Allen Barra pens a typically unfocused and wrongheaded piece for Slate concerning the BCS. He's not opposed to the BCS itself, but that the BCS no longer uses margin of victory as a part of its formula. Of course, until the seventh paragraph of the piece I thought Barra was opposed to the BCS altogether.
Putting aside Barra's failure state his theory until halfway through the piece, he presents unsupported statements of opinion as facts, citing no evidence whatsoever, and makes utterly incoherent arguments to support his points. Taking it point by point:
It used to be so simple. Beginning in 1936, the Associated Press would poll a group of leading sportswriters and broadcasters on who were the best teams in college football; United Press International, starting in 1950, would ask the same of the coaches. At the end of the season, the team with the most points was the national champion—that is, the unofficial national champion, since there was no provision by the NCAA for a champion in college football. The bowl games weren't set up to pick a champion; for the most part, they were simply considered postseason rewards. It wasn't until the 1970s that both AP and UPI finally agreed to postpone their final votes until after the bowls.
Emphasis is mine. The bowl games are still, and always have been, nothing more than postseason rewards for every bowl team not in the national championship hunt; in other words, all of them save two or maybe three. Even under the BCS, a bowl win doesn't mean much unless a team is playing for the BCS Championship. (Aside: while it is technically possible to have a split national championship even with the BCS in place, i.e., the AP Poll No. 1 team is not the BCS Champion, such a scenario is highly unlikely.)
The whole system—to the degree that you could call it a system—didn't exist to resolve controversy but to cause it. As AP guru Alan J. Gould said in an interview, "It was a case of thinking up ideas to develop interest and controversy between football Saturdays. Papers wanted material to fill space between games. That's all I had in mind, something to keep the pot going. Sports then was living off controversy, opinion, whatever. This was just another exercise in hoopla." And one that worked very well. That the season sometimes ended in controversy with fans of schools in different parts of the country all arguing about who should be No. 1 didn't detract from the interest. It enhanced it.
Fair enough. But what does this have to do with the BCS formula not taking margin of victory into account?
Now we have the Bowl Championship Series, the inevitable result of, in the words of former coach Bill Curry, "Trying to fix something that wasn't broken in the first place."
Two points. First, and I'll get to this in a minute, the system was broken before the BCS. That's why we have a BCS. Second, who cares what Bill Curry thinks? He was a terrible coach, which is why he's doing color on ESPN's third string broadcast team. And he sucks at that, too. Listen to Curry do one quarter of a football game then tell me if you put stock in anything he says about anything.
Real college football fans never complained about the lack of a playoff system. They might have complained now and then when their team didn't win, but that's a different matter. The ones who harped for college football playoffs were mostly from the northeastern media, writers who saw college football as an appendage to the pro game and couldn't understand why all the hicks in Norman, Tuscaloosa, South Bend, and College Station couldn't see it that way, too.
Nonsense. College football fans, real and imagined, and many in the media (not just in the northeast) have been complaining about a lack of a playoff system for years and years. And it's not just "when their team didn't win," but every single year. The impetus for the BCS was the constant carping by fans and writers regarding how the national champion was determined. As for whether college football is "an appendage to the pro game," no one can seriously deny that it is. The media coverage the NFL combine draws every year shows this is so. So does the Maurice Clarett fiasco: Clarett was prepared to make the an antitrust argument in his lawsuit against the NFL seeking early entry into the draft that college football serves as a de facto minor league for the NFL, which exploits the unpaid college athlete. At least one legal scholar agreed that Clarett has a strong case. So to suggest that the college game is not an appendage to the NFL (not for every player, but for some) is, at best, disingenuous.
So, now we have the BCS, whose major contribution to college football has been to erode interest in all bowl games except the one that has the championship, while much of the old controversy remains intact.
Where is Barra's evidence that interest is declining? Is attendance declining? Are ratings falling? Are bowls losing sponsorship dollars? Further, if Barra is right that interest in all bowl games except the BCS Championship game is waning, I would bet that that has to do with the sheer number of bowl games, 28, by my count, which means nearly half of all Division I-A teams will get a bowl bid.
Last year, the BCS screwed up by devising a system that kept perhaps the best team in the country, the University of Southern California Trojans, out of the title game. Southern Cal had finished the regular season with a 10-2 record against one of the toughest schedules in college football history.
So, Barra's saying that it's the BCS's fault that USC lost two games? Please. USC was not the best team in the country. You wanna know how I know that? They LOST TWO GAMES. Finally, note the irony that Barra's poster child for all that's wrong with the BCS would be an appropriate poster child for a playoff system, which Barra dismisses in his third and fourth paragraphs (as would the 1989 Fighting Irish, the 1994 Nittany Lions, and the 2001 Ducks).
Last year I ranked teams for my column in the Wall Street Journal with Professor George Ignatin of the University of Alabama in Birmingham and our computer, Mad Max. Going into the bowl games, Max gave the Trojans an edge of nearly 10 points over eventual national champion Ohio State, even though Ohio State was unbeaten. Why? Simple: Southern Cal had faced nine teams whose collective won-lost records were, when not playing Southern Cal, 75-26. They lost to two of them, both on the road—to Kansas State by seven points and to Washington State by three in overtime. SoCal beat the other seven bowl-worthy teams by a staggering total of 175 points—25 points per game.
Tough schedule or not, USC lost two games. Ohio State lost none; Miami (FL) lost none. A team with two losses should not be rewarded at the expense of teams with no losses or even one loss (assuming those teams played a decent schedule). That's just silly. And the fact that USC blew out "seven bowl-worthy teams" is not that impressive considering how many bowls there are. Finally, how are we to grant Mad Max's formula any legitimacy when it gives so much weight to strength of schedule and margin of victory that a team with two losses can still be ranked ahead of teams with no losses?
Yet the Trojans got no credit from the BCS for their margin of victory, an absolutely absurd proposition when you consider that Team A and Team B could play the same schedule and have the same won-lost record, Team A could win each game by one point and Team B could win each game by 21 points, and, by BCS logic, have exactly the same ranking.
Finally Barra puts forth his argument, but attempts to support it with a thought experiment that is, um, "an absolutely absurd proposition." There are too many variables, even under Barra's criteria. For example, the same W-L record doesn't necessarily equate; Team A could have lost to different (weaker or stronger) teams than Team B. This will affect poll average, quality win points, the computer rankings, etc. such that Teams A and B will not be tied in the BCS standings but for margin of victory.
Or, let's say they're both undefeated. Suppose Team A is plays Team C in week 8; Team C is ranked 10th in both polls. Team A wins, and gets credit in the polls for beating the 10th ranked team. Team B plays Team C the following week; Team C is ranked 22nd in both polls. Team B gets credit in the polls for beating only the 22nd ranked team. So Team A will be ranked higher than Team B going into week 10. The difference in poll rankings dictates that Teams A and B will not be tied in the BCS. So to say that the failure to consider margin of victory leads to ridiculous results is fallacious. Weighing margin of victory so heavy that it is enough to overcome two losses, which is what Barra is advocating, is more likely to lead to ridiculous results.
This season, the Trojans have come close to getting shafted again. USC isn't as good this year as last—no one is disputing that undefeated Oklahoma deserves to be ranked No. 1 going into the bowl games—but they are still very, very good, and by nearly any solid power rankings I've seen (including George Ignatin and Mad Max), they are clearly the No. 2 team in the nation. In fact, they were clearly No. 2 before Saturday's games, though the defending champion Ohio State Buckeyes had snuck into the spot in the BCS. Ohio State earned the ranking by squeaking out a 16-13 win over conference opponent Purdue, the third game this season in which the Buckeyes had failed to score a touchdown on offense. Southern Cal was punished for beating conference opponent Arizona 45-0.
Nitpick: "snuck" is not a word. The past tense of "sneak" is "sneaked." The managing partner at my former firm would have crossed this out and written "high school English" in the margin. As for Barra's main point, Arizona is an absolutely terrible football team, while Purdue is a solid, talented, well-coached football team. USC should have beaten Arizona by 45 points. This proves nothing.
If the Buckeyes had defeated archrival Michigan last Saturday, they probably would have played Oklahoma for the national championship in the Sugar Bowl on January 4. They didn't win, of course: They got clobbered 35-21 by Michigan while Southern Cal thrashed UCLA 47-22. A victory that, as we go to press, looks to put Southern Cal back in the BCS championship picture.
So, Barra's saying that a 25 point victory over a dreadful UCLA team would have been more impressive than a 1 point win over a very good Michigan team (a trendy preseason pick to play for the BCS Championship). Sheer folly.
The question, though, is why an Oklahoma-Southern Cal title match was dependent on Saturday's games. There wasn't an odds-maker in the country who wouldn't have favored the Trojans over the Buckeyes if they had met head-to-head. For that matter, there wasn't an odds-maker in the country who didn't seem to think that the Wolverines wouldn't beat the Buckeyes by at least seven points, which begs the obvious question as to why the BCS, the New York Times, USA Today's Jeff Sagarin, and others whose rankings systems are included in the BCS ratings would go to such elaborate lengths to elevate Ohio State.
Barra's point seems to be that oddsmakers are the true experts on who should beat whom. Barra obviously doesn't understand gambling, or he would know that betting lines are set to make money for the bookies, and they move depending on the action. In other words, oddsmakers pick a starting point for the line, and the public's betting tendencies move the line. So, if Bally's has Michigan favored over Ohio State by 17 points on the Monday before the game, and there's a lot of folks betting on Ohio State plus the points, the line is going to move to, say, 13 points. The bookies recognize that they might be overestimating Michigan or underestimating OSU if enough people think OSU will cover the spread. So the bookies cover their asses by reducing the spread. The betting line does not quantify how much better one team is than another, as Barra would have us believe.
Another nitpick: Barra misuses the expression "beg the question." To "beg the question" is not to raise a question that follows from a declarative statement or proposition, as Barra uses it in this paragraph. "Begging the question" is to paraphrase that which must be proved as a declarative sentence or to argue in a circular manner. An excellent explanation can be found here.
The answer is that the Times and Sagarin and the others did not actually create those rankings systems; they're following the guidelines given to them by the BCS through the NCAA, and those guidelines are riddled with compromises and logical absurdities. For instance, try and figure the BCS's formula for "Quality Win Component": "The quality win component will reward to varying degrees teams that defeat opponents ranked among the Top 10 in the weekly standings. The bonus point scale will range from a high of 1.0 points for a win over the top-ranked team to a low of 0.1 for a victory over the 10th-ranked BCS team. The final BCS standings will determine final quality win points. Quality win points are based on the standings determined by the subtotal. The final standings are reconfigured to reflect the quality win point deduction."
Got that? Personally, I think any team with a player who can figure out what that means deserves the national championship right there.
Actually, this formula is quite simple: suppose Team A beats Teams B, C and D during the season. In the final BCS standings, Team B is ranked 3rd, C is ranked 7th, and D is ranked 10th before they get their quality win points. Team A gets 1.3 as its Quality Win Component (.8 for beating Team B, .4 for beating Team C, and .1 for beating Team D). Barra is merely taking a cheap jab at the BCS formula to make his position seem more credible.
Two years ago, the BCS eliminated margin of victory as a component in its rankings. The reason for their concern was obvious: They wanted to prevent powerful teams from running up the score on weaklings just to increase their standing in the BCS rankings. This was an admirable sentiment, but the BCS's solution to the problem was entirely unnecessary, as computer programmers long ago figured out a method for making sense of runaway scores. "It's called 'collapsing,' " says Ignatin. "If a team is favored to win by, say, 20 points, and they exceed expectations and win by, say, 30, they deserve some credit for that. But if they pile it on and run up the score by, say, 60 points, it's not a true reflection of their real strength, so programmers 'collapse' the core after it reaches about 10 points above the anticipated margin of victory. That way no team can get too much credit for one or two lop-sided wins." Southern Cal, though, got no credit for six consecutive wipeouts in which they scored 43 or more points.
I don't understand. Were the computer polls right or wrong? Barra argues that the BCS didn't need to eliminate margin of victory because the computer polls already solved the issue of encouraging teams to run up the score by collapsing. Then, Barra laments that USC got no credit six huge wins, wins that would have likely been "collapsed." Which is it?
Through no fault of their own, Southern Cal's schedule this year is weaker than last year's, weaker, in fact, than Ohio State's. But only a fool, an ideologue, or an NCAA executive thinks that how much you beat a team by doesn't tell you something.
It might tell you something, I suppose. Maybe that a team is deep, or that a team is well-prepared every week (e.g., Oklahoma this year), or that the team is coached by an egotistical prick (e.g., Steve Spurrier). But it doesn't necessarily tell you that a team is better than its W-L record indicates.