This is the reality. There is no consistent framework. It's whatever each year's random selection of biased humans decide to weigh. It comes down to simple mental biases: Texas has 3 losses compared to 2. Nevermind they played 5 current top-15 teams and one of those was a 7-point road loss to the defending champ who hasn't had another single-digit game all year. Vandy is named Vandy instead of Notre Dame or Oklahoma. They must suck! Utah's AD is the chairman of the committee.
The sad part is the message it sends about OOC games. Teams are way better off playing a middling ACC or Big 12 program than someone like OSU or Notre Dame. People will counter with examples of playoff teams that had tough OOC games (ND, A&M, tOSU), to which I say - where would that team be ranked if they didn't play that game and got an easy win instead? A&M would still be in the playoffs easily even with a weak-ass SEC schedule. ND would be top 5. OSU would still be #1. Basically, it wouldn't make a difference at all, especially if you have a tough SEC schedule on top like Texas/Bama.
Basically, if Texas had played Tartleton State instead of Ohio St, they would probably be ranked 8 or 9. If they had beat OSU, they would probably be ranked 6-8 (maybe pass ATM and/or OM). The risk/reward of playing that game is nowhere close to worth it.