512 Paths to the White House • Hacker News Board

Re: 512 Paths to the White House

#31

Post by thaumaturgy » Sat, Nov 03, 2012, 5:57 AM UTC

raldi wrote: post #22

gwright wrote: post #14

Earlier quoted context omitted.

A purely popular vote system has its own flaws. In particular, national elections would be reduced to politicians catering to voters in a few high-density regions rather than attempting to appeal to a geographically dispersed electorate.

Most of the country doesn't live in high-density areas. If you only catered to urban voters, you'd lose in a huge landslide.

You aren't kidding! I was skeptical about this, so I checked. Given the list of urban population areas from http://en.wikipedia.org/wiki/List_of_United_States_urban_are... and a U.S. population estimate of ~312m, you run out of urban centers with >1m people (at #42, Salt Lake City) before even reaching half of the U.S. population (131,980,346)! To put this into some perspective, San Juan, Puerto Rico would be an attractive stop for an aspiring U.S. President (at 2.1m people), if they had the right to vote!

Suddenly, U.S. transportation & communications infrastructure problems make a lot more sense. I knew the U.S. was a big place, but I vastly underestimated how spread out our population was.

Re: 512 Paths to the White House

#32

Post by jashkenas » Sat, Nov 03, 2012, 6:02 AM UTC

cedrichurst wrote: post #2

Thanks Jeremy, Shan and Mike. I'm continually blown away by the data journalism you're doing over at nytimes on the election. It's truly an inspiration.

I've got nothing whatsoever to do with this one -- it's a Shan + Mike joint.

But thanks for the kudos. We should have some fun stuff for y'all on Tuesday night...

Re: 512 Paths to the White House

#33

Post by drivebyacct2 » Sat, Nov 03, 2012, 6:05 AM UTC

leif wrote: post #15

tomasien wrote: post #3

I think this is a great way to explain to someone who's not intimately familiar with electoral politics why people think Obama is going to win despite razor thing polls. Predictions mean nothing of course but: If Obama wins Florida, Romney has exactly 1 path to victory: winning every other swing state. If Obama wins Ohio, Romney has only 11 ways to win. If Obama loses Florida, Virginia, NC, and Ohio he could still po…

Obama's going to win. The only reason you hear pundits talking about it is because they get paid to talk about it. The media controls the messaging these days, not the candidates or the people, and it's in the media's interest to maintain as close a race as possible until the bitter end. That's why even the most partisan pundits, even when everything's in their candidate's favor, keep claiming that "this is such a cl…

NO BUT [insert crude reduction of Nate's science down to the level of tv pundits]. I'd love to know more about Nate's model, but I trust it a hell of a lot more than the raw data of Gallup (seeing as Gallup's nearly the only one left giving him a chance. Dick back-peddaled today, Rasmussen did more or less too.)

Nate actually has a book about why this occurs and what it does to the conversation of politics in elections in the US.

Re: 512 Paths to the White House

#34

Post by waterlesscloud » Sat, Nov 03, 2012, 6:08 AM UTC

calinet6 wrote: post #30

waterlesscloud wrote: post #28

Earlier quoted context omitted.

So instead of your vote being lost in the mass of California or New York votes, it would just be lost in the much larger mass of national votes. You'd still be in a winner-take-all situation with your vote, but on a national level rather than a state level. I don't think this change would have the effect people imagine. At all.

It would not appear to, but in effect, it would make the entire election simply based on popular vote. Therefore your vote is not lost, it simply becomes equal. This is far better than your vote actually being lost in the mass of California votes, since you, living in California, would still have some impact on the national outcome. And candidates, looking for popular votes instead of electoral votes, would have much…

What specific concrete difference do you see happening in that case? What does it mean to "campaign to all people"?

Re: 512 Paths to the White House

#35

Post by usaar333 » Sat, Nov 03, 2012, 6:11 AM UTC

001sky wrote: post #24

mbostock wrote: post #17

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 possibili…

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.

Re: 512 Paths to the White House

#36

Post by lbarrow » Sat, Nov 03, 2012, 6:12 AM UTC

ninetax wrote: post #5

This is amazing! It's everything a visualization should be: simple, to the point, interactive.

It's also pretty deceptive, as it gives each path equal visual weight when they aren't all equally likely

Re: 512 Paths to the White House

#37

Post by agildehaus » Sat, Nov 03, 2012, 6:15 AM UTC

trueluk wrote: post #21

I'm not sure why Nebraska's second congressional district isn't included. According to FiveThirtyEight, Obama's chance (14%) of winning one vote from Nebraska is higher than Romney's chance (11%) of winning Nevada. As a matter of fact, the last FiveThirtyEight visualization I saw linked here ( http://news.ycombinator.com/item?id=4657826 ) didn't include the one electoral vote Obama won in 2008 from Nebraska either.

Maybe they're not counting that electoral vote here. Can anyone tell? This diagram needs an electoral vote count.

Re: 512 Paths to the White House

#38

Post by lbarrow » Sat, Nov 03, 2012, 6:20 AM UTC

leif wrote: post #25

drewrv wrote: post #19

Earlier quoted context omitted.

So what? Why should some peoples vote matter more than others?

It's all about what kinds of questions get discussed and addressed during campaigns. The point of the campaigning process is chiefly to elect new representation, but secondarily (and closely so) to debate and test competing theories on the primary problems of the time. In the electoral college model, the concerns of the swing states get the most attention, for reasons which are far too numerous and obvious to go in t…

The game theory is not complicated at all for a first-past-the-post election with two candidates: candidates will cater to the median voter. There's even a theorem proving it: http://en.wikipedia.org/wiki/Median_voter_theorem

"Skewing" a democratic political system towards regions with more people in them would be vastly, vastly preferable to our current system. A number of big warts in US policy can be explained, at least in part, by the biases and misincentives introduced by our election system.

Probably the most obvious is the federal government's support of corn ethanol: supposedly it's done in the name of energy independence, but in reality it's done because almost all US Senators believe that they're going to run for President one day, and none of them have to try to win an Iowa caucus with a vote against corn ethanol on their record.

Re: 512 Paths to the White House

#39

Post by lbarrow » Sat, Nov 03, 2012, 6:24 AM UTC

rayiner wrote: post #26

gwright wrote: post #14

Earlier quoted context omitted.

A purely popular vote system has its own flaws. In particular, national elections would be reduced to politicians catering to voters in a few high-density regions rather than attempting to appeal to a geographically dispersed electorate.

So you're saying that even the election system artificially subsidizes sprawl in this country?

It's not so much that it subsidizes sprawl as much as it subsidizes low-population states. They aren't quite the same thing: Delaware is reasonably densely populated, but it benefits quite a bit from the electoral college.

Re: 512 Paths to the White House

#40

Post by comex » Sat, Nov 03, 2012, 6:29 AM UTC

waterlesscloud wrote: post #34

calinet6 wrote: post #30

Earlier quoted context omitted.

It would not appear to, but in effect, it would make the entire election simply based on popular vote. Therefore your vote is not lost, it simply becomes equal. This is far better than your vote actually being lost in the mass of California votes, since you, living in California, would still have some impact on the national outcome. And candidates, looking for popular votes instead of electoral votes, would have much…

What specific concrete difference do you see happening in that case? What does it mean to "campaign to all people"?

Candidates currently essentially have no reason to pay any attention whatsoever to non-competitive states, but if the election were a popular vote, going from, say, 20% to 40% in a state would be a significant advantage.