Why should I care about your statistics?
The short answer is that you shouldn’t. The longer answer is that I try to do my analytics in a way that are easy to understand, that show my reasoning, and can be replicated by anyone with an interest in doing the work for themselves.
Where do you get your data?
I am currently using historical data from Electoral-vote.com and current polling data from Huffington Post Pollster.
Shouldn’t I get an analysis directly from the source?
By all means. Each of the following places has their strengths and weaknesses with respect to data and political analysis. That being said, regardless of their personal ideology, they each do a tremendous job at statistical analysis.
- Princeton Election Consortium
- Huffing post pollster
- Five thirty-eight
- Real Clear Politics
- Electoral vote
So what do you do differently from them?
I take the median of current polls to use as a base. This idea was pioneered (to the best of my knowledge) by Princeton Election Consortium. I then use each state’s history so that I can apply a Bayesian analysis to the results to try to get a better picture of how each state will ultimately vote. Finally, I employ a computational analysis called the Monte Carlo method to compute the likelihood of each candidate winning in November.
Do you do all this by hand?
Yes. I download the updated data, sort it, and apply all the mathematics (with the exception of the Monte Carlo computations) by hand.
Holy Shit! That’s a lot of work.
Not a question, but yes it is. That is the reason that I don’t update my results nearly as often as the more popular sites.
Are you more accurate?
Sometimes, but I’ve made no great leap in political predictions.
Each respected statistician comes infinitely closer to the ultimate election outcome than the political pundits we will be forced to listen to for the next three months.
Then why do it yourself?
I studied math and physics in college: I like it. More specifically, it is a personal competition to see if I can get a hair’s breadth closer to the actual election outcome than the other statisticians.
I understand. Can I safely ignore this political analysis crap?
Yes, but please don’t ignore politics. It is important that we elect a person that will help the United States be the kind of place where we can all live happily, thrive, and reach the pinnacle of our humanity instead of allowing it to turn into a steaming cesspool of authoritarianism with no other purpose than to grind people to dust for profit.