Tuesday, April 5, 2011

The Omnipresent Tournament

This isn’t going to be a “Butler wasn’t very good” or “UConn isn’t really the best team” essay. You’ll find those articles elsewhere today, no doubt, with people arguing to various degrees why those teams did or didn’t deserve to play in the championship game. They are invested in that question much more than I am.

More interesting, at least to me, is why we—as sports fans—put so much emphasis on tournaments as a way of testing a team’s ability. Ordeal by bracket is fully engrained into the constitution of every major American sport, and most of the minor ones. Yet while we love tournaments, we change the details in every sport. The NBA and NHL let almost everyone in, then play four rounds of best-of-seven series. MLB lets fewer teams in, plays a best-of-five series to start, then two rounds of best-of-seven series. The NFL uses a single elimination tournament, but some teams have to win fewer games than others. College basketball uses a single elimination tournament where (almost) every team has to win six games without losing any; college baseball teams are allowed to lose one game in their tournament. Bowling has an interesting ladder structure where better seeded players get multiple byes (the five seed plays the four seed, the winner of which plays the three seed, the winner of which plays the two seed, etc.)

The official reason for this variation is that no sport is the same, and that tournaments have to be fitted to the peculiarities of the game. Football is too strenuous for series instead of single elimination games, for example. I’m sure that’s true, but the tournament choices made in each league are peculiar, to say the least. College basketball uses a single elimination tournament, while the NBA uses best-of-seven series. Hockey follows the same pattern. Baseball is a sport with lots of variation from game to game—the best teams win only about 60% of their games during the regular season—but the opening round is only a best-of-five series. There is far less variation in the NBA, but the first round is a best-of-seven series nonetheless. The NFL has 32 teams and uses a bracket of 12; Division I has 120 college football teams and uses a bracket of 2.

There’s a cynical answer, which is also true, that playoffs are just as much about generating revenue as they are about finding champions. But that can’t tell the whole story, because even with that economic reality lurking in the background, we as fans invest these tournaments with legitimacy. By and large, these tournaments would not exist if we didn’t confer legitimacy; if we all stopped watching football after the regular season, the playoffs would just be a drain on NFL coffers. (I also find the exceptions to the rule interesting; I think most people treat college basketball conference season results as more important, at least in the major conferences, than conference tournaments. Why?) Tournaments might just be money grabs, but only because we consumers are offering our money for them. Leagues are giving people what they want to watch.

So why do we want tournaments? I have a few ideas, though I am probably missing others:
1.) Tournaments provide certainty as to which games are important. No league is headquartered in Lake Wobegon; some teams each season are below average. A few are awful. But we might not know which teams won’t matter, especially early in the season. We also don’t know which games will “matter” in the sense of telling us which team is truly the best (will this October Jacksonville Jaguars vs. Kansas City Chiefs game tell me anything about the best team in the NFL?) The tournament is a powerful signal: watch these games, because the league has guaranteed quality by filtering bad teams out. 

2.) Tournaments are shorter than seasons. Diehard fans will watch all season long, but lots of people aren’t interested enough to pay attention to the NBA from October to June, or MLB from April to November. Tournaments provide a condensed season within a season; a two month NBA season, a one month MLB season, a three week college basketball season. Beginnings of tournaments also provide convenient jumping-on points for casual fans that may never get involved if there was no playoff.

3.) Tournaments have one winner. Regular seasons can have winners, too, but there’s usually some room for interpretation. Who was the winner of the college basketball regular season this year? Ohio State was the #1 overall seed in the tournament, so they are a reasonable answer, but Kansas and Pittsburgh (just to name two) were plausible answers also. The question in this context also seems peculiar—how can a 30 game season have a winner when there are 340 teams and the vast majority will never play each other? In lots of sports, there just can’t be an answer; the regular season isn’t long enough (perhaps can’t be long enough) to give us enough data.

I think reasons 1 and 2 are valid, but number 3 is probably what really drives our love of the tournament. UConn won; no one else did. No interpretation is necessary. We have an answer, and we can look at Wikipedia (and maybe the NCAA record book) twenty years from now and find out who won the 2011 men’s basketball season.

Sports are as much about plot as they are about truth. You might make a movie with an indeterminate ending, but I doubt it would be popular (look how people wigged out at the blackout ending of The Sopranos—that’s how lots of people would feel if we didn’t have a winner at the end of seasons). Fans want sports to have a point, so the tournament becomes the point. The season had an end, and at the end, a good team won a set of games all good teams were trying to win. They were the champion. They overcame adversity and long odds and won when winning was important. The End.

Obviously, I’m skeptical, or at least won’t lose much sleep if the “who was the best team?” question has no clear answer. But what if investment in the useful myth of the infallibility of brackets what keeps leagues going? The English Premier League is plenty popular and uses only a regular season, but maybe all the above ruminations are culture specific. Maybe American sports leagues are in such competition with one another that a tournament-less sport is at a marked disadvantage against others (perhaps this explains the decline of boxing and horse racing through the years, though even horse racing has tried to implement a quasi-championship in the Breeders’ Cup). Paris was worth a mass; college basketball is worth a tournament.

Bonus Linkage! Brian at Mgoblog claims the NCAA tournament still works just fine, since the matchup is usually at a plausible pairing of the two best teams in the nation. I think this is a funny definition of working, where the tournament is useful only insofar as it confirms what we know from the regular season. It's not an additional justification for the tournament, besides from "it is fun." This may be enough. 

He also uses Kenpom a little too liberally here, since 1.) Pomeroy has changed his formula around to backfit tournament results in some cases (nothing wrong with that necessarily, if the formula becomes more accurate), and 2.) those teams rise in the rankings as they win their tournament games. The rankings of those teams after the season is not their rankings before the season. To the extent that tournament performance is just one more piece of evidence concerning a team's quality, that's fine. To the extent that it's being used to argue that the tournament confirmed what the numbers said all along, it's not fine.