I'm no football expert but if there is a formula out there that indicates we've played the most difficult schedule in the country this year, that formula is pretty worthless. We haven't played an awful schedule but let's get real.
I guess if that metric is so valuable to Heather, why did she pick us #3 and not #1?
As good a time as any to discuss the metrics and what they mean.
SOR: Strength of Record
What it is: Relative likelihood that a top-25 team would finish with a record as good or better than another top-25 team, if it played that team's schedule.
What it answers: How good is one team's record vs. that of other teams?
Why it helps: Great for comparing resumes of teams with a differing number of losses. Is 10-0 TCU more deserving than 9-1 Tennessee? SOR says yes, because if Tennessee played TCU's schedule, there's a less than 50% chance it would be 10-0 (and TCU a 51+% chance of having 9+ wins against Tennessee's schedule).
What it leaves out: Victory margin
GC: Game Control
What it is: ESPN statistic of the average likelihood of a team winning each of its game, aggregated across all points of all games, based on score & time remaining
What it answers: How quickly & consistently does a team get ahead and stay ahead of opponents?
Why it helps: It improves on margin of victory by (a) considering point margin over the whole course of games, (b) gives no credit for running up the score once a team is up by 25, and (c) factors in defense (I think); a team leading 17-10 in the 4th Q is more likely to win than a team leading 45-38 in the 4th Q.
What it leaves out: W-L record. The score at 0:00 is just another data point.
FPI: Football Power Index
What it is: ESPN's statistical model that predicts the average number of points by which a team would beat an average team on a neutral field
What it answers: How much better is one team than another? (What would be the point spread at a neutral site if the betting public behaved objectively?)
Why it helps: Helps measure not just which teams are better but how much better, given the expected points added by each team's offense, defense and special teams. Right now, it says K-State would be favored by 0.8 points over the Frogs in a CCG in, say, Tulsa.
What it leaves out: W-L record, but a bigger weakness is what puts in: performance from prior seasons, the impact of which diminishes as a season progresses
TCU has the best SOR of all teams because it has a better strength of schedule than the other 10-0 teams. That will change, though, once Ohio State plays Michigan. TCU's GC (15th) is low because the Frogs have played from behind and won close. TCU's FPI is 16th because it hasn't had a dominating defense--and also, to a degree, because last year's team with a lot of the same players had a losing record.
While ESPN doesn't divulge the details of calculation, it seems unlikely that there is any brand bias in the first two and very little in the third when calculated late in the season.
You could put TCU first based purely on resume to this point--because no 10-0 team has played a stronger schedule, and no 9-1 team would have a 51+% likelihood of being 10-0 against TCU's schedule. The reason most people have TCU 4th is because team-unit performances suggest that Georgia, Ohio State and Michigan would all be 10+ point favorites among a balanced betting public in a neutral-site game. Either Ohio State or Michigan will have the better resume once one beats the other, so it's convenient to keep them both ahead until that happens.
Heather is trying to split the difference, putting TCU third to give SOR a little extra weight. But that's saying that even though TCU has an equivalent record while playing a better schedule to date than Ohio State and Michigan, OSU as a 14-point FPI favorite should be ahead of the Frogs, while Michigan as a 10-point FPI favorite should be behind them.
And I don't disagree that Finebaum is an SEC hack, but he's not crazy to say that TCU has punched above its weight all season, winning close games, making them the most likely good team to slip up this weekend. We can all hate the way he says it, though.