking though scores and reading game reports...