No Panaceas

Tuesday, September 30, 2003
In search of the mythical championship: Back in March I wrote a couple of posts (here and here) about the way tournaments, such as college basketball's March Madness, don't crown a champion in any true sense. This is because of two problems: intransitivity and an insufficient number of games played between teams.

The baseball playoffs start today and so I thought I'd return to this subject as it relates to baseball. The bit about intransivity still applies and doesn't need repeating here. But what about the small-n problem? I think in baseball it's even worse. The reason is because baseball has an attribute that makes it notably different than virtually all other sports. In baseball it's the norm for a team's most important players to not play every day. Obvously I am talking about pitching. Imagine the impact on, say, the Lakers if they could only play Shaq every fifth game, their second best center every fifth game, their third best center every fifth game, etc.

That's what happens in baseball, at least during the regular season. Pedro Martinez or Mark Prior or whomever pitches about every fifth game. For those other games a team must field its second best, third best, and so on pitcher. Thus the probability of a team winning any particular game is dramatically dependent not just on the quality of the opposition but also on who is pitching. Barring injuries, this kind of variation does not occur in basketball, football, hockey, and soccer.

Yet, in the playoffs the dynamic changes. Let's say that Team A has two incredible aces and three stiffs for its regular season pitching. Let's say that Team B has a good staff but no aces. (We might think of Team A as the Cubs and Team B as the Braves, whatever). In a five game series Team A can play with a three man rotation, starting one of its aces twice. In a seven game series it probably can get by with a three man rotation, but probably go with a four, and it can start its aces at least twice each, maybe one of them three times. In the 2001 World Series the Diamondbacks used a four pitcher rotation but started Schilling three times and Johnson twice.

The point is that in a short series a team can ride its aces. So let's take the Braves and the Cubs. In a five game series the Cubs have a huge advantage because of Prior and Woods. But if the two teams played a longer series -- nine or eleven games, say -- the chances are very good, probable even, that the Braves would win. (I'm rooting for the Cubs, by the way.) My larger point is simply that the nature of baseball makes the small-n problem even worse. Five game series are silly and I suspect that even seven game series don't really help matters like they do in basketball.

This brings me to a final point. Lot's of people yip about baseball's wild card system. The yipping grew loudest last year when two wild card teams -- the Giants and Angels -- met in the World Series. First, in general I agree with complaints about wild cards. Personally I think that college basketball, pro basketball, hockey, and, to a degree, the NFL have ruined their regular seaon by letting everybody and their dog in the playoffs. My fondest hope is that the NBA championship is one day won by one 38-42 team beating another 38-42 team in the championship series. Now obviously what the MLB has done is no where near as bad as what the NBA has done. But the wild card does water down the regular season a bit and that's bad. Because the regular season is what should matter not this little thing they do at the end. Adding wild cards will help inferior teams win. But so did splitting the leagues into divisions back in 1969, thus creating the LCS. (How was the DH rule a worse idea than this?) Or, for that matter, creating the World Series back in 1903. At least the original world series was best of nine.