http://sportsillustrated.cnn.com/2011/wr...t=hp_wr_a2

Letting human voters having a say in this BCS system is like letting your 2 years decide whether to press the button for a nuclear launch or not.

Couple of the money quotes on how stupid and bias some of those people are. They should implement IQ test for these BCS voters. People with IQ with less than 80 should not be allowed to touch the voting button. Not only are some of these people bias, they are dumb as sh*t too.

Quote:The Stanford-Oregon and Virginia Tech-Clemson rankings are an ideal measure of how much voters in the Coaches' and Harris polls actually paid attention to the games. Oregon beat Stanford 53-30. The game was not close after halftime. Yet 45 coaches ranked Stanford ahead of Oregon and 90 of the 115 Harris Poll voters placed Stanford above Oregon.

As much as I like to see Stanford do well since I am minutes away from their campus and I have family that went there, but the fact is Stanford coach who cried about the BCS is voting with the very same intention to manipulate the BCS. Good job David! At least you know how to game the system. I don't expect Stanford to beat OSU. Boise vs. OSU would be far more entertaining.

Quote:Clemson, meanwhile, outscored Virginia Tech 61-13 in two meetings -- including one in Blacksburg. If we learned anything this season, it was that Clemson was definitively better than Virginia Tech. So how many coaches ranked Virginia Tech above Clemson? Thirty-four. Meanwhile, 51 Harris poll voters apparently had better things to do than watch or look up the results of those two games.

I did not read the article, but those quotes leave out a glaring omission: Stanford had one less loss than Oregon, and Va Tech had one less loss than Clemson. It's not like Clemson deserved to be a top 10 team, nor Oregon deserved to be a top 4 team: meanwhile Stanford shoudl not be dropped down further just because their one loss happened to be to a lesser ranekd team, same with Virginia Tech. By the author's logic, no coach in their right mind could rank Oklahoma State above Iowa State. Head to head has to be largely ignored when comparing national rankings, save for maybe when two teams have otherwise similar records and profiles.

Also the computersvoted both Stanford and Va Tech to be higher than their conference counterparts. So I guess the computers would have to be considered to be like 2 year olds as well? Basically this article makes no sense.

(12-05-2011 06:32 PM)adcorbett Wrote:  I did not read the article, but those quotes leave out a glarign omission: Stanford had one less loss than Oregon, and Va Tech had one less loss than Clemson. It;s not like Clemson deserved to be a top 10 team, nor Oregon deserved to be a top 4 team: those other two did.

Also the computers also voted both Stanford and Va Tech to be higher than their conference counterparts. So I guess the computers would have to be considered t be liek 2 year olds as well? Basically this article makes no sense.

The issue I have is humans are SUPPOSE to balance the computers. If nothing else, those humans are using the computer system to game the whole system. That's the problem.

As I said (after you replied) head to head has to be largely ignored when computing national rankings, save for maybe when two teams have otherwise similar records and profiles, because its not like everyone plays each other. That and you end up in a paradox if you try to always rank someone ahead of someone you beat. Head ot head only works in a very small sample size, say of 2-4 teams.

(12-05-2011 06:36 PM)adcorbett Wrote:  As I said (after you replied) head to head has to be largely ignored when computing national rankings, save for maybe when two teams have otherwise similar records and profiles, because its not like everyone plays each other. That and you end up in a paradox if you try to always rank someone ahead of someone you beat. Head ot head only works in a very small sample size, say of 2-4 teams.

This does not change the fact that the entire system is BS. I don't normally want government involved in any sports, but it is time for congress to act. There are simply way too many public institutions acting like street gang thugs with their conference realignment and backroom dealings.

The whole system blows. People just need to stop watching these pathetic match ups. If not, congress needs to step in and force a 16, 24 or 32 teams playoff. There are simply too much money at stake to allow this system to be as corrupt as it is.

10 team playoff any more and you risk diluting the regular season too much. Any less and its pointless.

(12-05-2011 06:52 PM)SF Husky Wrote:  This does not change the fact that the entire system is BS.

The system itself may be BS. However the example given in the article is a piss poor example. You cannot use head to head to determine the ratings of two teams, with different win loss totals, when then not accounting for every head to head matchup. As I said, you end up with a paradox for everyone save for LSU and maybe Alabama.

(12-06-2011 02:03 PM)adcorbett Wrote:  The system itself may be BS. However the example given in the article is a piss poor example. You cannot use head to head to determine the ratings of two teams, with different win loss totals, when then not accounting for every head to head matchup.

While your point is generally correct, in these specific cases it doesn't hold water. Clemson just didn't beat VT, they beat them TWICE, and both by decisive margins, and neither game was played at Clemson. If Clemson had 3 more losses than VT, then it would still make sense to rank VT ahead of them. But just 1 more loss? Surely 2 head-head wins trumps that.

And likewise, Oregon had one more loss, but who did they lose to? #1 LSU, whom Stanford didn't play. So again, Oregon's crushing win over Stanford has to trump the extra Oregon loss.

Yes but then by doing that you either have to rank Clemson and Oregon over other deserving teams just to put them over va tech and Clemson, or you have to arbitrarily rank stanford and va tech below their record, just so that they are belew oregon and Clemson. That is the problem. Then you have to rank Boise under tcu, which puts them well below what they should be worth, and Houston below USM, Kansas St below Oklahoma, Michigan below Michigan St, and I don't even know what to do with Wisconsin and Michigan St who split. By the time you are done you end up with Arkansas at number 3 by default, then 32 other rankes tied for number 4.

As I said, there are too many moving parts to rearrange things due to one head to head matchup. You have to look at each team independently based on their record, who they beat, and who they lost to, and rank them accordingly, but not adjust them in accordance to where the teams they played are ranked. Or your head will explode.

(12-06-2011 05:16 PM)adcorbett Wrote:  Yes but then by doing that you either have to rank Clemson and Oregon over other deserving teams just to put them over va tech and Clemson, or you have to arbitrarily rank stanford and va tech below their record, just so that they are belew oregon and Clemson. That is the problem. Then you have to rank Boise under tcu, which puts them well below what they should be worth, and Houston below USM, Kansas St below Oklahoma, Michigan below Michigan St, and I don't even know what to do with Wisconsin and Michigan St who split. By the time you are done you end up with Arkansas at number 3 by default, then 32 other rankes tied for number 4.

As I said, there are too many moving parts to rearrange things due to one head to head matchup. You have to look at each team independently based on their record, who they beat, and who they lost to, and rank them accordingly, but not adjust them in accordance to where the teams they played are ranked. Or your head will explode.

Again, while your point about transitivity is generally correct, it doesn't hold water in these cases. Oregon and Stanford are easy: They finished 4/5 so just reverse the order. Has no implications for other team's positioning. Even VT/Clemson, an 11/15 spread, is easy. The computers have the three times ranked between them - Baylor, OK, and Michigan - ahead of VT. So just move VT behind Clemson.

Problems solved, in these specific cases. Bottom line is that Clemson clearly merits being ranked ahead of VT, and ditto Oregon over Stanford, and it could be done without creating any crazy hall-of-mirrors ripple effects.