... 6) Boise dropped after losing at Nevada... but not by a whole lot. That was still a 3-point overtime road loss against a top 20 team; that's a very close loss in a very tough test. Boise has still been an extremely good team the whole year long, and a game that's the definition of a "quality loss" is hardly going to change that fact. While we're at it, I'm sure everyone will notice how Boise is rated over Auburn despite their loss. One interesting thing to note is that the model rated Boise's 3-point overtime loss at Nevada as a materially BETTER performance than Auburn's 3-point overtime win at home against Clemson. Nevada is a much better team than Clemson, it's far tougher to win on the road than at home, and there just isn't much of a difference between a 3-point overtime win and a 3-point overtime loss (both rate as being very close to ties, which makes sense given that they're overtime results, and especially close ones at that). Slice those games out of each team's resume, and you have an undefeated Boise and an undefeated Auburn; Auburn had the tougher schedule, while Boise was far more dominant. For what it's worth, I think that Auburn should be rated over Boise... but I can at least see what the model's doing, and while I disagree, I don't think it's unreasonable. When doing a power rating (which is what this system does), Auburn's nearly unending stream of close to very close wins makes a difference; there's nothing at all wrong with a close win at Alabama (that's still a very good win), but when you also have very close wins against Kentucky, Clemson AND Miss St... that does get factored in.