Hacker News Reader

Comment by mbostock | Hacker News Reader

A couple hidden features: 1. You can option-click on any of the buttons to see the transition in super slo-mo. (This was mainly for debugging, but it's fun to see how the transitions work in more detail.) 2. You can double-click on any part of the tree, and it will zoom in by one level.

Also, we did a variation that used state-level probabilities to weight the tree. This gave a sense not just of the logical possibilities, but of the likelihood of each, which I liked. However, the FiveThirtyEight state-level probabilities are not fully independent, so you can't multiply them together to compute conditional probabilities. Perhaps next election!

View on HN

That is cool. And thanks Mike, for your incredible d3.js library.

I like the graph, but it's a pity that the weighted version is not available :(. It was the first thing I looked for. It's more difficult to explain, but it much more useful that only counting the number of branches.

(I don't mind if it would use only the naive approach and just multiply the probabilities. But it could be a problem if the final probabilities are very different when they are estimated with the correct method and the naive method.)

However, the FiveThirtyEight state-level probabilities are not fully independent

This is interesting, are they linked in some manner to national polls, or is their something more nuanced?

538 derives their probabilities by running simulated elections. http://fivethirtyeight.blogs.nytimes.com/methodology/

Somewhat intuitively, the simulations themselves have concepts of "national movements", where a candidate can (with some probability) gain 1%, 2%, etc. across all states.

The state probabilities are then derived from how many simulations/total candidate A won. However, partly because these simulations actually had national movement effects, you'll find that odds of winning both Florida and Ohio (as in # simulations/total candidate A wins both states) are not the state odds multiplied (they odds actually should be higher).

That said, with the raw simulation data, the graphic could show probabilities for every combination.

I wonder if that national movement concept explains that weird spike in his histogram at 330 EV. Maybe it's because there's some "movement" concept that tips several states at the same point?

Couldn't you get the results for all the different simulations and add up how often each combination arises and get probabilities that way?

Yes

You should weight the branch-level possibilities.

Reply on news.ycombinator.com