A better way to do Top 25 rankings

[StevensTechU]

This is a very offseason kind of post, so let me squeeze it in before there's actual football to talk about.

For the last couple years, I've thought that the best way to do weekly rankings is to look at each team as being what their average performance is. If you looked like world beaters one week but terrible another, your ranking should reflect that rather than reflecting your 'potential' or overweighting one game. What I would want each voter to do is to rank teams 1-25 each week based on their performance that week, and then the actual rankings for each voter would be that and all prior weeks averaged out (or given a weighted average with earlier games given less weight).

This is borne out of frustration with voters completely forgetting how bad a team performed in prior weeks as soon as they have a perceived big win. In college football, every week matters, and the current approach does a poor job of reflecting that.

Thoughts?

[OregonFan4Life]

How would that work with strength of schedule?

[Duck07]

How would that work with strength of schedule?

We'd have dropped out of the Top 25 for our win over the Vandals.

I think what STU is getting at is some of the rigidity of the voting and how teams outside the Top 25 can take a longer climb to get where they should be ranked, whereas the preconceived notions about what a team should be over-value on field performance early on. There's still going to be bias of course, but more fluidity to the rankings early season wouldn't be a bad thing either.

[StevensTechU]

How would that work with strength of schedule?

It's directly factored in. Team A beats Alabama by 14 and Team B beats Mercer by 40, Team A gets ranked #1 and Team B gets ranked #2 because of how you believe they performed that week relative to their competition. Then those rankings live with you for the rest of the season.

Another element I like about it is you can take into consideration injuries. Sometimes beating the Week 3 version of a team is different than beatong the Week 10 version because their quarterback got injured or what have you. That doesn't get captured in formula driven strength of schedules and ratings, but it would be here.

[UOducksTK1]

Early in the season you gotta rely mostly on:
-# of Blue chip recruits/transfers
-Coaching quality
-Starting QB quality
-# of returning starters / experience (and if lots of returning starters, then can include their final ranking from last year to help justify their ranking)
This won't reflect as nicely for schools outside of the bigger conferences, but IDK how else you can accomplish this.

Generally it feels like if you're in the SEC / a historical powerhouse, you get ranked highly. Hopefully we can get away from that.