Smart Football has moved!

Please check out the new site, All future updates will be made there.

Wednesday, January 07, 2009

What does (should) it mean to be crowned "National Champion"?

There is much debate now whether Utah or Texas ought to share in the college National Championship in some way, though it is unlikely that they actually will). From the WSJ's Daily Fix:

[I]n an open letter to the 72 members of the media who choose the Associated Press No. 1, the Washington Post’s John Feinstein makes a strong case for them to place undefeated Utah at the top of their ballot.

“The reason to vote for Utah is simple: This is the one and only way you can stand up to the BCS bullies — the university presidents, commissioners, athletic directors and the TV networks who enable them — and, to renew a catch phrase, just say no,” Feinstein argues. “Say no to this horrible, hypocritical, feed-the-big-boys system. Say no to the idea that fair competition doesn’t matter. Say no to all the hype surrounding the power conferences and power teams. To co-opt yet another catch phrase, say yes to change.”

Feinstein isn’t the only one who thinks Utah didn’t get a fair shake at the national championship. Yahoo Sports’ Dan Wetzel agrees, pointing out that some voters in the Harris Interactive poll, which determines the title-game participants, admit they didn’t see the Utes play at all this season. “Even by the absurd standards of the BCS, having voters not bother to watch an undefeated team play a single game is a new low,” Wetzel writes. “Whether Utah deserved to be ranked No. 1, 2 or 25 isn’t the point of this argument. The Utes deserved to have voters at least see them.”

But not all agree:

Gregg Doyel disagrees that Utah deserves to be No. 1. “People, please. Utah is the same team that beat Michigan 25-23. Michigan went 3-9 this season,” he writes at CBS Sports. “Utah is the same team that beat Air Force 30-23. Air Force went 8-5. Utah beat Weber State 37-21. Weber State isn’t even in Division I. Utah beat New Mexico 13-10. New Mexico went 4-8.”

I'm not sure if, given a vote, I would vote Utah #1, but Doyel's argument seems bizarre to me. There's a lot of weaknesses with arguing that, because, say, USC beat the Citadel by 20, while Ohio State only beat then by 10, then USC is better. And his argument is just a series of those assertions. I mean, that kind of thing happens all the time and no one believes those predictions because they are entirely unhelpful. All the time one team gets blown out only to turn around and beat a team that blew out the team that originally beat them. Some teams just match up differently; football teams are complex, and some just match up with others better. To paraphrase Herm Edwards, it's about who won. Nevertheless, we saw some of this reasoning this year with Oklahoma and Texas, where Texas's win over Oklahoma eventually counted for less than the subjective impressiveness of Oklahoma's victories down the stretch.

But, more fundamentally, all this made me wonder what the designation "National Champion" is supposed to capture, anyway. The baseline that everyone -- including the President-elect -- seems to push for is a playoff. So we can use that to ask about each view.

Doyel's argument seems to be that Utah doesn't deserve to be #1 because -- "People, please" -- you wouldn't really expect them to beat Texas, OU, or Florida, right? I mean, just look at all their bare victories over mediocre or mid-level teams. In other words, one could phrase the Doyel view as the "National Champion" is the team that you think is the absolute best team in the sense that, were they to be matched up against any other team in the country, they would always be favored to win.

That can't be right, though. That's not at all what a single-elimination playoff gives you. Had the 2007 Giants played the 2007 Patriots the week following the Giants' Super Bowl win, would Eli and Co. suddenly have become the favorite? I think not. In March Madness, with teams playing every couple of days, do we really think that the better team always wins each game? No, and that's kind of the point of a playoff.

Indeed, series-based playoff systems, like with MLB or the NBA, are presumably based on the very idea that one-game is not enough to determine the best team. So, if we still think the playoff is the best solution, then it makes no sense to say that Utah can't be the National Champion just because you think the other teams might actually be better overall. Though, if you subscribe to the Doyel view of "National Champion," then the BCS probably does a better job for you than a playoff would, because the system is all about crowning the perceived best overall team. Although it lacks the precision of a playoff, it gives you fudge-factors so that Florida's and Oklahoma's (though not Texas's) losses can be overlooked.

So, maybe instead of crowning as National Champion the best team in absolute terms, that distinction is a reward for having the best overall season. I don't really watch racing, but that seems to be what they go for with their points system. And many BCS defenders say that it makes "every week a playoff," so the best overall season gets rewarded (let's just pretend like that is true). Well, a playoff doesn't give you that either: Exhibit A - the 2007 New England Patriots. They played unbelievably all year, blew everyone out, and then lost. No one -- not even them -- tried to argue that they should get a share of the Super Bowl via media vote or whatnot.

And that sort of thing happens all the time in playoff systems. It seems like a lot of the recent Super Bowl winners haven't been that great overall, or certainly were not considered the best teams going into the playoffs. Even when the Colts won the Super Bowl, it was with arguably their worst team in something like four or five years. Luck and circumstance play a huge factor, and again, the playoffs are decided by single, permanently binding, contests.

So what does a playoff give you, and why is it probably a better solution for crowning a National Champion? Let me say first that I think it would be a better system than the current BCS morass. But the advantage the playoff gives you is not anything metaphysically correct. It probably does not crown the best team. And it does not reward the best season (sorry Utah).

It merely gives you relative certitude. It's not perfect -- some clunker teams can be crowned, some historically great teams will get the relative shaft -- but, before the season, during the season, and in the playoffs, everyone knows what it takes to be the champion: you must get into the playoffs, and you must win every game once you're there. The Patriots couldn't lobby for votes, they couldn't say that they got jerked around, and they even couldn't say that they didn't get their chance. They played and they lost. They were probably better, they might only have had a bad day, but hey, you knew what you were getting into.

Which is really the issue here. No one has any idea what being "National Champion" ought to mean -- especially in college football where you have over a hundred D-1 programs and no team can come close to playing all the others. A playoff would simply lay some ground rules people could follow. As it stands, without a playoff, everyone may mount their high horse and argue past each other.

Update: Rocky Top Talk and the good Senator both weigh in on this issue. And Bill James says that all self-respecting statiscians should boycott the BCS, because the computer side of the BCS is irrational, incomprehensible, and, worse still, used only to justify the coaches poll: "Throughout the 11 years of the BCS, whenever the 'computer' rankings have diverged markedly from the polls, the consensus reaction has been, we have to do something about those computers. And they have; whenever the computer rankings don't jibe with the 'human polls,' they fix the computers."


Anonymous said...

Where was all of this clamoring for Utah to be in the National Championship game a week ago?
It wasn't there because MOST people thought that Utah was not worthy. It's not like an Auburn 2004 situation, Utah wasn't even favored against Bama. But now that they beat the #4 ranked team, they should jump to #1? That doesn't make sense. The only legitimate reason to give them the National Championship is if one believes them to be the best team in the country. Very few people do, even now.
To play devil's advocate, the BCS does its job quite well. It was put in place to split hairs, to decide based on both subjective and objective factors who the #1 and #2 teams should be at the end of the year. The problem, I contend, is with the people responsible for the subjective factors, not the BCS itself.
I try to see the pros and cons in both playoff and bowl systems and enjoy the game for what it is.

P.S. I do believe though that any playoff should also include a standardization regarding conference championship games.

Anonymous said...

The people who thought Utah weren't worthy are a bunch of know nothings who only know that they play in a midmajor conference and that’s it. The facts are that the Mount West Conference is the third best conference in the country. They are easily better than the ACC, PAC 10, Big 10, and Big East(complete and total joke). The only conferences that are better than the MWC conference are the SEC and Big 12. But if any PAC 10, Big 10, ACC, or Big East team had gone undefeated they would be playing for a national title despite playing in a worse conference.

Anonymous said...

Utah wasn't in consideration because the PAC-10 was viewed as an awful conference, thus the Mountain west's 6-1 record against them was dismissed, and it was widely assumed (thanks Hawaii) that Alabama would blow Utah off the field.

Oops... The "experts" were wrong. The PAC-10 went 5-0 in bowls and Utah blew Alabama off the field.

Championship games shouldn't be standardized anywhere. They're even more of a money grab than the BCS. 9 out of 10 times, one team doesn't deserve to be there.

Anonymous said...

I agree that in a sense the BCS does as good a job with finding the "Best" team as a playoff does.

The BCS selects a champion and only one champion the same as a playoff.

The BCS probably does as good of if not better job at picking the team that would be favored above all others on a neutral field.

The BCS also probably does a better job of picking the team that had "The Best Season"

The main problem with the BCS seems to be that it is subjective and arbitrary in that the whims of the voters can be incosistent. Also it does not give all teams a fair shake. A playoff does a much better job of this as the rules, win and advance, are clear and understood before hand. The only subjective decision is who makes the playoff to begin with.

I think their are two simple tweaks to the system that would solve most of these problems.

1) Create a simple objective formula for ranking teams. This formula would not be intended to say who would win on a neutral field, but instead is a way of saying how you keep score. Essentiall it would say this is how we rank seasons and everyone would know it before hand.

One such formula could be similar to the RPI formula in basketball and be:

(Wins*avg winning percetage of teams you beat+Losses*(1-average winning percentage of teams you lost to))/(total games played)

Everyone would know before hand that this is how you keep score. And their would be no subjective opinion in the rankings.

The matchups for the BCS games could be picked from conf champions plus highest ranked non champions.

2) Play a plus one game after the bowl games. You need this because conference don't play each other enough to get a good ranking, you have a much better idea after the bowls.

You could pick teams for the plus one in order of fewest losses first and then BCS ranking second. So if you had 3 undefeated teams the 2 with the highest ranking after the bowls would play.

This method seems to improve the current system while not changeng it drastically and would score well on most of the criteria of a sytem of appointing a champion. Such as.

Team that would win head to head,
Best Season, Certainty, and objective vs subjective.

Anonymous said...

Under my formula, Utah plays USC, or the winnie of (FLA/OU) in the plus one for the championship game.

Unknown said...

I think a good way to solve it would be for the top four bowls to be "play-in" bowls for the top eight teams. You have four games with the top eight teams, seeded by their respective ranking. The losers of the first four games go to the #3 and #4 bowls, the rest move on. The next weekend would determine the #2 bowl and the National Championship game. The rest of the bowls can be played as usual. Everyone wins.

Anonymous said...

Formula in my above post should have subtacted losses times the opponents loss rate

Anonymous said...

It's nice to believe, but the BCS doesn't even reward the team that had the best season. If that were true, Boise State would have played Utah for the championship because they were the only two undefeated teams at the end of the regular season (and you don't get a better season than undefeated).

More food for thought; Florida, who is playing for the championship squeaked by Alabama in the SEC Championship. Utah destroyed the same Alabama team and is the only undefeated team in the country and people still scoff when it's suggested they be ranked number one. It doesn't make sense, does it?

Also, the people who argue that the regular season is a playoff are wrong as well. If it were, neither Oklahoma or Florida would be playing for the championship because they both lost during the regular season. It's not really a plyoff if you are able to advance after losing.

Anonymous said...

Great points.

A playoff doesn't crowd the best team at the end of the regular season, considering each team's body of work as a whole - just gives you the last team standing at the end of a 6, 8, 16, whatever team tournament. To me that is just as imprecise as taking results from a computer or polls - you get two evenly matched teams, play 100 games, they will each probably win 50. You play one game, it could go either way - might as well flip a coin, it basically just comes down to luck or any of hundreds of different factors. People say that's the only way to truly determine the best team in the nation, which is not true.

I'm a supporter of the polls because it introduces the human element - the stats never tell the whole story, the computer do not know what is actually happening on the field. If you want to have a BCS-style system, the human subjective element has to be a part of it.

Anonymous said...

First, I agree with Chris' prior point that there is no such thing as a Platonic Ideal, so my support for a playoff is not about selecting a "true" national champion. It's more an issue of fairness. Realistically, a team like Utah cannot possibly put together a much more impressive resume than they have done this year. If they can go undefeated and still not win the national championship, what chance does any team outside of the BCS conferences and Notre Dame have? Why keep other conferences in the same division as the big boys if they have no realistic chance at the title?

My proposal: a 12-team playoff, with the first two rounds played at home sites, so that the top 8 seeds get an extra home game. Then the semifinals and finals can be played at neutral sites on a rotating basis. The BCS system or something like it could be used to rank teams nationally and determine the field. Under that system this year OU, Florida, Texas and Alabama would have first-round byes. The first-round games would have Cincinnati at USC, TCU at Utah, Ohio State at Texas Tech, and Boise State at Penn State ... a healthy mix of traditional powers and up-and-comers, in my opinion.

Bowl games could still exist for the teams that don't make the playoffs.

Anonymous said...

By standardizing the conference championship games, I didn't mean that all conferences have one, I meant that either all have one or all don't, not some have one and some don't. The way it is now isn't fair to the conferences that do.

Also, what it boils down to now is that majority rules. The teams that most people think are the 2 best at the end of the year will usually end up playing for the national championship. That is what the BCS gets right.

And once again, there is a reason tht Utah was ranked 6th last week. The reason is most people didn't think they were that good, so there was no controversy that they didn't get a shot at the NC. And now, one week later, having beaten the 4th ranked team by one more score than 2nd ranked team did, they should jump 5 teams to the top spot as the best team in the contry? Nope. Doesn't make sense. The system isn't perfect, but it isn't THAT screwed up.

I don't think anyone who really watches football can argue that Utah's team or resume for that matter are better than whoever wins thursday night.
Deserving....maybe for being undefeated, but if you think the best team should win the NC, I still don't think you vote Utah #1 friday morning, regardless.
And most people who vote will agree with me. Majority rules.

Anonymous said...

Also, comparing the Utes with Florida based on the performance against Alabama is futile. Utah did win by more points, but go back and watch both games.

Against Florida, the Bama team i saw was one who came out and played the best they had all year, by far going blow for blow against a Florida team playing without Harvin. florida ende dup pulling it out in the end by 11 points.

The alabama team in the Sugar bowl as missing their best player, one of the best players in the country, and they came out flat. They were not the same team we had seen the past 5 games. They got down 21-0 early to Utah, but from that point on outscored them 17-10 while missing 2 field goals and playing very mediocre football.

Irrespective of points, florida's performance was significantly more impressive in my eyes.

Judgements like these where the subjective portion of the BCS fails. We could play the "but Utah beat Michigan by 2, New Mexico by 3..." game all day and get nowhere.

SJT said...


I have two problems with your arguement. The first is that it seems to be based on the assumption that what people "think" is right should be what happens. People "think" Team A is superior to team B, so that's who the champ should be. This turns what should be sport into something else: a beauty/popularity contest.

The second problem is that, even if the first assumption is correct, the people who are supposedly choosing the most worthy team are incapable of anything more than an educated guess (at best), and that's not what sports should be about.

The appeal of sports is that they (ideally) shouldn't leave any doubt about the outcome. Football should be played on the field, not on a ballot. Especially considering that there is no way that the voters will ever get to watch enough football to make a truly informed decision, and they will admit as such. So they take tons of information shortcuts in their voting and are often proven wrong.

The fact that Utah wasn't even favored against Alabama proves this point: clearly, what people think is not always reflective of reality, especially in a game as complex as football. People are irrational, biased, short-sighted, short of memory, lazy, stupid, and all sorts of other things. The great thing about sports is that it doesn't matter what people think. You line up and play, and may the best team win.

As for Championship games, they hardly penalize those teams that have to play in them. Florida and Oklahoma are in the BCSNCG precisely because they got to play in a conference championship game. The added exposure of a last minute big win on national television is often enough to push one team ahead of another in the final polls, especially given the status those games had this year as "national semi-finals" of sorts.

Anonymous said...

Only thing that needs to done is to use the damn 16-team playoff system that NCAA uses for Div I-II-II.

The thing I hate about the top NCAA Div is that, like Jay said, many teams do not get a fair shake. No MAC or Sun Belt teams are ever going get to the top BCS game. Why are they in the Div?

I like the way the colleges (NCAA) support fairness.

Anonymous said...

When I read these things or watch people support the BCS on TV it kinds of comes across like the games really don’t matter. When they say "The two best teams" what they are saying is the "Two most talented teams." Not really caring about what they actually did in the season. I'm sure Jerry Jones would love it of the most talented team, *cough*and the largets fan base *cough* got an automatic bid to the Super Bowl.

Anonymous said...

It was just fine with me when the bowls were determined by who won their conference. The pollsters made their choices and we all argued about it. I don't see that much has really changed except for the fact that some schools are now automatically excluded from national championship consideration.

Anonymous said...

Regardless of whether one supports the BCS or not, there can be no question that the most incredibly stupid arguments about sports are put forth be the annual rants of BCS bashers. Since almost no one argues for it, the bashers never feel the need to think before opining.

The BCS is merely a two team playoff. It is a significant improvement over the prior setup. Unless the college presidents in the NCAA decide to go with a playoff, it is likely to continue.

Bill James' criticism is correct. But the blame goes to the BCS bashers whose criticism invariably attaks the wrong parts of the system. It is the critics who keep forcing change in the wrong direction.


david said...

How is the BCS better than the old system when good teams like Utah and Boise State will never be able to participate? In the old system, if enough voters thought they were the best team then they would win the MNC. Now, they don't have a prayer.

Ted Seay said...

Chris: >And they have; whenever the computer rankings don't jibe with the 'human polls,' they fix the computers."

Why should this surprise anyone? It's exactly how the computer forecasts of global warming have been managed...