That’s fine. They explained the decision according to the criteria. What should they have done?
Did they fully explicate how they grouped nearly comparable teams?
I don't find Bama and FSU this year to be nearly comparable.
The first three that are nearly comparable are UW, Michigan, and FSU.
Then out of Bama, Texas, Georgia, tOSU it's Texas then Bama because of conference champ and H2H.
Even if you group FSU and Bama, I see that as a draw.
-H2H = n/a
- conf. champs = tie
- SOS = the slimmest of margins for Bama, no metric for determining SOS is prescribed. I reject the ordinal rank of SOS because the size of the league for FBS is so big. The raw values aren't as disparate as one would believe from looking at the ordinal rank for FSU and Bama. Bama has the better top teams on their schedule, FSU has more good but not great opponents. Quality losses count less than quality wins so Bamas SOS advantage is discounted a bit.
- Quality of teams w/ injury considerations = FSU by slimmest of margins, over the past two weeks both teams played in state rivals and legit conference championship opponents. FSU won both by larger margins against marginally worse opponents. So it's a wash based on available data about the teams. If we're discounting for FSU's QB injury, we also have to rebate them a little bit for the opportunity to have 3 weeks of practice with Rodemaker finally getting first team reps and modify their gameplan accordingly. All told FSU slim margin.
So that's a draw. Did you read the protocol in a way that tells you what to do now?
The fundamental problem with the grouping, then comparing process is that FSU gets penalized twice for the injury. They're not in Bama's comparable group if they have JT. So because losing JT brought them down to Bama's level we now ding them for the injury again to rank them within group.
The fact that the published protocol neglects to weight factors, define the size of groups that are comparable, or how the voters should chose the ranks in the rounds of polling to get comparable groups, or how SOS should be calculated is no accident. It's all the window dressing of an objective process without any of the important meat in the bones to justify making decisions that benefit the earning potential of their system.