Hey, I have an idea. My intuition tells me that there should be an easy way to mathematically show that the GOP calling the election with only 84% of the votes in, with the numbers being so close, is not a sound argument. EVEN assuming those numbers are correct. I need someone who is better at statistics than I am. Any volunteers? At the very least, this is a good exercise for a stat student:
Here's what someone can do:
1. Take the published results
http://www.mainegop.com/wp-content/u...us_results.pdf
2. Treat each precint as a sample of the total true population.
3. Convert votes to % in each precinct (Where there are votes)
4. Find the mean, standard deviation, etc for each candidate so that we have something like Romney 39% +/- 5%, Paul 36 +/- 5%
5. Use a t-test or something to test for statistical difference between Romney and Paul
6. Show how if 16% of the votes still need to be cast (the number being thrown around earlier), which is a lot, that variances in that population could have Paul going as the winner, maybe 10,20,30,40% of the time.
Any takers?