First off: a note on third parties. Polling, especially at this point, drastically overstates the strength of third party candidates. While this could change this year, it is unwise to bank on that prospect. Therefore, I am considering the odds of a third-party White House win to be as unlikely as Trump winning the District of Columbia. If this changes, I will begin to display the numbers accordingly. While the third parties influence the margin, they do not have a defined % chance of victory. I am considering them to be infinity small and therefore statistically negligible. I don't see the need t add a column for a candidate if that column will just contain zeros or "<1%"
There are three new items now added to the site. The first is a simple histogram. This displays the distribution of Electoral Votes based on a simulation of 50,000 elections. Second, there is a table displaying individual win probabilities for each state. These win probabilities for the specific state treat each state as its own individual election. For example, the probability in California does not care what the outcome in Alabama is. Therefore, running a simulation based only on these probabilities would yield some very unlikely scenarios. For example, Clinton winning Mississippi and Georgia while losing Florida, North Carolina, Ohio and Pennsylvania. The histogram on the other hand treats the simulations as national elections. Therefore, if a Clinton were to win Pennsylvania, Massachusetts, California, Illinois, Mississippi, Arizona, Virginia, North Carolina, Florida, Minnesota and Ohio: she wouldn't be losing New York. While there are absolutely upsets happening in these simulations, the quantity of upsets are reduced to a more reasonable level. Therefore, Trump winning Iowa doesn't mean he will win Minnesota, but it does make it much more likely that he carries Kansas and Utah.
Finally, a table to display all of this. This table consists of an overall win probability which is drawn from the same 50,000 simulations I ran for the histogram. I will note here that electoral ties are presumed to go to Donald Trump. The House would decide the winner in a 269-269 tie. Even if Democrats regained control of the house, they would be exceedingly unlikely to do so at a level that would allow them to vote in Clinton. An election in the House would allocate one vote for each state. Therefore, the District of Columbia does not get to vote. Also, states where congressional districts have been drawn favorably to one party or the other will skew that count away from the 2-party congressional vote state-wide. Examples would be Virginia and Ohio. Also, winning a seat in a red state does not help. You must hold a majority of seats in the state as the state's delegation determines how to vote.
The second section of the table is a vote margin. This is the margin I am projecting that the leading candidate is currently leading at the national level. The third section is where the electoral college would currently sit if all favored candidates won the states in which they are favored. This negates any form of a toss-up and forces states into someone's column.
There will be more data and graphics implemented as we move forward, but I will be rolling those out at a later date. Any feedback is welcome, however keep in mind that the numbers can be very volatile at this point and should be taken with a grain of salt until we start getting closer to the first presidential debate this fall. For example, Trump could clearly still win. However, Clinton is much more likely to win at this point based upon the data at hand.