Vandy had a great night last night, taking down its first ranked opponent since Jefferson-Pilot was in its infancy.  But perhaps we should be looking at the pollsters who continued to rank South Carolina after their abysmal offensive performance the previous week.  

I think they probably just looked at the score, and there is something striking about a 34-0 victory over an ACC team.  Or perhaps there isn't, considering the ACC was mauled all week.

But anyway, I come not to praise the Gamecocks, or to bury them, but to use them as a starting point to once again rail against pre-season and early-season college football polls and rankings.  Or at least, to rail against how they work, which is to say, they don't work.

The other morning, I was listening to a sports talk show on my way to work, and one of the hosts commented that some pollsters moved the Alabama Crimson Tide from unranked on their ballot all the way to #2.  The word that came out of his mouth and into the microphone was "ridiculous".  Personally, I find it very defensible, and I would appreciate if we saw more of it.

Put simply, preseason rankings are guesses based on no data.  Rankings after the first week are educated guesses based on some data.

No Data vs. Some Data?  Which one is better?  If your answer is "some data is much better than no data", you and I see eye to eye, and we probably both believe it is justifiable to make radical changes to a ranking the first time we have access to some data.

My problem with the radio host's comment is that the person discussed polls and made a huge value judgment in how a few pollsters voted without even once discussing the standards used, or what standards should be used.  His criticism was along the lines of, "How could they not be one of the 25 best teams in the country a few days ago and now be #2 in the country?"  The answer to the rhetorical question is actually easy.  Their previous ranking was based on nothing, and their current ranking is based on actual performance.

It all comes back to standards.  The standards to apply in ranking late in the season is easy.  Late in the season, you rank according to accomplishments on the field during the season.  While I think a lot of voters apply that rule incorrectly, I hope we can all agree that this is the rule that should be applied.

Early in the season, or in the pre-season, it's a different story.  There has been little in the way of "accomplishment" on the field for most team, as most halfway decent teams have played overmatched cupcakes in the opening week.  What standard do we apply to those teams?  How, at this point, do we rank a USC team that thoroughly dominated a team that almost has a pulse against a Georgia team that thoroughly dominated a team that doesn't have a pulse?  Do we try to determine a ranking based solely on that?

If we go with "accomplishments on the field", it is perfectly defensible to rank USC #1 and Bama #2 right now.  Heck, we might even by duty-bound to rank them #1 and #2.  After all, those two teams pretty clearly had the most impressive wins of the first week of the season.  

While I detest early season rankings, I appreciate it a little more when we see voters actually rank teams based on what they have accomplished, but as LSUJonno pointed out recently, the standards seem to change to benefit favored teams.  As he points out, USC jumps Georgia in the polls by thrashing Virginia, but LSU did not jump USC in the polls by thrashing Virginia Tech last year.  Last year, teams said of USC "they're #1 until someone beats them."  At least, they said that with their actions if not their words.  This year, it's a different story.  Georgia doesn't get the "they're #1 until someone beats them" treatment.

I think there simply are no standards, and people apply whatever standards suit their needs, or they don't ever even think about the standards to apply.  I think it's the lawyer in me that wants rigorously applied standards.  When we rank teams, let's at least talk about what we're ranking.  When we rank teams early in the season, are we ranking "perceived ability"?  Or are we ranking "expected performance"?  Those are two different things, and using different standards produces different results.

And when and how do we mix in "actual performance"?

If it's the lawyer in me that wants standards, it's the former engineer (Master's Degree of University of Delaware; Go Fighting Blue Hens!) in me that appreciates actual data.  That's why I loathe and refuse to participate in or consider pre-season rankings.  They are based on no data.  They are at best educated guesses based on past performance and other factors, but in no predictable way.  Their entrenchment ("USC is #1 until someone beats them") leads to problems and injustices, or at least petty aggravations.  Or in the case of the aforementioned radio host, their lack of entrenchment leads to confusion.

Anyway, congratulations to Vandy for beating its first ranked opponent in quite some time, even if that team probably shouldn't have been ranked.  It was actually a really entertaining game, and I'm glad I stayed up to watch it.  Unlike Poseur, I could really give a crap about our strength of schedule at this point.  My goal for LSU is just to win the SEC.  I don't really care how we do it or where we end up ranked if we do.  Besides, Bama's win last week looks like it will elevate our strength of schedule just a little.  So, I like to see entertaining games, and I have a soft spot for Vandy and like to see them do well.