Liquor Dealer

The following is currently on ESPN's website:

http://sports.espn.go.com/ncf/news/s...=/bcsroad/1116

Vast Conspiracy

You thought the BCS treated the Pac-10 poorly two years ago when No. 2 Oregon got aced out of the Rose Bowl by No. 3 Nebraska? Listen to this scenario, which isn't that far-fetched.
Ohio State beats Michigan.


LSU loses to Ole Miss or Arkansas or the SEC East champion.


Oklahoma loses to Texas Tech or at the Big 12 Championship Game.


USC beats UCLA and Oregon State.


Ohio State rises to No. 2 in the polls and No. 1 in the BCS ranking.


Oklahoma drops no farther than No. 2 in the BCS.


And USC, No. 1 in both polls, is No. 3 in the BCS ranking and doesn't play in the Sugar Bowl.



Implausible? Sure. But the math works. If that happened, the national championship spotlight would shift to the Rose Bowl, three days before Ohio State and Oklahoma play. The Rose Bowl would not get to host an Ohio State-USC game for the title. But given that the nearby Trojans would benefit, the Tournament of Roses would probably get over it.



Ivan Maisel is a senior writer for ESPN.com


While this is an extreme it shows the fallacies and shortcomings of the current BCS. The current BCS ratings are now posted and this was in reply to some of the changes made in the BCS that are totally contrary to all of the rest of the major polls.

__________________
Life isn't always a bowl of cherries, sometimes it's more like a jar of Jalapenos --- what you say or do today might burn your ass tomorrow!!!

gov135

This is interesting, liquordealer. Two observations, however:
1. If Oklahoma loses, they are going to fall farther than #2 in the BCS. Writers and pollsters drop teams pretty far after late losses, and the computers would not be happy with the Sooners for losing to an unranked opponent.
2. USC is so close to Ohio State, noone really knows what will happen even if OSU wins.

Liquordealer, the scenario presented is not, as the article points out, outside of reason. But I think you need to be careful when looking at arguement like this because: with a couple of assumptions, you can make the numbers do what you want.

Artermis

I disagree with you in this case gov.

There are currently only 3 one loss teams of any merit. I do not see OU dropping more than 2 places if they were to lose a game.

They would still only have one loss over all the 2 loss teams.


Art

RoadRage

At least one of the computers (Colley) would still have Oklahoma #1 if they had lost to BAYLOR. Try it for yourself. http://www.colleyrankings.com/playgod.html Certainly no drop here losing to either Texas Tech or Kansas State/Missouri.

In Colley, Oklahoma is .965 and #2 Ohio State is .891. The difference is .074, and .074 below Ohio State is just above #5 Tennessee. It is reasonable to assume that the larger the gap over the #2 team, the less likely Oklahoma is to drop with a loss. Testing the other computer polls with this example should be a reasonable test of whether Oklahoma will drop with a loss. I'll set the cutoff at #7 to allow for variances in computer's procedures.

Anderson-Hester has #1 Oklahoma at .821, #2 Ohio State at .770. That .051 gap is the same size gap as between #2 Ohio State and #13 Michigan. No drop.

Massey has #1 Oklahoma at 4.752, #2 Ohio State at 4.322, 0.340 differential. That far below Ohio State is between #14 Michigan and #15 Purdue. No drop.

Wolfe has #1 Oklahoma at 8.569 and #2 Ohio State 7.529. At 1.04 below Ohio State is #18 Mississippi. Oklahoma stays #1 with a loss.

Sagarin has #1 Oklahoma at 100.21 and #2 LSU at 93.52, a 6.69 difference. At the same distance below LSU (86.83) is just above #10 Florida. No drop here.

Billingsley has #1 Oklahoma at 334, #2 Ohio State at 308. That difference (26) below Ohio State is less than a point over #4 LSU. Oklahoma should lose a spot or two here if they lose, but Billingsley is Oklahoma based and Oklahoma-biased, so one spot at the most.

I can't test the NY Times because I'm not a subscriber. Since the lowest computer ranking is thrown out, we'll just throw this one out.

With a loss, Oklahoma is still #1 in 5 polls, no worse than #3 in a sixth, unknown and dropped in a seventh. Computer average drops from 1.00 to 1.33 at worst. SOS is 10th currently and can't help but rise playing against a 7-4 Texas Tech and either a 9-3 Kansas State or a 7-3 Missouri.

Hypothetical BCS after Oklahoma's loss in the Big XII title game to Kansas State, with all other game involving these three teams won:

Oklahoma: 1 (loss) + 0.32 (SOS) + 1.33 (computers) + ??? (polls) - 0.6 QW (Texas, assuming they beat Texas A&M) = 2.05 + poll

Ohio State: 1 (loss) + 0.36 (SOS) + 1.67 (computers, #1 from Billingsley and NY Times) + ??? (polls) = 3.03 + poll

Southern California: 1 (loss) + 1.28 (SOS) + 3.16 (computers, gaining one spot in either Billingsley or NY Times) + 1.0 (poll) - 0.2 QW (Washington State, moving up to BCS #9 with Michigan's loss) = 6.24

[NOTE: Ohio State beating Michigan prevents Michigan from being in the BCS top 10, so no quality win.]

As long as Oklahoma's poll average doesn't drop to 4.5 or worse, they still go to the Sugar Bowl with one loss.

__________________
Join TFP Team SETI
43K workunits complete, 34 members, more of each needed.

phyzix525

So won't you vote OSU above USC if it beats Mich? therefore the #1 team in human polls would not be USC. Hell I bet it would still be OU

__________________
It's hard to remember we're alive for the first time
It's hard to remember we're alive for the last time
It's hard to remember to live before you die
It's hard to remember that our lives are such a short time
It's hard to remember when it takes such a long time

The_Dude

RoadRage, what bout the pollsters? AP & Coaches? I can see the coaches putting OU up there (top 3 for SURE), but AP wont be that kind.

__________________
"Luck is what happens when preparation meets opportunity." - Darrel K Royal

RoadRage

Distilled from above posting:

Oklahoma: 2.05 + poll average
Ohio State: 3.03 + poll average
Southern California: 6.24 with a 1.0 poll average

6.24 - 2.05 = 4.19

If my assumptions hold, OU can have a poll average of 4.0 and still be in the top 2.

In there a 2-loss team that COULD be above Oklahoma in the polls if they lose? Yes, Michigan could be, but for Michigan to be there requires them to beat Ohio State, thus meaning all Oklahoma has to do is to keep from slipping behind LSU to stay in the top 2.

LSU and Southern California winning out, Ohio State losing to Michigan, Oklahoma beating Texas Tech and losing to Kansas State (Big XII Championship) :

LSU: 1 (loss) + 1.00 (SOS, assuming beating Georgia for SEC Championship) + 3.33 (computers, assuming up 2.5 spot average in all computers) + 2.00 (poll) = 7.33

Oklahoma: 2.05 + poll

7.33 - 2.05 = 5.28

Nice thing about where Texas is in the polls is that it gives Oklahoma an absolute floor in falling. Ohio State losing to Michigan brings Texas up to 6.0 in the polls, forcing Oklahoma to no worse than 5.0 in poll average.

Likely polls in this second scenario: Southern Cal #1, LSU #2, Michigan #3, Oklahoma #4, Texas #5, Ohio State #6 (remember that LSU beat Georgia in this scenario)

Revisiting the first scenario, the likely polls there: Southern Cal #1, LSU #2 (assuming winning the SEC against Georgia), Ohio State #3, Oklahoma #4, Texas #5

Either event seems to put Oklahoma in, but it requires that Oklahoma drop no further than 1.50 in the computers for the first scenario or 1.67 in the second.

__________________
Join TFP Team SETI
43K workunits complete, 34 members, more of each needed.

Sion

"But I think you need to be careful when looking at arguement like this because: with a couple of assumptions, you can make the numbers do what you want."


100% correct. like this assumption:

"And USC, No. 1 in both polls, is No. 3 in the BCS ranking and doesn't play in the Sugar Bowl."

sure, USC would PROBABLY be No. 1 in both human polls, but not NECESSARILY.

__________________
He's the best, of course, of all the worst.
Some wrong been done, he done it first. -fz

I jus' want ta thank you...falettinme...be mice elf...agin...

bond007

which is the "more likely scenario" in determining the #2 team???

A. the polls are wrong.
B. the BCS rankings are wrong.

in other words... if USC, LSU and Ohio State win their remaining games... how in the world would USC deserve to be ranked #2???

RoadRage

Actually, if you have to assess fault, you have to fault a system that tries to determine the best 2 teams out of 117 given only 11-13 games in which to determine those teams.

To fix the system, you have to either:

1. increase the number of games played,
2. increase the number of "best teams" to determine, or
3. reduce the number of teams from which the "best teams" are selected.

In the end - barring losses by TCU and two of three of LSU, Southern Cal, and Ohio State - someone will cry about being left out.

__________________
Join TFP Team SETI
43K workunits complete, 34 members, more of each needed.