BCS Computer Rankings - time for another tweak

[Alum81]

5 of the 6 computers have Missouri ranked ahead of OU. Since they were so thoughtful to remove margin of victory out of the computers, maybe they could be intelligent enough to program the computer to recognize that when two teams only have one loss, and one of them's loss is to the other, then they should be ranked accordingly. I think people get that (see coaches poll)

http://cfn.scout.com/2/582114.html

Just sayin..........

[Vegas Vic]

Missouri lost to OU, who lost to Colorado, who lost to Missouri. The computers weigh the totality of the entire schedule, not just the head to head matchups. If OU had lost to Texas instead of Colorado, the computers would have OU ranked ahead of Missouri.

All of this will be moot anyway if OU wins the South and plays in the Big XII Championship game.

[XingTheRubicon]

Exactly, Mizzou has a better loss. It's not that complicated.

[douxpaysan]

Exactly, Mizzou has a better loss. It's not that complicated.
Absolutely...but it is still illogical. I have always thought it unfair for some conferences to have a CCG and others not, but this year I am truly thankful.:rolleyes:

[tommieharris91]

Don't worry guys. The 2/3 human brain component will never let a 1-loss OU get passed by a 1-loss Mizzou. Besides, this issue, as well as Kansas being ahead of OU in the BCS, will play itself out.

[kernalsanders]

well if you think about it, its possible that kansas may still be ranked ahead of us even if we beat them in the conference championship game (assuming ou and kansas win out till the championship), a loss it OU is alot better than a loss to colorado making them outrank us in the computer polls my gosh the bcs sucks

[Ground_Attack]

Look no further than 2000. Miami beat FSU in October while Miami lost to Washington in September. Both teams had 1 loss but FSU got the berth in Championship game. Everything except logic prevails in college football. By the way, that Washington team that beat Miami was also 10-1.

[colleyvillesooner]

that's just a TAD different than a team losing it's CG and going to the NC instead of he team that you just lost to.

I say we waste time complaining about teams higher than us in the computers that we can't do something ourselves.

We have no one to blame but our performance in the last 1 1/2 quarters in boulder. I'll tell you what though... Sitting in that stadium watching that kick go through the uprights, I never thought we'd be talking about stuff like this they way we played that day. Here's to Sam and the boys righting the road woes and tearing Tech a new one!

[Stitch Face]

Exactly, Mizzou has a better loss. It's not that complicated.

This is another example of emotion and opinion and so-called "logic" being more important than what happens on the field. It makes sense for Missouri to lose less 'points' than does OU due to a loss against a stronger opponent, but that shouldn't be more important than the fact that they actually lost to OU in honest-to-God, on-the-field, head-to-head competition.

The computers are basically saying that who you lose to is more important than actually beating one another. It seems a little strange to say 'you may have beaten us in an actual game, but you're a better team than the team that beat you so we're still better than you are.' WTF?

You can compare strength of schedule and margin of victory and "quality wins" or whatever all day, but OU and Missouri actually played each other. OU won. That would seem to indicate that OU is better than Missouri. Call me crazy. Unfortunately who actually wins on the field doesn't seem to matter anymore. We should let the Xbox choose our "champion."

[Alum81]

that's just a TAD different than a team losing it's CG and going to the NC instead of he team that you just lost to.

I say we waste time complaining about teams higher than us in the computers that we can't do something ourselves.

Except in 2003 after the CCG, OU was 12-1 and K-State was 11-3, so while they beat us, they didn't have as good a record as us. My point is when everything else is equal, the computer doesn't know to keep the team that lost below the team that won.

And to your second point, I agree. I'm not complaining about where we are and I think when we take care of business, we'll be in the NC game at the end. I only say that they tweaked the computers once, they can tweak them again - preferably by throwing them off the nearest cliff and going to a playoff! :D

[SoonerKnight]

OU beat missouri. OU will either play KU or MO in the Big XII Championship. OUr loss in Boulder plays a big part of this. I think if OU beats the snot out of TT and okie lite then we will be ranked #3 KU or MO will be #4. The winner of the Big XII should go to the title game. I watched part of the KU game. The two gay announcers were talking about if the quackers were to win out they should play lsu in the title game. They were slobbering all over themselves.

[Leroy Lizard]

Look no further than 2000. Miami beat FSU in October while Miami lost to Washington in September. Both teams had 1 loss but FSU got the berth in Championship game. Everything except logic prevails in college football. By the way, that Washington team that beat Miami was also 10-1.

Except that Miami won on its own field with a missed FG.

[Leroy Lizard]

This is another example of emotion and opinion and so-called "logic" being more important than what happens on the field. It makes sense for Missouri to lose less 'points' than does OU due to a loss against a stronger opponent, but that shouldn't be more important than the fact that they actually lost to OU in honest-to-God, on-the-field, head-to-head competition.

Why do you say that? Appy State has an 8-2 record and beat Michigan, which has a worse record. Should Appalachian State be ranked above Michigan?


You can compare strength of schedule and margin of victory and "quality wins" or whatever all day, but OU and Missouri actually played each other. OU won. That would seem to indicate that OU is better than Missouri.

Sure, just like Colorado is better than OU. And Stanford is better than USC. And Navy is better than Notre Dame. (Wait, that last one doesn't apply.)

OU didn't prove it was a better team than Missouri any more than Colorado proved it was a better team than OU. The purpose of the games is not to prove which team is better.

[Crucifax Autumn]

Huh?

[Stitch Face]

Sure, just like Colorado is better than OU.


The purpose of the games is not to prove which team is better.

Right, I forgot - it's to entertain/distract us rednecks while you college presidents take care of more weighty matters.

Colorado's record is not the same (as good) as OU's at this point, therefore they may have beaten us during our meeting but they are clearly not as good overall.

Right now OU and Mizzouri have the same records in the same conference, and OU beat Missouri head-to-head. OU is therefore the better team until our records shift and/or Mizzou beats us in an actual game.

[Crucifax Autumn]

Yep!

[yermom]

Mizzou also beat Illinois out of conference, their SOS is a bit better than ours

it really doesn't matter where they are ranked though, the only way they are above us when it counts is if they beat KU and then us

of course this also means we aren't all going to be in the top 5 for long, either team will drop like a rock with their loss

[sooneron]

Why do you say that? Appy State has an 8-2 record and beat Michigan, which has a worse record. Should Appalachian State be ranked above Michigan?



Sure, just like Colorado is better than OU. And Stanford is better than USC. And Navy is better than Notre Dame. (Wait, that last one doesn't apply.)

OU didn't prove it was a better team than Missouri any more than Colorado proved it was a better team than OU. The purpose of the games is not to prove which team is better.
Ok, I feel much more retarded now.

[XingTheRubicon]

This is another example of emotion and opinion and so-called "logic" being more important than what happens on the field. It makes sense for Missouri to lose less 'points' than does OU due to a loss against a stronger opponent, but that shouldn't be more important than the fact that they actually lost to OU in honest-to-God, on-the-field, head-to-head competition.


Trust me, the emotion and opinion is all on your side of this issue.

Example:

OU and Mizzou have identical records. Let's say for arguments sake Mizzou's SOS is 4th and OU's is 80th. Mizzou has 4 top 15 wins and OU has 1.
OU loses at home to Iowa State and struggles vs. weak competition, and Mizzou loses @OU in their only poor game of the year.

Is OU still the better team?


Hope this helps.




The computers score the body of work.....without emotion, pride, or prejudice. The process is a bit more convoluted than "We beat theyam, we beat theyam, whyyy would thay be ahead of usss."

[The Maestro]

^^^^^Why losing to Colorado sucks^^^^^

[Stitch Face]

Trust me, the emotion and opinion is all on your side of this issue.

Example:

OU and Mizzou have identical records. Let's say for arguments sake Mizzou's SOS is 4th and OU's is 80th. Mizzou has 4 top 15 wins and OU has 1.
OU loses at home to Iowa State and struggles vs. weak competition, and Mizzou loses @OU in their only poor game of the year.

Is OU still the better team?


Hope this helps.




The computers score the body of work.....without emotion, pride, or prejudice. The process is a bit more convoluted than "We beat theyam, we beat theyam, whyyy would thay be ahead of usss."


I get the unemotional balancing stuff, but to use one of Leroy's arguments: "if the season ended today" who between OU and Missouri should advance to a championship/better bowl/etc.? According to the computers Missouri should. Do you agree?

[Leroy Lizard]

Right, I forgot - it's to entertain/distract us rednecks while you college presidents take care of more weighty matters.

I cannot believe that in this modern world people still believe that a game between two teams proves which team is better.

There are only about a gazillion examples of where a lesser team won. The point of the game is not to prove which team is better.

If you disagree, ask yourself this, "Before OU played Baylor, which team was better?"

How sure were you?

If you were absolutely sure, why bother playing? If the point of a game is to prove which team is better, and we already know which team is better before the game is played, then there is no point in playing the game.

Is Colorado a better team than OU? If you say no, then how could they have possibly won? If the point of the game is to prove which team is better, and Colorado won, then Colorado is better. Period. That is a logical statement that cannot be refuted, but you can try.

The purpose of the games is to compete and to say who won. Nothing more, nothing less. I play in a golf tournament every month against a guy who is clearly better than me. But I still play him. Even if I beat him, he is still the better player. Proving which player is better is not why I play. The same applies to football.

[Stitch Face]

The purpose of the games is to compete and to say who won. Nothing more, nothing less. I play in a golf tournament every month against a guy who is clearly better than me. But I still play him. Even if I beat him, he is still the better player. Proving which player is better is not why I play. The same applies to football.

That's so noble and mature- I just cried a little bit.

Every collegiate athlete should read this so they, too, can know the true meaning of competition. Thank you for your wisdom.

Thank you.

[douxpaysan]

Another something for all you moday morning quarterbacks...Just checked the schedules and six of the top seven ranked teams plays on the road this week. This may be another saturday of surprises.:hot:

[Leroy Lizard]

Every collegiate athlete should read this so they, too, can know the true meaning of competition. Thank you for your wisdom.

Well, it beats your definition of competition. After the Colorado game, our players would have had to turn in their cleats, having proven that Colorado is a better team what would be the point of playing on?

Has it occurred to any of you that the better team may not always play as well as they should? No, it's beyond you. I can tell.

[Stitch Face]

Well, it beats your definition of competition. After the Colorado game, our players would have had to turn in their cleats, having proven that Colorado is a better team what would be the point of playing on?

Has it occurred to any of you that the better team may not always play as well as they should? No, it's beyond you. I can tell.

I think it's safe to say that none of us simpletons understand this sport as you do.

We may watch football. But you see it.


Man, I just got a little choked up again.

[Leroy Lizard]

Another post. Another dodge.

[Stitch Face]

I think in the course of your self-congratulatory windbaggery you've missed my very point - that OU is the "better" team overall (in relation to, say, Missouri, despite what various computers say) even though we failed to show up for the Colorado game. I just don't buy into the notion that we lost to a crappy team and they lost to a better team (us) so we're somehow not as good in the computers' eyes. it's a bogus algorithm if it says Missouri is "better" when we have equal records and we won head-to-head. So, yeah, professor, I get that the "better" team can lay an egg sometimes but still be the "better" team, as may very well be the case with OU right now.

I know the notion of even playing to determine relative skill is beneath you, but that's my argument. I think OU has an edge on Mizzou, in part, because we've shown it on the field once already. You may be too high-minded to care about winning and losing while your golf buddy kicks your *** every week, but fans and athletes and alums do. It's a crazy pride-by-proxy thing. Kinda drives the entire sports industry.

[Stitch Face]

Has it occurred to any of you that the better team may not always play as well as they should? No, it's beyond you. I can tell.

BTW, it's posts like this that keep you squarely in the red. We could send you out to sea on a little raft and you'd survive for weeks on your own sense of superiority.

[Leroy Lizard]

I think in the course of your self-congratulatory windbaggery you've missed my very point - that OU is the "better" team overall (in relation to, say, Missouri, despite what various computers say) even though we failed to show up for the Colorado game.

Okay, so at this point you are acknowledging that head-to-head is not necessarily an indicator of which team is better. I agree.

You haven't explained why OU is better than Missouri.

By the way, it isn't that we failed to show up for Colorado. The buffs beat us. Fair and square.


I just don't buy into the notion that we lost to a crappy team and they lost to a better team (us) so we're somehow not as good in the computers' eyes. it's a bogus algorithm if it says Missouri is "better" when we have equal records and we won head-to-head.

Okay, now you switch back to using the head-to-head argument as the definitive difference between OU and MU.

What if three teams have each beat each other and have identical records otherwise? Say Teams A, B, and C are all 11-1, with Team A beating Team B, Team B beating Team C, and Team C beating Team A. According to your logic, Team A can't be ranked below Team B, but Team B can't be ranked below Team C. But Team C can't be ranked below Team A.


So, yeah, professor, I get that the "better" team can lay an egg sometimes but still be the "better" team, as may very well be the case with OU right now.

Let's use your logic and apply it in MU's favor: MU laid an egg against OU. Therefore, despite the loss, it can still be the better team. After all, if you are allowing OU a mulligan on its loss to Colorado, why can't you do the same for MU? Both games were on the road. MU was playing the tougher opponent.

Don't get me wrong: I am an OU fan. I am just trying to show you why the computers are not being illogical at all when ranking MU over OU. The same thing applied in 2000, when Miami lost to Washington but beat FSU. Even though Miami beat FSU that year, FSU was probably "better" than Miami that year. As I have stated, and you have agreed, head-to-head does not always demonstrate which team is better.

[Leroy Lizard]

BTW, it's posts like this that keep you squarely in the red.

Well, I usually don't make posts like I did until I receive posts like the following:


Right, I forgot - it's to entertain/distract us rednecks while you college presidents take care of more weighty matters.

Up to that point, I had not written any snide comments, so your comment was out line. Fine by me. Just don't snivel when I fire back.

[Stitch Face]

Okay, now you switch back to using the head-to-head argument as the definitive difference between OU and MU.

I don't think it's "definitive." I think it's the best indicator when comparing two teams of equal record from the same conference. Isn't that why a head-to-head win is the tie-breaker for going to the conference championship game when two teams have equal records? If OU and Texas had the same records in the BigXII South at the end of the regular season but OU had won the RRS would you advocate Texas going to the CCG instead if the computers' rated them higher? It's an academic question, obviously, since computers don't play into conference champions, but it illustrates my concern with using computations of SOS, etc. to determine who's "better." I think it's good to balance out the relative lack of "logic" in the human polls, but that doesn't mean the math is always optimal.

[Leroy Lizard]

Isn't that why a head-to-head win is the tie-breaker for going to the conference championship game when two teams have equal records?

No, and I can prove it. Another tie-breaker is the record against teams in your own division. But that cannot indicate which team is better since both divisions have the same rule. In other words, if a North team is given more credence for beating North teams, then why would the South team be given the same benefit for beating South teams?

When teams have identical records, SOMETHING has to decide which team plays for the CCG. Since we want to base it on the outcomes of games in some way, we set up rules that are systematic and easy to understand.

Another way of looking at it: If OU has a 4-1 conference record, with one loss to Kansas, and Missouri has a 4-1 conference record, with its single loss also to Kansas, there is no logical basis for concluding that MU's record is less impressive. They both lost to the same team.


If OU and Texas had the same records in the BigXII South at the end of the regular season but OU had won the RRS would you advocate Texas going to the CCG instead if the computers' rated them higher?

Who did OU lose to? If OU followed up the RRS with a loss to Army, should they still be ranked ahead of UT? IMO, no.

Both teams have one loss. UT's loss is understandable: when two good teams meet the outcome can be decided on a blown play, or a bad call. However, OU' loss is inexcusable: Even with a few blown plays and bad calls, OU would have had no business losing to Army.

Why should the understandable be punished more than the inexcusable?

So there is nothing illogical about the way in which computers rank teams in this regard.

[sanantoniosooner]

Leroy Zepplin broke out the long posts.

You guys are toast...........or bored.

[Stitch Face]

So there is nothing illogical about the way in which computers rank teams in this regard.

That may be going a bit far. Why then do they tweak the formulas and use the average of several different algorithms? If one calculation could be 'completely logical,' so to speak, then why the ongoing modifications?

[Leroy Lizard]

Computers are tweaked because humans have a desire to produce results that reflect their own biases.

If computers picked Hawaii as national champion, then all of those who feel that their team was more worthy complain. So they go back and try to tweak the system to produce a more popular champion. If taking SOS out of the equation means that USC would have gotten the nod over Hawaii, then that is what they do. The following year, the same tweak puts an unpopular team yet again at the top of the polls, so there is a demand to tweak it once more.

The problem isn't the computers; it's our own biases. I say let the computers do their job and quit screwing with them.

[Leroy Lizard]

By the way, since my own algorithm ties all teams together in multi-dimensional space, then there is no problem with the head-to-head argument. And since margin-of-victory is logarithmically scaled, then that problem dissipates as well. I think it is the best algorithm available for doing this sort of ranking.

Another idea of mine was to mask the identify of the schools and mutate the final scores by minor values so that it becomes very difficult to determine which team is which in the polls. All the fans would know are the teams' final records and a rough idea by how much they won by, probably not enough to figure out the teams' identities, especially since the labels would change each week. Then before the bowl teams are chosen, reveal the identities and let the shock begin.

[Stitch Face]

The problem isn't the computers; it's our own biases. I say let the computers do their job and quit screwing with them.

There are multiple, different formulas which have gone through several changes over the years. Which was the 'right' one we should have just left alone?

[Stitch Face]

Another idea of mine was to mask the identify of the schools and mutate the final scores by minor values so that it becomes very difficult to determine which team is which in the polls. All the fans would know are the teams' final records and a rough idea by how much they won by, probably not enough to figure out the teams' identities, especially since the labels would change each week. Then before the bowl teams are chosen, reveal the identities and let the shock begin.

Still wouldn't stop Lynn Swann from voting USC #1 every year.

[Leroy Lizard]

There are multiple, different formulas which have gone through several changes over the years. Which was the 'right' one we should have just left alone?

There is no "right" one. Nor does there have to be. As long as the algorithm is sensible and everyone knows what they have to do each week (namely, win), it doesn't matter which one you use. Just pick one and stick to it.

Now, if there was a clear "correct" result, then we could settle on a computer algorithm that is most accurate. But no such standard exists. If Hawaii was ranked #2 right now by a computer, who's to say it would be wrong? Just our own personal biases.

[Leroy Lizard]

Still wouldn't stop Lynn Swann from voting USC #1 every year.

Unless he can morph into a computer subroutine, it won't matter what Lynn thinks each week.