Predicting the US Election 2012

November 5, 2012
By

Here is my take on who will win the  US elections, and by how much. It is based on an  analysis of the available information on November 5, 2012, using weighting factors to assess the importance of each bit of information.

 

Winner: Barack Obama

National Vote: 50.4%

Electoral College: 293 seats

 

My prediction model, in essence, considers a number of factors, or criteria, which I believe, will affect the election results. Selection of these criteria is limited to elements that can easily be quantified – such as polling results. I have compiled these criteria from a variety of sources; with a wider net, individual biases will tend to even out. Some factors, such as the impact of Hurricane Sandy, cannot be easily quantified and these are excluded. The effects, would, however, be reflected in other criteria, such as polls, that can be quantified.

To each criterion, we have assigned a weighting factor, based on our best judgment about its relative importance. To keep the math simple, all the weighting factors add up to 1.00. Each numeric indicator (a percentage) is now multiplied by its weighting factor to produce another percentage for each candidate. These percentages are added up for each candidate, to produce an overall rating, as a percentage. The candidate with the highest overall rating is predicted to be the winner, at the time of the analysis. Of course, these numbers change every day and the prediction is only valid for the day it is made.

The analysis of the Electoral College votes follows the same logic. We have taken the data from several reliable sources and assigned weighting factors. We even added a prediction by the Republican Karl Rowe, obviously not a pollster. However, since we know he is a biased source, we assigned a low weighting factor to him.

Data from the Five thirty eight blog has been given a higher weighting factor, since this is arguably the most reliable source of data. Nate Silver, who runs this blog, has developed a very sophisticated model and he has an excellent track record, predicting the 2008USelections with a high degree of accuracy.

National Polls

 

Indicators Data Source Weighting Factor Obama Rating Romney Rating Obama Composite Index Romney Composite Index
National Polls Average RealClearPolitics 0.1 48.8% 48.1% 4.9% 4.8%
Five thirty eight 0.3 50.6% 48.5% 15.2% 14.6%
Politico 0.1 48.0% 48.0% 4.8% 4.8%
NBC/Wall Street 0.2 48.0% 47.0% 9.6% 9.4%
Public Policy Polling 0.2 50.0% 47.0% 10.0% 9.4%
Twindex twitter polling index 0.1 59.0% 53.0% 5.9% 5.3%
Total 1 50.4% 48.3%

 

Electoral College Votes

Indicators Data Source Weighting Factor Obama Vote Romney Vote
Electoral Vote RealClearPolitics 0.1 303.0 235.0
Five thirty eight 0.3 307.0 231.0
Electoral vote 0.1 294.0 244.0
University of Virginia 0.1 290.0 248.0
Economist/You gov 0.1 303.0 235.0
Reuters/Ipsos 0.2 294.0 244.0
Karl Rove 0.1 259.0 279.0
Total 1 292.9 245.1

 

4 Responses to Predicting the US Election 2012

  1. vaidy
    November 5, 2012 at 7:31 pm

    Interesting analysis and prediction. Hope you are right as that’s what is needed.

    I wonder how you decided on the weightages apart from your gut feel

    • Shavak
      November 6, 2012 at 2:50 pm

      Hi,

      Very interesting analysis. Now let’s hope this comes true. I would hate to see Romney in power. Shudder…

      Shavak

  2. Akshay
    November 7, 2012 at 1:02 am

    That’s an interesting analysis. It will be a tight race.

  3. Akshay
    November 8, 2012 at 1:18 pm

    Mausaji that was a very accurate prediction!!!!!

Leave a Reply


Hit Counter provided by Los Angeles Windows