Pretty pleased with this week's effort: better than average amount of work, comparison and revision, and nobody is inadvertently left out. Yes, after hours of looking at wins and losses side by side, I'm more confident than ever that ranking college football teams by any method is one of the most completely arbitrary, contradictory endeavors ever undertaken by man.

Update [2007-10-10 12:22:28 by SMQ]: As promised, notes and justification:

First things first: I have to confront what appears to be an unmistakable SEC bias. It would take a rabid anti-SEC bias to deny LSU the top spot right now – in the first half of the season, the Tigers are still the only team to beat either South Carolina or Virginia Tech, came from two scores down in the fouth quarter to knock off the defending mythical champion and shut out a Mississippi State team that is subsequently 4-1 with a win over Auburn – but it's less obvious the league deserves six other spots, especially when those include once-beaten South Carolina ahead of four undefeated teams and Auburn, Florida, Tennessee and Georgia comprising fourth-fifths of the poll's two-loss membership, the former pair ahead of four high-profile teams with one loss apiece.

I can only answer by saying the teams are getting a lot of value from playing one another. Most teams in the poll have something like one quality win and one quasi-quality win, usually one (like Nebraska, for instance, or Louisville) that looks far less valuable at the moment – in fact, Kentucky drops substantially based on the diminished value of beating Arkansas and Louisville – and the SEC is getting a lot of credit for beating up on one another. In the top five, I roughly measured the value of South Carolina's schedule against those of South Florida, Arizona State, Ohio State and Boston College:

So. Florida	So. Carolina	AZ State	Ohio State
W. Virginia	Kentucky
Auburn	Georgia	Colorado	Purdue
	Miss. State	Orgn. State	Washington
N. Carolina		Stanford	N'western
	LSU	Wash. State	Minnesota
Fla. Atlantic	UL-Laft.	S.D. State	Akron
Elon	S.C. State	S.J. State	Y'twn. State

The actual system I had slightly more nuance, because it included more teams, but you get the basic idea: assuming losing to the undisputed number one team in the country by less than four touchdowns is somewhat akin to beating Washington State by three or blowing out a barely functioning, 1-5 Minnesota team, South Carolina's resume compares favorably with teams that don't have two wins of anywhere near the quality of Kentucky and Georgia – though both of those teams count for significantly less than they did a week ago (even beating Kentucky was worth less on Saturday than it was on Thursday, after Louisville's umpteenth – that is, fourth, in six games – defensive collapse against Utah Friday night, though the Cats are still a better pelt than Purdue, Ohio State's best, or Georgia Tech, the tarnished gem of Boston College's rapidly flattening schedule).

The same method led me to shuffle spots 15-20 a couple times, eventually deciding in favor of the SEC; in actuality, no matter what method you want to employ, there's barely a hair's worth of difference between the resumes of Auburn, Kentucky and Florida right now and those of West Virginia, Wisconsin and Virginia Tech. Every one of those teams is currently serving as someone's "big" win while also holding cards of fluctuating value. Is Wisconsin's escape against Michigan State a good a win after the Spartans fell to Northwestern Saturday as Kentucky's comeback at Arkansas? Where does Florida's admirable effort in Baton Rouge fall in? I had the SEC trio there 18-20 in the first draft and moved it to the front in the last; I see arguments against that decision and can't raise them much.

USC plummets with an atrocious loss and a lot of questions now about the value of its win at Nebraska. Do the Trojans have what might qualify as a "quality win" right now? Certainly there is nothing that negates losing at home to Stanford. There is a valid argument for booting the Trojans out of the poll altogether; I keep them in at 21, ahead of fast risers Colorado and Tennessee and sinking Georgia, all of them with at least one good win (a great win, in the case of Colorado) and "acceptable" losses to their credit.

I cannot and will not bring myself to rank UConn until it beats Virginia this weekend (the Huskies' best win is Pittsburgh, Pittsburgh, and close call or not, it shouldn't come down to one play to beat Temple). Hawaii will not be in the poll for much longer than that. Winner of Texas A&M-Texas Tech Saturday might be in the top 20 next week.

When, as always, everything will be completely different.