Post

Jupyter notebook illustrate how to simulate elections based on polling data

http://nbviewer.jupyter.org/github/ahoaglandnu/election/blob/master/Assigning%20State%20Probabilities%20and%202016%20Election%20Forecast%20Simulations.ipynb

Reading, it occurs to me that elections have some lotto-like qualities-if the variables are random, every 10,000 election we should see something as unlikely as winning pick 4