BCS

[airmail]

Ok, i think i'm with you guys on a play-off.

However, who gets in? Auto-bids? Simply the top 16 get in?

How do you rank the teams? IMO it needs to be a computer and only a computer. Humans are biased.

Good point there... I guess there has to be some selection criteria for the field... you're maybe right about "human error"...

[The Sicatoka]

Precedent (which is what the polls are based on) shows that late season losses, especially in championship games are supposed to hurt worse than early season losses.

They need a playoff, however, a partial, de facto, one was played Saturday:

#1 Oklahoma got rocked by #13 Kansas State

#2 USC knocked off the (unranked?) OSU Beavers

#3 LSU knocked off #5 Georgia

#4 Michigan didn't play (?)

When it was all out there for the taking, an undefeated season and an uncontested championship, Oklahoma couldn't get it done. (Dare I say North Alabama here?)

That tells me that in this "what is available in terms of games left to play" tournament, the championship game should be USC and LSU.

That argument is based on on-field results, not any sort of naivity.

PS - Just so you know my biases here, I can't stand OU, USC, or LSU. Go Longhorns!

[Bisonfan1234]

Can't disagree with you more on the feeling that late season, championship games should mean more than early games. Why? So what if they're championship games?

Each school had one loss, Oklahoma to KSU, LSU to Florida, and USC to Cal. Just on that alone you can see why OU and LSU are in it. Strength of schedule follows suit.

[Smoggy]

Even the computers are flawed. Nobody understands them which makes it hard for anyone to argue with them. They are controlled by a human and if that human likes TCU better, then they get in. A couple of the computers have USC ranked 4th. 4th!!!! How can that happen? Who's ahead of them that has one loss? Boise State? Dicky V said this morning that the winner of the Rose and the Sugar should play. Why not?

Run a playoff with the lesser bowl hosting early rounds and the major ones rotating the ending rounds. Each conference champ gets in, that way conferences can keep their money making championship games. I heard there are 11 conferences. That leaves 5 at large bids for the independents and the other good teams. Use the BCS to determine who gets the at large and who plays where.

The whole thing is comical. You should hear the b!tching out here.

[Bisonfan1234]

The computers can't lie unless the humans make them do it.

If the humans made them do it it would be because they tweaked the formula to reflect one team favorably. But noone knows the criteria so we can't check it publically.

Obviously we would have to make the criteria public so this couldn't happen.

[Smoggy]

The computers can't lie unless the humans make them do it.

If the humans made them do it it would be because they tweaked the formula to reflect one team favorably. But noone knows the criteria so we can't check it publically.

Obviously we would have to make the criteria public so this couldn't happen.

But there would be very few of us that would understand it. Even now, we have to have BCS 'experts' just so us regular folk can understand it. If I had the time I suppose I could figure the dumb thing out. Actually that wouldn't be a bad job. I do have a math degree, so that might help. Hello? ESPN, sure I'll be your computerologist!

[Smoggy]

What gives you the power to say that USC and LSU should be 1/2 over Oklahoma? How can you justify this? If it is simply because they lost to Kstate by 28, then you are indeed naive. So what that they lost by 28? LSU lost to Florida and USC lost to UC-Berkeley. There is no possible way for you to justify what you say unless you can prove that LSU and USC are better than OU. The only way you can do this is by creating a point system based on mesurable quntities (e.g. number of wins, number of losses, number of opponant's wins/loses, number of opponent's opponent's wins/loses, etc.) and then tally the score.

Oklahoma didn't win their own conference. Enough said.

[Bisonfan1234]

That's your opinion, not mine.

So what they didn't win the conference? USC didn't even play in a conference championship game and their conference is weak. The only top 25 team they played was Wash State.

[Shawn-O]

Kirk Herbstreit made a good point today. As far as these conference championship games go, either have them across all BCS conferences, or don't have them at all.

The current system is definitely broken. The frustrating part is that it wouldn't take that much to fix it. Do six autobids from the BCS conferences, and two at-large teams based on the BCS criteria. That would give us Oklahoma, LSU, USC, Michigan, Ohio State, Florida State, Miami, and Kansas State. Bracket 'em, run a playoff within the existing bowl structure, and extend the season two weeks for the semi-finals and finals. Hell, rotate the games among seven or eight sites if needed. Fact is, we are stuck with the current system for two more years after this, and I'm not optimistic that the current system will change much after that.

[Smoggy]

That's your opinion, not mine.

So what they didn't win the conference? USC didn't even play in a conference championship game and their conference is weak. The only top 25 team they played was Wash State.

USC can't play a conference championship game. There is only 10 teams in the Pac 10 and you need 12. If they could've, they would've. It's a big money maker for the conferences, but also a big price to pay if you lose.

USC is only 4 spots behind LSU strength of schedule and the only reason LSU jumped was that they played Georgia for a second time. Yes they won them both. They also beat some really poor DI schools, where USC played a rounder schedule (actually USC played 4 preseason ranked teams, which doesn't mean anything now). LSU is lucky the SEC was so strong this year.

All three teams are deserving of going, but facts are Kansas State is the B12 champion and that's how it'll read in the record book. Oklahoma is a great team that plays in a conference that wants money. They lost and not going is the price they should pay. It sounded like the BCS chairman and other folks were going to be making the change that you need to win the conference, but it wouldn't be fair since some conferences can't play a conference championship game.

[Bisonfan1234]

[Smoggy]

You're missing the point. ONLY the SEC and Big 12 can have a conference championship and they choose to at their own risk. That's why the ACC expanded. It isn't fair, but it's a choice each conference makes.

LSU played in nonconference:

Louisiana Monroe

Louisiana Tech

Western Illinios

Arizona (which USC also had to play)

USC played in nonconference:

Auburn

Brigham Young

Hawaii

Notre Dame

Oklahoma played:

North Texas

Alabama (which LSU played)

Fresno State

UCLA (which USC played)

USC's nonconference finished much better than LSU's did, but USC can't have a conference championship and plays in a weaker conference that they can't do anything about. I'd be willing to bet that if USC played Washington St. again for the conference championship and had won, they'd be in instead of LSU.

Conference championships do have a bearing on the national championship, because if LSU didn't get to play one, USC would be in. I've already said this, but someone from the Big 12 (Kansas State I believe), lost in the conference championship to like Texas back in 2001. NEBRASKA got to go to the national championship and wasn't even a DIVISION champ in their own conference.

[CoteauRinkRat]

Kirk Herbstreit made a good point today. As far as these conference championship games go, either have them across all BCS conferences, or don't have them at all.

That is what I have been thinking for a while now. It doesn't make sense that some conferences play them and some don't.

[sprig]

It should be completely up to the computers to rank the teams.

One of those 'puters that thinks for itself

[Bisonfan1234]

You're missing the point. ONLY the SEC and Big 12 can have a conference championship and they choose to at their own risk. That's why the ACC expanded. It isn't fair, but it's a choice each conference makes.

LSU played in nonconference:

Louisiana Monroe

Louisiana Tech

Western Illinios

Arizona (which USC also had to play)

USC played in nonconference:

Auburn

Brigham Young

Hawaii

Notre Dame

Oklahoma played:

North Texas

Alabama (which LSU played)

Fresno State

UCLA (which USC played)

USC's nonconference finished much better than LSU's did, but USC can't have a conference championship and plays in a weaker conference that they can't do anything about. I'd be willing to bet that if USC played Washington St. again for the conference championship and had won, they'd be in instead of LSU.

Conference championships do have a bearing on the national championship, because if LSU didn't get to play one, USC would be in. I've already said this, but someone from the Big 12 (Kansas State I believe), lost in the conference championship to like Texas back in 2001. NEBRASKA got to go to the national championship and wasn't even a DIVISION champ in their own conference.

It's USC fault for being in a conference that doesn't have one.

Conference champion games *shouldn't* have any meaning on the national championship anyway.

The best idea i've heard so far was adding in opponent's records for the past 5 seasons to the SOS. That way teams wouldn't be penalized for playing teams that should've been good when they scheduled them, but flopped.

[NDSU grad]

It's hard to make the argument that Oklahoma shouldn't be in just because they didn't win their conference when they're sitting at 8-1 in conference play and K-State ("the conference champions") are 7-2. Somebody on here said they need to take conference champions from the "Big Six" and two at large bids. I agree with that except with one caveat, get rid of the conference championship games. The SEC and Big XII should be able to make up the revenue lost from these when their representatives play in the playoff games (the SEC and Big XII should always have at least one team in the playoffs). They could use the current BCS system. This year the eight participants would have been OU, USC, LSU, Michigan, Ohio State, Texas, Miami, and Florida State. Or you could screw the autobids and just give spots to the top eight teams (this is what DII used to do).

[lawkota]

Let's be realistic - we are not going to see a true playoff system anytime soon. The bowls mean everything to the BCS schools. Its a week of pageantry and fun for the team and alumni. Sure, its easy for the college fan to cry for a playoff, but the schools themselves don't want it if it takes away the chance of spending a week in Pasadena soaking in Rose Bowl week. You can argue that its easy to incorporate the bowls into a playoff system but that ignores the week leading up to the game which, to the big time schools, is just as important as the game itself. The best we can hope for is a 4 team playoff after the bowl season, which means extending the season into mid January. After a few years of this, maybe the bowls will lose their luster as a primary goal. When it is the propect of playing for a national title, and not playing in your conference's bowl game, that drives a school, then a true playoff will result.

While the BCS isn't perfect (it's not even very good), it is no worse than the old system of letting the polls decide who is national champion. As a fan, I am taking something positive out of this mess. I get to watch a very good, no, a classic Rose Bowl matchup that could very well produce the AP national champ. 2 days later I get to see Oklahoma and LSU battle in New Orleans for at least a share, if not the outright, national championship. The last 6 years, only one game has mattered. At least this year, 2 games matter.

[Bisonfan1234]

Just taking the top 8 would not work very well, at least until the ranking are not bias (which will never happen).

This is exactly my point for having an all computer ranking system. Computers can never be biased because even if the criteria the computers use is skewed a certain way, it gets skewed for every team! So the team that has the best numbers is the team that gets ranked the highest.

Also i think the tournement should be the best 16 teams, not the best 8.

[The Sicatoka]

Computers can never be biased because even if the criteria the computers use is skewed a certain way, it gets skewed for every team!

Humans have biases.

Computers are programmed by humans.

The criteria and formulas that the computer uses to make its calculations are all based on what, and how, a human told it to figure (how the human programmed it).

In the programming trade it is called "garbage in, garbage out."

Now re-read statement 1 and try to trust "the computers."

A simple example: The DII strength of schedule criteria. Those are biased, skewed, to favor playing other DIIs and were applied evenly to all .... and the computers figured out the numbers exactly that way. Yet, no one will disagree that the computer came out biased against certain schedules.

Why? Because of the criteria, the programming, set by .... humans.

Never be biased indeed.

[Smoggy]

Just heard on the radio that the coaches MIGHT boycott voting in their own poll which is conducted by USA Today. Not sure what this would mean, but the coaches are mad that they HAVE TO vote for the BCS champion and feel that they should actually be able to vote.

[CoteauRinkRat]

Just heard on the radio that the coaches MIGHT boycott voting in their own poll which is conducted by USA Today. Not sure what this would mean, but the coaches are mad that they HAVE TO vote for the BCS champion and feel that they should actually be able to vote.

I heard those grumblings also.

[Bisonfan1234]

Humans have biases.

Computers are programmed by humans.

The criteria and formulas that the computer uses to make its calculations are all based on what, and how, a human told it to figure (how the human programmed it).

In the programming trade it is called "garbage in, garbage out."

Now re-read statement 1 and try to trust "the computers."

A simple example: The DII strength of schedule criteria. Those are biased, skewed, to favor playing other DIIs and were applied evenly to all .... and the computers figured out the numbers exactly that way. Yet, no one will disagree that the computer came out biased against certain schedules.

Why? Because of the criteria, the programming, set by .... humans.

Never be biased indeed.

Numbers, however, don't have biases.

The only thing that matters are things like the number of wins and loses by yourself and the wins/loses of your opponents and opponent's opponents.

Things that don't matter include: Oklahoma shouldn't get it because they lost their conference (although they have a better record than everyone else in it), it shouldn't be USC's fault that they scheduled teams that flopped this year, etc. Things that humans perceive to try and rationalize their biases.

Computers can only be programmed to look at numbers, and like i said, numbers can't lie. Perhaps they can be skewed, like you said, but they will be skewed for everyone which negates the purpose of trying to skew the numbers in one team's favor.

Couple this computer only ranking system with a 16 team play-off and those most bitch we'll ever hear is from the 17th ranked team who has no realistic chance at winning the title anyway.

[PCM]

Numbers, however, don't have biases.

But they are easy manipulated.

[Bisonfan1234]

But they are easy manipulated.

Manipulated, as in skewed?

If you manipulate the numbers for the formula, you manipulate the numbers for every team. They all get the skew evenly, which negates the effect of skewing.

However (i think this might be what you meant), if you manipulate the numbers for simply one team, then this is cheating. One way to combat this would be to make the forumla public so it could be checked by ESPN and whoever else to make sure there wasn't a watergate happening.

[redwing77]

Manipulated, as in skewed?

If you manipulate the numbers for the formula, you manipulate the numbers for every team. They all get the skew evenly, which negates the effect of skewing.

However (i think this might be what you meant), if you manipulate the numbers for simply one team, then this is cheating. One way to combat this would be to make the forumla public so it could be checked by ESPN and whoever else to make sure there wasn't a watergate happening.

ESPN would be a poor regulating source. They favor big schools especially out east and double especially in the New England and Florida (Florida State and Miami) regions.