For those that use previous season(s) in their rating algorithms, even when the "dimensioning" returns stop echoing, the imprint is still there because it influenced the starting point, which forms at least some of the basis for where each team currently rates. It is unavoidable, as far as I can tell, unless the service forgoes such data.
That is a legit open question. I'm maybe too quick to assume that they're conscious of that, and that previous ranking in the season is not one of the variables... rather, that the same variables that added up to ranking for week 8 are the same variables that add up to a somewhat different ranking for week 9. Sagarin being an MIT grad, I suppose, makes me inclined to think the best, but then, there certainly is no shortage of others whose education may or may not lend similar confidence.
While the committee may function something akin to a deliberating jury, bear in mind that the committee (like a jury) cannot un-see and un-hear and disassociate any feelings or thoughts from the seen and heard despite any successful "objection" lodged in opposition.
Right. We agree. But then again, it's one thing to "hear" evidence that says there is a 1:1 billion chance that the DNA is wrong and make a decision about that, and be mandated for whatever reason by the court to disregard that... and at least somewhat another to "hear" evidence of what coaches and sportswriters think, and disregard that... the latter seems more like the legal equivalent of hear-say... would you agree?... and I am inclined to assume it's not that difficult for intelligent people to set that aside.