Originally posted by TheBucsFanI know you explain this poll a lot, but can you - if you care enough to do so - defend Notre Dame at No. 6?
I'll try my best to explain it. The short version is that ND faced teams with pretty tough SOS's (percentage wise) and that's a big component in my formula. Teams are rewarded more for beating high SOS teams than low SOS ones and over the course of the season, that rating for the game will be in flux as the teams win/lose over the course of the season.
Loads of other stuff here doesn't make sense to me either, but that's the most ridiculous of them. Ohio State jumps 25 spots with a win over Indiana?
They jumped so high because of their SOS. I calculate it like how the old SOS component of the BCS was calculated. If Team A gives Team B their only loss of the season, Team B counts as 11-0 for Team A's SOS.
Ohio State faced four teams with one loss. It wasn't so much as the Indiana win, as the other one loss teams they beat keep winning.
USF jumps 10 spots with a win over No. 65 Syracuse?
I'm pressed for time, but at first blush I don't see anything obvious. An educated guess is that the teams ahead of them as teams they beat lose, bringing down their SOS. USF's isn't that good to begin with because Wofford, WKU, and Charleston Southern haven't beat one FBS team among them. I can dig deeper later.
And Boise State was actually No. 1 last week?
I know you've been doing this for several years, but I don't remember your poll ever being as absurd as this one right here. Even if the answer is, "it's early and the poll will sort itself out," there are other computer polls out there that would never have Boise State ranked No. 1 at this point because Boise State clearly does not deserve a No. 1 ranking anymore than Notre Dame deserves to be No. 6.
And yes, with very little data to work with at this point, it will hopefully work itself out as the season goes on. Teams will fluctuate less and less and you won't be seeing all these 10-20 rank moves.
I can't stress this enough, my poll didn't say that by BSU being #1 could beat every other team and ND was the 6th best, just that based on the data at hand, their wins were better than all the others, and that ND did enough to come in at 6th most accomplished. That's all.
Looking back at last season, my results were nowhere near the BCS computers had them this early in the season. Week by week I came closer and closer to what the "pros" had. This system is different than what I used last week and I came in pretty much dead last in internet consensus last week, but I published a week earlier than I wanted.
Did anyone see what happened here? I notice it in the Transactions on the sports page, but I hadn't heard of it and this was all I could find online (for as much as I felt like looking): (from af2.com)