Monday, November 5, 2012

Capturing Predictions for Monday Morning Quarterbacking on Wednesday


I'm bookmarking Electoral College predictions from various media outlets.  Always fun to compare after the election.

By the way, this is not a political posting, but rather one discussing math, more than politics.

One interesting site is the New York Times 512 Paths to the White House.  It looks at the various swing states, and what remaining paths there are for a candidate when certain ones go into their column, or their opponents.  I'm not going to be able to stay up tomorrow night, but if I could, this would be a fascinating site to use.

Given that each candidate has states that almost 100% will win, with only nine states that are toss ups, the numbers say there are 512 combinations of winning or losing those nine states.

Obama has 431 paths to victory, while Romney has 76.  There are five paths leading to a tie.

A lot of people have demonized Nate Silver of fivethirtyeight.com for his polling aggregation and statistical analysis, it is interesting to note that 431 is 84% of 512.  That is very close to the 86% chance of winning Silver says Obama has of winning.

Of course, that still leaves Romney a 16% chance of winning.  I wouldn't make a sizable bet of any kind with a 16% chance of loss.  A Romney win would not vindicate Silver's critics.  For Romney, its like rolling a die and winning if a five comes up, but losing if any other number came up.  It can happen.  So Obama supporters shouldn't think it is in the bag, nor should Romney supporters think all is lost.  Vote!

No comments:

Post a Comment