The Playoffs Are Not a Crapshoot

 "My job is to get us to the playoffs. Everything after that is f-ing luck." - Billy Beane, quoted in Moneyball.

Imagine 2 teams with slightly different strategies in terms of applying their superior financial resources; in other words, imagine 2 teams who have exactly the financial resources of the Yankees. One team's strategy is to use all its resources to win 110 games a year; let's call this team the Bosses. The second team's strategy is to hold back a bit and try to win 95 games a year, enough to make the playoffs; let's call them the Crapshoots. In this thought experiment, each season both teams win their target number of games, and meet each year in either the League Division or League Championship Series. The Bosses are always the top seed.



The numbers of 110 and 95 wins were not pulled from my ass, but rather come from the following passage of a Bob Klapisch article that articulated an oft-repeated theme, which is that once you get into the playoffs whether you win is pretty much a matter of luck:

"If nothing else, the Yankees figure they’re good enough to win 95 games right now, even with A.J.Burnett and Phil Hughes in the same rotation. That puts them in the playoffs. Once there, remember what Billy Beane once said about October’s crapshoot....There’s no need - no point, actually - in building a 110-win roster....'The idea now, for all of us, is just get there,' said a rival GM."

This attitude is absurd. When the Bosses and Crapshoots meet in the Division Series, each team has approximately the following odds of winning the series:

• Bosses: 69.5%
• Crapshoots: 30.5%

Note: Series odds in this post were calculated using the teams' winning percentages and Bill James' log5 method. If you are not familiar with log5, chancesis.com provides a good introduction.

If they were to meet in the League Championship Series, the odds shift slightly in the Bosses' favor:

• Bosses: 71.9%
• Crapshoots: 28.1%

In other words, the Bosses are over twice as likely as the Crapshoots to win a postseason matchup, suggesting talent is a factor, not mere luck.


Now, a reasonable question is, do these log5-calculated odds reflect reality? We cannot test these precise numbers - there has never been one postseason series between teams with 110-52 and 95-67 records, and we would need numerous such series to form a sample size. However, to get a reasonably large enough sample size to test this out, I looked for all postseason series in which the difference between the winning percentages of the favorite and underdog are in the range of .07  to .115; this should get all series where the favorite-underdog matchup is similar to that of Bosses vs. Crapshoots.

Note: The odds in each of these series would have been slightly different based upon the exact differences in team wins, whether it is a 5- or 7-game series, and the home-away format of the series; but for what we are estimating these differences are insignificant.

In MLB history there have been 40 such series. This is not a huge sample size, but it is large enough that if the log5 odds are correct and if more than mere luck is involved, we could expect the better teams to have won in the range of 60-80% of the series.


Which is what happened: the favorites won 29 of the series, 72.5% of them. Winning in the postseason is not entirely due to luck; talent clearly is a huge factor.

So as a fan, given these numbers and facts, would you rather have the Yankees continue to run their business a la the Bosses? Or would you like them to become the Crapshoots?