If you have no priors whatsoever, you start out the season after one week by saying Utah St is WAY better than USC. After all, USU beat Robert Morris by 22 and USC only beat LSU by 7. And without priors, there's no reason to believe LSU is any better a team than Robert Morris. Even if we all know that to be a dumb assumption. And that's why USC was rightly favored against USU in week 2.

Computer rankings are (at least) as much predictive as they are assessment based, and historically it's been found useful to include preseason priors. The important thing to note is that preseason value falls off as the season goes on, so the eventual ranking systems do not remain biased by preseason once there's enough data to make better judgements.

Using only this season at this point would tell us...basically nothing. BYU beat FBS winless Wyoming(#112 by Sagarin) and an FBS winless FCS foe (SIU, #118)... but also SMU. But is SMU any good? If we're only basing it off of this season, SMU barely beat a terrible Nevada team (1-3 and #128 by Sagarin) and one of the worst FCS teams (Houston Christian).