Predicting a 35 point blowout as "best case" sounds like you ran 1 scenario through your model. Obviously a best case would have BYU winning the game 59-0 or something similar. LIKELY case is BYU loses, maybe even by 35, but best case? That's plain silly.