The New York Times created this web site UpShot. It is poll consolidator of poll consolidators. It takes data from Dr Sam Wang, Nate Silver, and some electronic prediction markets and prognosticating punditry and comes up with an estimate for “Probability of victory for Hillary or Trump”.
It is a very well done site, from technical standpoint. It presents data very nicely and it is hours of fun to play in this site. At the bottom they have created a tree of path to victory, based on the 10 closest races. As of now, based on the 10 closest races, Hillary has some 1000 ways to win and Trump has 4 ways to win. You can click and give the state to the blue team or red team and watch the paths to victory dynamically change. Hours of fun.
Yesterday, at the peak of the bounce if you give any one of the eight states in the closest 10, it was game over. Today any one of the six means game over for Trump.
Curiously, they have listed Colorado, 86-14 for HRC in the closest 10. But not Indiana 16-84, Miss 17-83 or Alaska 17-83. I was wondering why. If you give Colorado to team blue, there are no paths to win for Trump. Just one way to tie. And HRC has 511 ways to win.
Is it possible they are throwing in very deep blue states in the mix so that at least some paths to victory shows up for Trump? I mean, they spent all these effort to get a cool tree display and dynamic animation working. Seems like a waste if just says, “why bother? she got it in the bag?” May it is done this way to keep suckers like me to keep coming back giving them the clicks.
With all that coding power behind that site, it would be trivial matter to dynamically update the 10 closest races. It might even be more work to maintain 10 closest list manually. Well, it was fun. If Nevada, now 78-22 becomes bluer than Indiana (16-84) or Miss (17-83) or Alaska (17-83), then there will be no paths to victory for Trump.