27 October 2015

Royals or Mets? Flip a Coin. No, Really: Go Ahead and Flip a Coin.

This post is part of a series about the 2015 MLB Postseason. Follow along here.


Want to simulate the outcome of the 2015 World Series? Flip a coin. No, seriously: find a coin and flip it.

My simulator—based on Matt Swartz's home field advantage calculations, the Log 5 equation and my own RPScore true record estimates—has the Kansas City Royals winning 51% of the time. Thanks to the work of Persi Diaconis, Professor of Mathematics and Statistics at Stanford University, we know that if we vigorously flip a coin, there is a 51% chance that the coin will land on the same side it started out on,* conveniently equivalent to the Royals' chances of winning it all.

*h/t Smithsonian Magazine

So, go ahead: find a coin and note whether it's heads up or tails up. Give the coin a good flip. Did it land the same way it began? If so, you just picked the Royals. If not, you picked the New York Mets to go the distance. There's your prediction.



From the chart above we can see that the Mets and Royals have experienced relatively similar postseason journeys. Both teams went five games in their respective Division Series. The big difference of course is how the LCS played out. The Royals never surrendered the probabilistic lead in the ALCS after their game one win against the Toronto Blue Jays, but their road to the pennant wasn't a smooth one, the Jays holding on in game five and close to forcing a game seven.


While the Mets experienced as much of a seesaw in the NLDS (at least on a game-by-game basis) as the Royals did in the ALDS, it was comparatively smooth sailing for New York in their pursuit of the pennant. In fact, considering the Cubs never led in the NLCS, the Mets' line underestimates their dominance (though if it took into account inning-by-inning chances, the line itself would look more like a series of steps in New York's favor).


The reason for the Kansas City's slim, statistically insignificant edge is entirely home field advantage. RPScore considers the Mets a slightly better team than the Royals. Accordingly, had the Senior Circuit bested the Junior Circuit in this past summer's All-Star Game, the Mets would win the simulation about 52% of the time. Put another way, despite all the hand wringing about the Midsummer Classic and whether or not it should or does "count," the effect of the result is only +/- 3% of a World Series victory.

As it stands today, the modal outcome is a Royals win in seven games, whereas the Mets' peak is a win in six. There is a nearly 2:1 chance that the series goes the full seven and only a 12% chance that one team sweeps the other.

Of course, the vertical scale of the graph exaggerates the differences between a Mets win in five, six or seven and a Royals win in six or seven. What your takeaway should be is that it's more likely to be a long, close series than it is that either team will dominate the other. Regardless of who wins, that's a win for everyone (well, except the loser).


So go ahead and flip that coin. Then, tune in tomorrow to see how the outcome of game one of the 2015 World Series affects the odds for both teams going forward.

PS: the penny I flipped started out heads and landed tails. It could be the Mets' year after all.

No comments: