Listen, you have a system that already has the deck stacked against it in a subdivision of 134 teams only playing 12 regular season games and most times that's only 11 against other FBS opponents.

If whatever system you come up can be empirically proven to provide the most accurate forecasting of games at the end of the season, you can't just summarily throw it out because it offends your very subjective sense of "fairness", whatever that means.