I want to see Auburn beat Alabama, Miami beat Clempson, Michigan to beat tOSU and tOSU to beat Wisconsin. That will stir the pot in this voting, eye test, recruiting stars, MNC extravaganza.
The best way to pick four teams is to take the best four conference champions. If you are going to use a committee of "voters" then use "must be a conference champion" as the first criteria. Who would complain about the results on the field?
I can, because conference championships are only based on conference games. This makes them a poor indicator of overall quality because they ignore a lot of "results on the field".
First, it eliminates the second best team in the country just because they happen to be in the same conference as the best team in the country.
Second, pretending that non-conference games never happened skews the results. Let's take an example, Team A and Team B. Both are in the same major conference. Team A schedules its three non-conference games against Alabama, Clemson and Oklahoma and wins them all. Team B schedules its three games against Happy Valley Community College, South Eastern Rhode Island State and Vassar and loses all three. Team B loses one conference game and Team A loses to Team B by a point.
This makes Team A 11-1 with wins against Alabama, Clemson and USC and Team B 8-4 with losses to three cupcakes. But since non-conference games don't exist as far as conference championships are concerned Team B is the conference champion and under your scheme would go to the playoff over Team A.
Madness.
Conferences and division winners work fairly well in baseball, pro football and basketball because there is a relatively small number of teams, the skill levels are far closer than in college football and most of the teams play each other. As currently implemented in college football they are a poor indicator of quality, which is why the playoff system only uses them as a tie breaker.
