BCS Rankings Released
by cocknfire on Oct 18, 2009 5:29 PM EDT
in SEC Football
Email
PrintIt's time for the insanity to begin again as the initial BCS rankings of the 2009 season have been released. [PDF] Florida is first and Alabama is second. Your Top 10:
1 Florida
2 Alabama
3 Texas
4 Boise State
5 Cincinnati
6 Iowa
7 Southern Cal
8 TCU
9 LSU
10 Miami
Boise State can never complain about the polls again, as the only reason I can see for them being fourth in the country is the "they won, move them up" philosophy that non-traditional powers have decried for years. And, no, the computer rankings don't "fix" that because they can't take into account things like MOV and the fact that Boise State barely defeated Tulsa.
Iowa is one of the biggest gaps between humans and the robots in the Top 10, which have the Hawkeyes at No. 3 while the flesh-and-blood voters put them at Nos. 7 and 8. Another notable is Southern Cal, which hasn't done much but is No. 4 in both human polls and No. 11 according to the chips. (Win to the computers on both points.)
Houston is behind Oklahoma State -- again the blame goes to those who are supposedly able to recognize things like head-to-head -- and Ohio State is only in the poll at all because the human inexplicably decided to keep the Buckeyes in the Top 25 after losing to Purdue.
Your only other SEC team in the rankings is South Carolina at No. 24, in part because the coaches decided losing by 14 to the No. 2 team in the land isn't a firing offense and in part because the computers have them No. 22.
No. 25 Kansas is also in thanks to the sentient beings. The one thing you can say in favor of computers is that they tend to recognize things like losing to Colorado don't happen to ranked teams.
For more on the BCS rankings, visit BCS Evolution, the SB Nation blog devoted to the most haywire set of sports rankings ever devised by man.
10 comments
|
0 recs |
Do you like this story?
Comments
That doesn’t look so bad overall – I have the first eight in my MP ballot, and nobody lower than 14th. (Going by the computer ranks, I have the first six and nobody lower than 13th. The exclusion is LSU – I check my subjective impressions against FEI, and FEI has really not liked LSU so far.)
Boise’s definitely flying higher than I think they should – Cincy’s looking like the pick of the little five + Big East pack – and I agree that Iowa is probably a little low… though given the two narrow victories over tomato cans, perhaps it’s just as well for Iowa that MOV isn’t considered by the computers.
They are 5-2
A tough FG loss to a better-than-expected Huskie team and a road loss to a top ten squad…decent wins against the Cardinal, CMU and Ore. St….Not a bad body of work. And, this year, it’s hard to find 25 quality teams.
"Hollywood made a movie of my life. The film had me proposing to my wife on the football field. I would never misuse a football field that way." -Crazy Legs Hirsch
by Stuck in the Plains on Oct 18, 2009 9:09 PM EDT up reply actions
Shot At The BCS?
I think they’re gearing up to do another fine job…
"A player who conjugates a verb in the first person singular cannot be part of the squad, he has to conjugate the verb in the first person plural. We. We want to conquer. We are going to conquer. Using the word 'I' when you're in a group makes things complicated." ~ Wanderley Luxemburgo, 1999
Margin of victory
And, no, the computer rankings don’t “fix” that because they can’t take into account things like MOV and the fact that Boise State barely defeated Tulsa.
But that’s the whole point! Computer rankings aren’t supposed to take into account margin of victory because of the philosophical bias that a win is a win is a win, right? So it’s a feature, not a defect of the present system.
More generally, this is the sort of cat-and-mouse game that has led to a lot of evolution of the BCS without any sense that we’re getting better results from one year to the next. Last year people were complaining that the BCS was flawed because it didn’t pick Texas over Oklahoma! They had decided beforehand that Texas was more deserving — rather stupidly also deciding that Oklahoma needed to be penalized for beating Tech too badly — and complained that the BCS hadn’t been jimmied to fit their preconceptions.
The whole thing is lunacy, deemphasizing the human polls and then trying to tweak the computer rankings to simulate the human polls.
I think the problem in this instance is that the computers don’t simulate the human polls – it’s rather silly that the algorithms are barred from incorporating a useful datum that the humans aren’t, and the algorithm writers generally agree.
Of all the year-to-year tweaks – remember the separate SOS component, and the bonus wins or whatever they were called? – suppressing MOV was probably the worst, and complaints about that decision have been pretty wide-spread right from the start. To the extent that the “win is a win” bias exists, it’s on the part of the present BCS administrators, and not the programmers.
Consider how Sagarin rates Iowa – in the version used for the BCS, it’s #3. In his preferred ‘Predictor’ system, based on MOV, it’s #11.
Oh yeah
Obviously Sagarin or whomever* aren’t the source of the “win is a win” bias.
The philosophical notion of leaving out MoV has some basis, I guess. A 1-point win and a 40-point win count exactly the same in qualifying for the NCAA basketball championships. Obviously once you get to an actual tournament, regardless of sport, no one cares about margin of victory.
On the other hand, every other sport has a substantially larger sample of regular season games to determine what teams are worthy of contention. And of course, the Boise problem mostly goes away because other sports invite more than two teams to their single-elimination phase. :)
==
*But Colley does take the “win is a win” approach! I forget if any others intentionally leave out MoV.
The sample+tournament size comment hits it on the head – there aren’t enough games in a season to really sort out two teams based solely on number of victories and SOS. (Unless you get lucky like in 2005.) So the eyeball test – and its statistical cousin, MOV – matter more than in most sports.
And yeah, I wish they’d stop mucking around with the formula. It isn’t as bad as it used to be, I guess – in the good old days they’d change something after every year in response to whatever the crisis du jour had been. I think that’s what happened to MOV, actually – too many tomato cans got crushed too thoroughly, and people started complaining about bad sportsmanship and running up the score etc. (Never mind that ‘style points’ – ie, touchdowns and lots of ’em – still count with the human voters, or that most if not all of the algorithms capped or telescoped MOV.)
And
Point taken that the complaints about this one aren’t really representative of the other incremental changes everyone pines for every year.
Heh, I forgot that cocknfire had commented on that post last year
So maybe the B12S thing last year is my hobby horse, but it’s just the latest next-latest example of what happens every year with people’s dissatisfaction.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
