By this point, I think it’s pretty clear that I, like most people, don’t like the Bowl Championship Series. I stated most of my case last season, when I suggested a 12-team playoff replace it. I’m more onboard with a 16-team variation now than I was then, but that’s not necessarily original and doesn’t require a whole lot of explanation.
But when someone, in this very space, deigns to point out any good things about the BCS, well, then it’s my turn to chime in and refudiate those notions.*
*What’s that? Sarah Palin jokes are also unoriginal and gratuitous? My bad. Change that to “repudiate” then.
In his mildly pro-BCS screed, which is really just a “Hey, it’s not pure evil” argument about the system, John writes, “The BCS is no more unfair than, say, the absence of a salary cap in baseball, or the fact that there will be a playoff team from the NFC West this season, but it draws exponentially more ire than either of these injustices.”
But John, comparing the systemic unfairness of the BCS to the aberrational one of the NFC West is disingenuous. Some years—in fact, most years—the NFC West produces a competent playoff team. In the last decade, three teams from the NFC West have reached the Super Bowl, tied with the NFC South for the most of any division in football (and the fourth, the 49ers, only won 4 titles in a 13-year span from 1982 to 1994). This is very different from the intrinsic unfairness built into the BCS and MLB’s salary structure, where imbalances NEVER cycle.
John also attacks my running series, “Every Game Counts”: “But this is because Tim is using ‘count’ to mean a very specific thing, namely to have a direct impact on the national championship. Of course, in this respect, it is impossible for every game to ‘count’ in any system—every league with a playoff has games between teams that have been eliminated, both officially and unofficially, from playoff contention.”
The key difference between my point about college football games that don’t count and your point about professional football games that don’t count is that I am approaching the season a priori. I can even quote you quoting me: “[C]ertain games do not–indeed, cannot–count. Many teams cannot control their own destiny. These are games that, regardless of what happens before or after them, will have no bearing on the national championship picture.” The important parts are the “cannot count” and the “regardless of what happens before or after” (my emphases each), whereas in the NFL, no one could say with certainty that this weekend’s Broncos-Cardinals game would not count. In every other sport on nearly every level, at the start of the season, every game can conceivably mean something. This is not true in college football.
More on the issue: “When you introduce a playoff, you are effectively saying that some games count more than others: Do-or-die elimination games count the most, followed by non-elimination playoff games (in leagues with playoff series), followed by divisional or league games that determine playoff positioning, followed by regular season games between non-playoff teams. This is why teams do things like rest players in Week 16 instead of trying to go undefeated, and why great players take games off, and why regular season records don’t tell the whole story in these leagues.”
I’m not quite sure what you’re using “count” to mean here. If you mean some games in major professional sports matter more than others, well, that’s obviously true. Jets-Patriots last week mattered more than Broncos-Cardinals this week. But that’s also the case in college football, where Oregon-Stanford mattered more than TCU-San Diego State.
But if your point is that Oregon-Portland State counts just the same as Oregon-Stanford, because one loss—any loss—can derail the Ducks’ title hopes, that’s also the case in professional football, where something like the Texans’ Week 1 loss to the Jets last season can cost Houston a playoff berth.
Let me reiterate a point I made last season: College football’s regular season is the most exciting largely because it has the fewest games. If the NFL played a 12-game schedule, almost nothing would ever be clinched before the last week of the season, either. And if college football played a 16-game schedule, there’d be a much better chance of the Cam Newtons of the world taking more second halves off down the stretch.
Also, please do not trump up college football’s regular season by comparing it to the NBA’s, which is nothing more than an overhyped preseason. (Here’s a fun experiment: Take a snapshot of today’s NBA standings, and see how different they are at the end of the season. I’ll set the O/U on “Teams Who Will Eventually Make the Playoffs Who Wouldn’t Qualify Today” at 2, and challenge you to take the over.*)
*I’ve got Portland and…that’s it.
And I can honestly say I have no idea what you mean when you imply that regular-season records DO tell the whole story in college football. I hope hope hope you find that concept as laughable as I and the 2008 Utah Utes do.
On underdogs: “In college football, though, an underdog doesn’t just need to pull off one upset, or even a series of upsets—an underdog in college football needs to prove it is one of the best two teams in the country. This can be an especially difficult task since teams like Boise State play a demonstrably weaker schedule than teams like Alabama.”
I only suggest one change here: “[E]specially difficult” should be replaced with “impossible.” It is impossible for a team such as Boise State to “prove” it is one of the nation’s two best teams during a single regular season. If TCU, for instance, won each one of its games this season by 50 or more points, it would likely still not be playing in the title game. And think of how far the Broncos and Horned Frogs have traveled to reach this peripheral point: BSU has been a dominant program for a decade. The Broncos have been every bit as good as Gonzaga has been in college basketball, but this is the first year they’ve had a chance at the title, even though it’s been far from their best team. TCU, meanwhile, has flourished under Gary Patterson, finishing in the top 10 in four of the past five seasons. Think of it this way: If you’re a freshman enrolling at a non-AQ school that hasn’t made national noise yet, there is no chance you can play for a championship in your four years.
As an aside, I would also argue that it’s “especially difficult” for any team, regardless of context to “prove” it’s one of the two best in the country during a 12- or 13-game season. Right now, we can say Auburn deserves to play for the championship more than TCU based on its schedule; but can we definitively say that the Tigers are better than the Horned Frogs? It was three weeks ago when Vegas had TCU as the favorite in a potential meeting with Auburn.
And I also disagree with your main point that the BCS is considered worse than other methods simply because it is different: “Instead of introducing a hierarchy of games and gradually weeding teams out, the BCS evaluates the entire regular season to determine the two best teams to battle it out for the National Championship. It may get those two wrong, but that’s merely a flaw in the execution, not the principle behind the system.”
The flaw is that the principle behind the system rarely allows it to be properly executed. How many times has the BCS given us, definitively, the nation’s two best teams? Three times! In 13 years!* It doesn’t matter what the methodology is; if it’s used to select only two teams to play in the championship based on a 12-game and totally unbalanced schedule, it will fail more than it succeeds. There simply aren’t enough games to make an informed decision. What made 2008 Oklahoma more qualified than 2008 Texas? Or those Longhorns more qualified than 2008 Texas Tech? It was impossible to create clear distinctions between those three teams.**
*FSU-Virginia Tech in ’99, Miami-Ohio State in ’02, USC-Texas in ’05. That’s it.
**Sure, you can say “Well, the Sooners demolished Texas Tech, so the Red Raiders are out.” But that would be like saying the Jets should be barred from this year’s playoffs because they lost to the Patriots by 42 on Monday night.
And that was as easy as threesomes get in the BCS. Think: Those teams played in the same conference! They had games against one another! There were bases of comparison! It wasn’t as easy in 2004, when Oklahoma, USC, and Auburn hadn’t played each other. How could you figure out a second team in 2007, when Hawaii was undefeated, Ohio State and Kansas each had one loss, and LSU, Oklahoma, Georgia, Virginia Tech, USC, Missouri, West Virginia, and Arizona State were all BCS conference teams with two losses?*
*I bet you, like me, forgot Kansas only had one loss that season.
This gets to my main point, which is that a BCS-like system can work in some sports. A BCS system could work well in baseball, where every team plays every other team in its league several times. It could work in basketball, as well; it might even spice up the regular season (is Pierre listening?).
But college football is the absolute worst sport to use such a system. Why? Nobody plays anybody! Non-conference tilts between high-ranking BCS-conference teams are an endangered species. I’m not entirely positive about this, but my count of non-conference games played between BCS teams that finished in the Top 25 is…two. LSU-West Virginia and Arkansas-Texas A&M. I forgive you for not staying up for those showdowns.
And so there are no meaningful bases of comparison. Our conceptions of conference strength are at best estimates and at worst assumptions. It’s easy to say that Auburn has played a harder schedule this year than TCU. But how much harder? Would a 12-1 Auburn team be more impressive than 12-0 TCU? What if Auburn lost two games? How about Oregon vs. Ohio State? What if the Ducks dropped one game? Would their schedule within the Pac-10 be better than OSU’s in the Big Ten? What about Wisconsin’s Big Ten schedule, which was different even than Ohio State’s? What if Oklahoma went 13-0 in the Big 12? How would that have compared to Auburn’s 13-0 and Oregon’s 12-0? And how would the comparison this year differ from the one made in 2004, when Oklahoma was 13-0, Auburn was 13-0, and Pac-10 USC was 12-0?*
*Answer: They’d be ranked Oklahoma, Auburn, Oregon in that order. Why? Because Oklahoma is Oklahoma and the SEC is the SEC. Oregon doesn’t have the tradition USC did in ’04.
The problem is that in the large majority of its operative years, the BCS cannot do what it purports to do, which is to give us a legitimate championship game between the nation’s top two teams. It’s not that it gets it wrong those years; it’s that there’s no way for it to get it right. In 2008, BCS proponents cannot legitimately justify why Oklahoma deserved to be in that title game over Texas, just like they wouldn’t have been able to legitimately justify the vice versa. The same could be said about ’07 (LSU over a host of teams), ’06 (Florida over Michigan, Louisville, Wisconsin), ’04 (USC and OU over Auburn), ’03 (OU and LSU over USC), ’01 (Nebraska over CU and Oregon), ’00 (FSU over Miami and Washington), and ’98 (FSU over K-State, Ohio State, and UCLA). Bill Hancock et. al—and perhaps John—may say I just don’t get it because I’m still viewing things through a playoff prism, but if TCU wins the Rose Bowl this season, it will be the third consecutive a year a team finishes the season undefeated yet without a title. And that can’t be satisfying to anyone.
If you design a system that doesn’t properly execute its function more than half the time, then the system is a failure. And even Richard Billingsley’s computer rankings know that’s not a good thing.