I was intrigued after reading a post last month by Chris Anderson on his Soccer By the Numbers Blog. The post compares the competitiveness of different football leagues in Europe. You can find it here. Anderson talks about "uncertainty of the outcome" as a measure of parity. This makes sense, as leagues where the outcome is not a sure thing are more equal.
With an uncertainty of the outcome in mind, I took another approach to analyzing the parity in a league is by looking at the amount of champions. In the past 10 years, only 3 clubs have won the English Premier League. In baseball, 9 different teams have won the World Series. Of course, this has flaws and is not a complete look at the league. This does suggest that the outcome is not fixed for baseball though.
Does this mean that professional baseball has a more balanced league than the EPL? If you've read Moneyball by Michael Lewis (if you haven't you should) you would know that MLB is facing payroll disparities similar to the one's in the EPL. So why the large difference in the number of winners? The answer is the playoffs.
In baseball, the 6 division winners plus 2 wild cards make the playoffs. There is one best of 5 series, followed by 2 best of 7 series. This adds up to only 11 wins to take home the World Series. Most people say that the playoffs are different than the regular season. They say all previous records are thrown out the window and any team can beat any other team. While there may be some change in the way a team plays when it comes to the playoffs, there is a more important factor at work: a small sample size of games. With such a small sample, it is not uncommon for a less skilled team to simply get lucky and beat a better team. Assume a team has a 30% chance of beating another team in a playoff game. For a best of 5 series, that team has a 16.3% chance of winning. For a best of 7 series, the team has a 12.6% chance of winning. All in all, upsets are not uncommon in the MLB playoffs. These upsets are the force behind the multitude of World Series winning teams this decade.
In contrast, we can look at the EPL. The EPL has no playoff system, and the winner is determined by the most number of points after each team plays 38 games. Effectively, you can look at this as just being one long playoff. Here, the sample size is much bigger: 38 games. Historically, teams have to win above 25 games to win the league (with the exception of last season). If we look at an above average team, what is their chance of winning more than 25 games? Let's take Liverpool from last season. For simplicity's sake I will only look at wins for the analysis. This may hurt a team with a lot of draws, but it makes the analysis a lot simpler. Last season they finished 6th with 17 wins. This means they win about 45% of their games. I am also assuming that Liverpool's record is an accurate measure of their ability to win games. In other words, Liverpool really does have a 45% chance of winning a game. The probability of Liverpool winning more than 25 games last year, if they have a 42% chance of winning each game, is .3%. For a team that won 25 games, or 65% of their games (in the past 10 years it has been ManU, Chelsea or Arsenal), the chance of winning more than 25 games is 42% Because of the bigger sample size, upsets are much less likely in the EPL. Even with a good team like Liverpool (I don't think anyone would say Liverpool winning the league is an upset), the probability of it happening is very low.
Baseball's smaller sample size of games in the playoffs allows for upsets and gives the appearance of parity with numerous teams winning the World Series. The EPL's larger sample size and lack of playoffs vastly reduces the chance of an upset which leads to the same powerhouse teams winning over and over again. John Henry already has two championships this decade with the Red Sox. The way the leagues are set up, a third championship with Red Sox is more likely than his first with Liverpool.