To take an example we're all likely familiar with, consider BYU and Utah.

Utah has a slightly higher viewership score... But part of that is likely the better opponents.

So who gets "credit" for the BYU/Utah viewership? If I'm going off of initial scores, Utah is slightly ahead so gets credit.

But if I remove both teams results against higher drawing teams, then BYU may (haven't actually calculated it) be slightly ahead. So then I have to recalculate it, assigning "credit" for that matchup to BYU. Which can then bump us ahead of other teams (or Utah behind). So every ripple causes other changes which can then come back and because the first

And that's not even considering the situation where the "higher-drawing team" actually depends on which team gets credit. This could well be the case for BYU/Utah (again haven't run the numbers yet), but I can already near-guarantee it will be the case for Army/Navy.

And of course, I don't want to run the numbers and then decide how to handle things like that because then I'll be tempted to put my thumb on the scale. So I'm stuck right now trying to figure out the best way to go do all that....