Should the Best Team Win Each Year?

The Cubs won the 2016 World Series. Though that hopefully isn’t news to anyone, it is still interesting for a variety of reasons. Notably, it was the Cubs’ first World Championship since 1908. I have nothing new or interesting to add to the conversation about the Cubs’ accomplishment. The reason I want to talk about the Cubs now is because not only are they World Champions, they were also clearly the best team in the MLB this year.

Most fans recognize that those two statements are saying vastly different things. The Cubs won more games in 2016 than any other team, had the greatest run differential and had the highest team WAR total, so it is fairly safe to say that they were, in fact, the best team in 2016. But in 21 seasons from 1995-2015 (wild-card era) the team with the best regular-season record (or tied) has only won the World Series four times: the Red Sox in 2007 and 2013 and the Yankees in 1998 and 2009. That’s a 19% success rate. Also since 1995 only three teams that have led the major leagues in team WAR have won the World Series: again the 2007 Red Sox and 2009 Yankees, and also the 2010 Giants. That’s 14%. So that raises the question: is this a problem? Should the World Series champion more frequently be the best regular-season team? Should MLB change things to fix this problem?

The short answer is, no, MLB’s goal is to make money. Baseball is a business. This is why MLB implemented winner-takes-all wild-card games. This is why there is a postseason tournament instead of just playing 162 games and giving the World Series title to the team with the best regular-season record.

The postseason isn’t going away any time soon, but I wanted to explore what it would take to more consistently crown the best team in baseball. I wrote some Python code to simulate a very basic, very over-simplified 162-game season. To do this, I assign each team a random “true” skill level drawn from a normal distribution and normalized to a zero to one scale. Then for each day in the season, I assign each team a random opponent and use more random numbers to “play” the game. Obviously this ignores the structure of the league as well as some of the more important and interesting aspects of the game of baseball, but in the end, each team has a win-loss record that resembles one that would come from a real MLB season. And the benefit is that these seasons take a lot less time than the real ones.

I simulated 1,000 162-game MLB seasons and in 41.7% of these seasons, the team with the highest “true” skill level also had the best record. If my simple simulation is to be believed, this means that if you played the 2016 season out two more times, you’d only expect the Cubs to have the best regular-season record in one of those seasons, and it would be more likely that they have the best record in neither season than it would be for them to have the best record in both seasons. Even taking injuries and other extenuating circumstances into account, this 42% is a solid increase over the 19% we’re seeing under the current system. Can we do better though?

Well, if you increase the length of the season to 300 games, then you can get the probability of the best team having the best record to over 0.50. That’s a very short off-season, but theoretically it’s possible, I guess. Moving fully into the realm of the impossible, it would take more than 1,620 games to get this probability greater than 0.75.

You may now be asking (or you may not be), “what if we look at the top two teams?” If, at the end of a season, MLB took the two clubs with the best records and played a World Series, how often would this series contain the best team? Running the simulation again, I find that in nearly 62% of seasons, a World Series played by the teams with the two best regular-season records, regardless of league, would contain the most talented team. Assuming from there that the teams have a nearly even chance of winning the series, we get a probability of around 0.31 that the best team is crowned World Champion, which is worse than just crowning the regular-season champion.

Anyway, the point is, next time you hear your Bostonian friend criticizing a team like the Indians for getting lucky in the playoffs, remind him or her that baseball is doing its best to give its fans an entertaining product to watch, and that it’s impossible and/or boring to ensure that the best team wins every year. And I would say based on Game 7 of this year’s World Series, the current system does a pretty good job.

newest oldest most voted

This is a great article! It’s not necessarily “the best team doesn’t win a majority of the time because the playoffs are a terrible system.” Instead, it’s probably just as true to say “the best team doesn’t always win because: Baseball.” Wonderful.