Fri Nov 2, 2012, 09:52 PM
Shivering Jemmy (900 posts)
Swing State Win Probabilities
It's too late in the season to build a statistical model of the election, but I am interested in doing so. If anyone wants to put together a DU model for use in the next electoral cycle(s) please inbox me. It could be really fun.
But I did write a simple script to calculate probabilities of a win based on the last week of polls for arbitrary states. I assumed normal statistics and simply calculated 1 the probability Obama  Romney was less than zero. I call that the probability we are really ahead as of this moment. Simple really. Anyhow here are some probabilities we are currently leading in several swing states: Virginia 83 Ohio 99 Wisconsin 98 Florida 91 Colorado 26 Iowa 94 Interestingly, our probability of being ahead in Ohio is higher than anywhere else. This is in part a consequence of the high polling density that Ohio has experienced. on edit: sorry, I forget that "normal statistics" isnt' selfexplanatory to nonstatisticians. Basically I assume that the error in a measurement is distributed according to a bell curve. So mean values are likely to be "close" to the center of the distribution. Large excursions from this value are assumed to fall off, in their probability of occurrence, as the bell curve falls off. Using this curve, it's possible to calculate the probability that the "true mean difference" is actually on the other side of zero (that Romney is really ahead). Compare the valuefor Florida from a month ago, in the week after the first debate, to see how far we've come. Then there was only a 26% chance that we were in the lead.

4 replies, 1023 views
Always highlight: 10 newest replies  Replies posted after I mark a forum
Replies to this discussion thread
4 replies  Author  Time  Post 
Swing State Win Probabilities (Original post) 
Shivering Jemmy  Nov 2012  OP 
courseofhistory  Nov 2012  #1  
Shivering Jemmy  Nov 2012  #2  
stevend56  Nov 2012  #3  
Shivering Jemmy  Nov 2012  #4 
Response to Shivering Jemmy (Original post)
Fri Nov 2, 2012, 09:55 PM
courseofhistory (801 posts)
1. Interesting! Thanks and can
you explain a little how you arrive at these numbers?

Response to courseofhistory (Reply #1)
Fri Nov 2, 2012, 10:06 PM
Shivering Jemmy (900 posts)
2. I put a little more explanation into the body
basically I'm just using MATLAB's ttest function to generate a pvalue (probability value) assuming that the data are really random and dsitributed according to a bell curve. Long story short: random fluctuations cannot explain Obama's lead in Ohio.
I edited the post with a somewhat qualitative explanation. To win, we just have to hope the polls aren't systematically wrong. Hopefully evangelicals are answering their phones as much as anyone else. 
Response to Shivering Jemmy (Original post)
Fri Nov 2, 2012, 10:17 PM
stevend56 (36 posts)
3. Thanks!
Yea, it's been a few decades since I did any probability calculations. But, I could understand your post. We could be looking at another electoral college landslide.

Response to stevend56 (Reply #3)
Fri Nov 2, 2012, 10:26 PM
Shivering Jemmy (900 posts)
4. I'm suspicious of the Florida numbers
it's only just recently swung our way. Throw in polls from earlier, and it goes away, sad to say.
But the only thing keeping me up nights now is that our turnout machine gets cocky and doesn't perform. 