2012 Presidential Election

Category: Politics

Published: 11/06/2012 02:47 a.m.

Since the 2008 election, I've been a follower of the political blog FiveThirtyEight. The author, Nate Silver, has a methodology for interpreting poll data and was very accurate in his 2008 predictions. I like stats, and this site gained my interest.

In the 2012 election, there have been other polling sites that, with similar poll data, have come to strikingly different conclusions. Below are two examples of predictions, followed by my own that I made at 270towin.

UnSkewed Polls

I'm not sure how I heard of UnSkewed Polls, but I'm glad I found them. On top of adjusting most polls to unskew the liberal bias (with the exception of QStarNews, who runs UnSkewed Polls), they made a bold prediction:

Romney 359 - Obama 179


The predictions at FiveThirtyEight have changed throughout the course of the campaigns. At one time, President Obama had a commanding lead, only to see that fall to almost a tie after the first debate. Since then, his lead has increased to the results we have on Election Day.

Obama 332 - Romney 206

Incredibly, this is a 153 electoral vote swing from the Unskewed Polls prediction, including the following states:

  • Florida - 29
  • Pennsylvania - 20
  • Ohio - 18
  • Michigan - 16
  • Virginia - 13
  • Wisconsin - 10
  • Minnesota - 10
  • Colorado - 9
  • Oregon - 7
  • Nevada - 6
  • Iowa - 6
  • New Mexico - 5
  • New Hampshire - 4
If these guys are looking at the same polling data, someone is going to be way off.

My Predictions

My map is almost the same as the FiveThirtyEight, except I have Florida going red.

Obama 303 - Romney 235

We will see soon!


Looks like Nate Silver was right on and Unskewed Polls does some explaining.