Tuesday, November 13, 2007

Editor's Note on BCS

As it happens, I pretty much agree with my guest blogger when it comes to the BCS. I agree the whole thing is an illegitimate system that only pretends to add objectivity quality-control to the decision-making process of choosing the national champion. It is, in short, the Kansas of ranking systems: a pretender no one takes seriously when they really think about it.

I actually liked the formulation they used in 2003 and before, where they de-valued human polling and put great emphasis on quality wins. That system was changed because it didn't give results the humans liked.

The current formulation defeats the entire purpose of having a BCS formula, which was to have an mostly objective formula for choosing who would be the teams playing for the title at the end. It emphasized computers, strength of schedule, and quality wins, but it kept USC out of the championship game in 2003 despite the fact that the humans rated them #1. It did, therefore, do exactly what it was designed to do. It used dispassionate formula to trump human biases. The humans protested being overruled, and that system is dead.

They've backtracked just a little on that, by assigning BCS value to the total points teams get in the polls, rather than just the absolute rankings. Therefore, teams that are closely separated in the polls will be closely separated in the poll-portion of the BCS, and therefore easier to jump with solid computer rankings.

Here are some things I believe I will never change my mind about, vis a vis the BCS:
  • Only conference champions should be considered. Any system in which teams that failed to win their conference are allowed a chance to win the national championship invites chaos and sacrifices legitimacy in the proceedings. Will this penalize teams that play in conferences with tough championship games (i.e., everyone other than the Pac-10 and the Big 10)? Yes, but those conference aren't required to have a championship game, and can get rid of it if they choose. Those teams also get the benefit of a late-season win against a team most likely to be rather good. This pleases voters and computers, and I think on balance conference championship games normally help a team get to the BCSCG more than they hurt it. But if you think about it, who are the most ridiculed teams to make it to the BCSCG? Nebraska in 2001 and Oklahoma in 2003, one of whom lost its conference championship game, and the other of whom failed to even get to its conference championship game. Both were subsequently beaten in the BCSCG. Last year, a lot of talking heads wanted a rematch of a game that had been played late in the season between Ohio State and Michigan. Both of those teams got exposed as pretenders in their respective bowl games.
  • The teams have to be chosen based on accomplishments on the field. I don't want to hear about who has the most talent. I want to hear about who wins the games against the best teams. Your team can be as talented as the Patriots, but if they don't get it done on the field it doesn't even matter. And how would you even know if they're as talented as the Patriots? If you're stating a case for your team, your argument should begin and end with what it has done on the field, and should eschew all talk of things like recruiting, potential NFL talent, All-Americans, personality, and coaches. Now, if there are circumstances that serve to put on-field accomplishments in context, it's fine to bring them up, contextually. For example, if LSU had beaten Georgia in 2005, we very much should have talked about how much we accomplished in the face of Katrina and a shifting schedule. As it happened, we didn't have a legitimate claim to a spot in the BCSCG (see above), but those circumstances definitely serve to highlight and emphasize the accomplishments of the season.
  • I agree with Poseur about how those accomplishments should be organized within your arguments. Start with your wins. In particular, start with your best wins. Talk about how you won. Did you win tough games? Did you blow out good teams? Did you win close ones? Then discuss your loss(es), if any. Defend your loss. Feel free to use context, such as the temporary loss of a key player, or something of that nature.
  • I have no faith in human polling. I think it's getting a little better than it's been in the past, as voters are getting more apt to move teams around a little. In the past, the polls were a ladder in which teams moved down when they lost, and teams moved up to replace the teams that lost. That was the only movement in the polls, at least at the top. This season, we've seen the voters be a lot more willing to move a team around based on quality of performance. It's an improvement, but you still hear absolute garbage being spewed like, "They're #1 until someone proves they aren't by beating them." This attitude is nothing more than an excuse not to think. Personally, I believe any voter who ever utters this phrase should immediately have voting privileges taken away. When you think about it, this ladder method is nothing more than a modified computer ranking, with the simplest and least legitimate formula ever devised. Win your game, whether it's by 30 to a power or by 1 to a 1-AA team, and you move up to replace those above you who lost. Lose your game, and you move down. Whenever I hear a talking-head say this, I want to jump through the screen to throttle him. I've seen way too much of this in the past to have any faith in the polls to get things right.
  • No team should EVER base their season on the BCS. It's too arbitrary, too subject to having to decide between legitimate teams, too apt to choose illegitimate teams. I think teams should judge their season based on conference championships. Every team controls its destiny in its conference, and the conference champions are decided objectively, on the field, without ambiguity.
  • Despite all these complaints, I oppose a playoff. It just doesn't solve anything. A 4-team playoff just shifts the argument from "who's #2" to "who's #4". An 8-team playoff shifts it to "who's #8". And what happens when the #8 team that no one thought should be in it wins? Will that be considered legitimate? I'm doubtful.
  • What's the solution? Use the BCS for fun, but get judge the season based on conference championships. The BCS makes for great arguments if you're into that sort of thing, but it is not and should not be the end-all, be-all of season accomplishments.

4 comments:

Poseur said...

I think we need to stress this point: when the BCS actually did what it was supposed to do, overrule the human polls, it was seen as wholly illegitimate and the system was rigged to make that almost impossible to happen again. That's because of two things:

A) The powers-that-be don't want the human polls overruled
B) No one views the BCS as a legitimate criteria to overrule anything.

I am a playoff advocate. I want eight conference champions, which would do several things that would make college football better.

1) It would make a legitimate objective criteria. Win your conference or shut the hell up.
2) It would give the little guy a shot. Either Boise St is D-1 or they aren't. They went undefeated and everyone conceded they had no claim to the title. That's absurd. Let it be decided on the field or kick them out of D-1.
3) It doesn't go overboard. I'm not encouraging the formation of smaller conferences. It gives the best MAC team a shot maybe, but not the Sun Belt.
4) Most importantly: WE WOULD GET BETTER REGUALR SEASON GAMES. Right now, there is absolutely no incentive for a team to play a tough OOC slate. In fact, the incentive is to play a weak OOC slate to avoid a loss. Schedules keep getting worse, and that's a direct result of the BCS. If the only criteria was winning your conference, the regular season would still be a playoff while at the same time, teams could play a tough OOC game to toughen up for the conference slate. There would be no risk to, say, LSU and Michigan playing.
5) Everything gets decided on the field.

Anonymous said...

Your 2nd and 3rd points contradict each other. And besides, what's wrong with saying that Boise State is a good DI school, but not one worthy of a national championship. They aren't this year and weren't last year - just like BYU back in the 80s. Boise needs to schedule GAs, Arizona State's and the like on a regular basis first.


I think your 4th point is more problematic. There's a big risk in LSU playing Michigan in terms of higher risk of injury, for one, and if they don't end up winning the SEC, a lesser bowl if they lose. I think the only real way to boost OOC schedules is with a system that takes schedule strength into account, whether it's the current BCS or a plus one.

shelly said...

Unless you're willing to completely reshuffle the conferences, I want no part of a conference playoff system. As recent bowl results have shown, the top three teams in one conference can be better than the top team in three other conferences in head-head play.

There's not perfect system. They all succumb to that least perfect of all influences - humanity.

Let the BCS system stand and tinker around the edges if you must. Truth told - I enjoy the melee' when there's controversy. Don't you? If you think about it, that's what all this is about anyway - fanaticism! Gotta love it.

Poseur said...

They certainly do not cotradict each other. I want to give the little guy a shot without going overboard. Leaving two slots open for five small conferences accomplishes the goal of giving the little guy a shot without encouraging them to form ten smaller conferences.

Can the third place team in a conference be better than a conference champ? Almost certainly. Texas is the fourth best team in the Big 12 and I believe they would win the ACC outright. But here's the thing: we already know that third place team does not deserve the national title because they lost their conference. If the goal of the system is to find the best team, we already know there is one team better than them: the team that won their conference.

And the BCS has an SOS component and schedule difficulty cratered. College football had much better inter-regional games in the 70s and 80s. SOS is always an imperfect measure, and actually encourages teams to schedule three mediocre teams. If OOC games had no real value, we'd see coaches and AD's, who are incredibly risk adverse for obvious reasons, be able to justify good OOC games again.

And, no. I hate the fact the title comes down to lobbying efforts. It makes a mockery of the sport. We find plenty to talk about in the other sports which don't have a rigged system for crowning their champ.