12-06-2016, 07:42 PM
That isn't the only criteria they use. In fact, it isn't one at all. It is like a tiebreaker if two teams look close to each other when they rank them.
In 2014, when Baylor and TCU got jumped by Ohio State after the conference championship, when comparing the three, you had a Baylor that had beaten two ranked teams, TCU and Kansas State, and lost to a mediocre West Virginia team, you had TCU who had beaten two ranked teams, Minnesota and Kansas State, and lost to Baylor, and you had Ohio State who had beaten two ranked teams before the championship game, Michigan State and Minnesota, and lost to a mediocre Virginia Tech team. After the conference title game, Ohio State had beaten three ranked teams, had won 11 in a row, was a conference champion, and looked better than TCU and Baylor (they do use the eye test too). If the Big XII had a title game, Baylor would have played in it because they beat TCU, and they would have had the chance to add another ranked win to their resume, and that may or may not have put them above Ohio State who had just beaten a ranked team 59-0.
This year, comparing Ohio State and Penn State, Ohio State beat Big XII champion #7 Oklahoma and Big Ten runner up #8 Wisconsin on the road, #6 Michigan at home, and lost by 3 points on the road to #5 Penn State. Three top ten wins. Penn State had two top ten wins, lost by 3 to 8-4 Pittsburgh and lost by 39 points to Michigan. They weren't close enough for a tiebreaker to matter, even after the conference title game.