Wednesday, October 07, 2009

Playoff Musings

For the sake of discussion, let's assume that we can estimate a team's true W% by taking 40% of their actual W%, 20% of their EW%, 20% of their PW%, and 20% of .500. Let's also assume that there is no quality difference between the NL and AL, no home field advantage, and that all games are independent of one another with the probability of a given outcome constant across games. This will allow us to estimate the quality of the playoff teams as follows:

We can couple this with the binomial distribution (since we assumed independence a constant W%) to figure the probability of each team winning the Division Series, and then the LCS and World Series (P(DS) is the probability that a team will win the Division Series, and so on):

Why am I doing this? It's obvious that while these estimates might be reasonable, they are could be improved fairly easily. We know there's a home field advantage, we could incorporate the actual pitching matchups, we could come up with a better combination of the various W%s--or better yet, we could look at projections for the team's actual personnel rather than using aggregate season W%s. So why bother, especially when you can find playoff odds reports elsewhere on the net that do take some of those factors into account?

As I said, they're a reasonable starting point without getting more involved. For the rest of the piece I will treat them as more than that, for the sake of discussion. More importantly, though, they illustrate what sabermetricians usually mean if they say something like "the playoffs are a crapshoot". The most lopsided first round matchup still yields a one-in-three chance for the underdog, and no team has a greater than 40% or less than 10% chance to win the pennant.

In fact, the source of a lot of the differences is the uneven first round matchups between NYA/MIN and LA/STL. Here are the probabilities of each team winning the World Series given that they make it out of the first round:

Here the probabilities range only from 16 to 36%. The Cardinals move ahead of the Phillies and Rockies; their first round matchup with the NL's top team drags down their chances, but if they get past LA, they are the strongest remaining team in the senior circuit.

Suppose that for some reason the Phillies' place in the playoffs (including seeding--obviously they would actually play the Dodgers and not the Rockies, but that's besides the point) was taken by the Nationals. What would the probabilities look like in that case? Washington was last in the majors with a crude strength estimate of .408--plugging that in produces these results:

Even the Nationals have a 1:4 chance of advancing to the LCS, benefiting from the five-game series. It gets tougher in the two seven game series, but they again have a roughly 1:5 chance of winning that series should they get to either one. Even looking at the playoffs as a whole, Washington has a 1% chance to win the World Series given these assumptions.

1 in 100 may not sound like a lot, but considering that they were the worst team in MLB, it's not that bad. Do you think the average mainstream media member would give them a 1% chance in that scenario? Do you think they'd say they had a 25% chance to beat Colorado in a division series?

I'd have to guess that, no, they wouldn't. I could be wrong, but it seems as if the public in general is far too confident in their ability to project the results of the postseason. If you can't state with more than 80% certainty that the Nationals wouldn't advance past the Dodgers, then what can you say with that kind of confidence?

This is all just a long way of saying that I avoid making predictions about the playoffs. I don't think that I'm smart enough to tell you anything with a high enough certainty level to even make it worthwhile. So instead I offer anti-analysis, something completely personal and not entirely rational--my rooting interests.

I am definitely rooting for the Yankees to win the World Series. First, they have Nick Swisher, who is my favorite player in the game. Second, I actually like George Steinbrenner, and I would love to see him win another championship. Third, I would love to see A-Rod silence the critics that have attempted to brand him as a choker (although even a superlative performance throughout the playoffs by A-Rod would not make that meme disappear completely). Fourth, when the Yankees win there is always a hue and cry from the crowd that constantly wrings its hands about competitive balance, and I really don't agree with their position at all and am amused by their lamentations. Cold and petty? You betcha.

The Red Sox also offer the competitive balance angle, as well as a front office that is easy to root for. I've never been an Angel fan in the past, but Bobby Abreu has always been one of my favorite players and Chone Figgins leading the league in walks doesn't hurt. I also have to admit to being a little partisan towards the AL due to the DH issue. Albert Pujols alone is enough to make the Cardinals likeable, and I have no real problems with the other NL clubs. So:

1. Yankees
2. Cardinals
3. Red Sox
4. Angels
5. Rockies
6. Phillies
7. Dodgers
8. Twins

What I'm really rooting for though is some competitive series and compelling games (although I don't know how likely it is that the drama of the AL Central playoff will be topped). There hasn't been a six-game World Series since 2003, so that would be a great place to start.


  1. "Second, I actually like George Steinbrenner, and I would love to see him win another championship."

    Geogrge does have 6 already. A lot of NBA fans were rooting for the Lakers to win the championship last year so Jackson would get his 10th ring as a coach. But Jackson already had 10 rings if you count his one as a player with the Knicks.

    Steinbrenner probably has the most interesting past of any owner or former owner in professional sports. He attended military school, played college football, joined the Air Force, worked for Woody Hayes at OSU. He owned the Cleveland Pipers of the short-lived American Basketball League in the early 60's. He also invested in some Broadway shows. He did all of this before he purchased the Yankees. Did I miss anyting?

  2. He also was an assistant football coach at Northwestern or Purdue (maybe both). I believe in those roles he was actually a coach, whereas at OSU he was just a graduate assistant-type, although I might be off on the details.

  3. I dug out my copy of Steinbrenner! by Dick Shaap. Yes, he was a graduate assistant to Woody Hayes. And he was an assistant football coach at both Northwestern and Perdue. You have a good memory.


Comments are moderated, so there will be a lag between your post and it actually appearing. I reserve the right to reject any comment for any reason.