Purple map, 2012 edition | Hacker News Reader

Purple map, 2012 edition

kryogenix.org

43 pointsmindstab43 comments

It would be great to get even more granular and get to the county level to show how even the "divided" counties are purple as well. My guess is that you might get some counties that are 100% Red or 100% Blue.

These colors are great, but almost ultraviolet to my eyes. I can only look at the map for 5 seconds. Also, the color doesn't really tell me how far from 50/50 the states are.

The burning, it hurts. I think the purple is like #17f00017f or something. Seriously, you can't just mix RGB like this; perceptual color scales like Cynthia Brewer has helped popularize are the right choice. http://colorbrewer2.org/

For anyone who wants to do this, the data's available: http://www.guardian.co.uk/news/datablog/2012/nov/07/us-2012-...

I'm going to do it right now!

I stumbled upon the same dataset, so did the following map:

http://i.imgur.com/8YjOg.png

Here is some notes:

https://gist.github.com/4034954

Edit:

If anything, I'd say that the counties themselves are more divided than I thought. The one which surprises me the most is the deep south of texas, which is quite blue.

Also, I do not claim that I did this flawlessly, there is a chance I stuffed up and it is not showing the right data, but I don't really have the time to play with it for too long :)

The first thing I noticed on your map (which is really nice, by the way) was that it's far more divisive than the per-state version. Most interesting.

I think this just highlights what people have been saying for quite a while now, that the political divide in this country isn't red states versus blue states, but urban versus rural. If you go to a level that separates those, as per-county does but per-state doesn't, you see a much bigger divide.

South Texas results aren't too surprising. There's a huge Hispanic population down there.

Ah, of course. Thanks for pointing that out.

That's what's fun. The electoral college is red blue, but the states are purple. But the counties are red and blue again (which some purple).

The most hotly contested states have (generally dense urban) blue city enclaves balancing against (generally sparse rural) red counties

I wonder how many people didn't bother to show up at the election rooms because their vote wouldn't mean anything. Being pro-Romney in a strong "blue" state or pro-Obama in the "red" one.

Well yeah, but being pro-Romney in a strong red state, or pro-Obama in a strong blue state, your vote still doesn't mean anything. The question is whether the "don't bother voting he's going to win anyway" effect is stronger than the "don't bother voting he's going to lose anyway" one.

Nice, but I wonder how objective `halfway between red and blue' is when people evaluate color. Halfway might appear to be 2/3 to some people.

All it needs is a key showing what shade is 50/50. Unfortunately one isn't provided...

Good call. Map updated to include a key!

The key isn't that helpful. Suggestion: a color slider showing continuity between two (asymptotic) extremes.

Better yet would be quantizing it into 10 or 20% buckets so people could better guage a given state more easily.

Sure, a key is provided. Look at the colour of Florida!

I don't really think that the colors are working out that well. I'm having a hard time imagining the continuum of colors between red and blue. Mapping the variable (percentage of votes for either party) to saturation or brightness of a certain hue would be more straightforward and intuitive. Although, it would make for a pretty lousy political statement as compared to purple :)

Another way to make the two basic colors more interesting could be the delta between the state and 50% and mapping it to saturation in both directions, i.e. CA dark blue, OH light blue, UT darker red… but I'm pretty sure some news outlets already did that.

Red and blue contrast well against each other, purple and blue contrast well against each other, purple and red do not however, so the map visually looks very republican heavy.

This map is kind of silly. How much of a victory does a candidate need to get red or blue to appear? Utah was 72% for one candidate. If all the map shows us is that nobody got 100%, or that 72% red and 28% blue is a shade of purple, then, well, duh.

This map makes the point that the United States is not the divided nation that pundits claim. Most states were fairly close to a 50/50 split vote wise (aka purple). Even in Utah the vote was only 3:1 Romney:Obama; they are not as different from blue states as people think.

The fact that most states are close to 50-50 means the nation as a whole is more divided, not less: there are lots of people in every state on both sides.

If people in the poorer states like Alabama voted the same way as people in richer states like Maryland, that means they have more in common. If people from a small state with low income inequality like New Hampshire are evenly split, that means whatever divides the candidates is not economic, geographic, or cultural reasons.

That's not how data works.

All this purple map shows is that the division between republican and democratic voters is prevalent in the majority of the states.

It says nothing about the demographics of the voters.

We already know the demographics of people in the various states though. If people above a certain "richness" level liked one candidate over the other, then the majority in Maryland would have voted one way and Alabama the other. (Or both the same way, if the necessary level were very high or very low.) The fact that about 50% of Alabama voted each way and about 50% of Maryland voted each way means income level does not strongly change the way people voted.

This isn't about Democrats vs Republicans, because about 40% of Americans don't identify with either party. http://en.wikipedia.org/wiki/Political_party_strength_in_U.S... Party affiliation isn't nearly enough to explain how close this election was.

This isn't about Democrats vs Republicans, because about 40% of Americans don't identify with either party.

Exactly. I've heard some studies that point to a majority of young voters (18-24?) self describe themselves as fiscally conservative and socially liberal. IMHO, neither of the 2 parties fit that bill right now so many people vote along with their particular social issue. A gay friend mentioned to me how hard voting was because she is fiscally conservative, but no way could she vote GOP (not that they are the picture of fiscal responsibility either at this point).

Another friend of mine described his feeling about the election as, "I hated to see Obama win, but I'm equally afraid of the crazy republican right wing. I want a party that is fiscally conservative that also believes in evolution."

It seems that with the vote always being near 50/50 that one of the parties could figure out that sweet spot and satisfy more than 50% of the populace. That would mean taking chances and really challenging the status quo, so it will likely never happen. After last night though, the GOP is at a crossroads and now is probably the best time if there ever was one to really make some changes. Very unlikely though.

But you don't know the demographics of who voted in each state. Just because a previous phone poll shows which party they admit to being privy to, doesn't mean that those were the demographics that went to the polls this week.

I agree that whatever divides the candidates doesn't correlate to any easy demographic measure. All I was saying is that "the entire country is purple" does not, IMO, show less division than "some states are red and other states are blue". It shows more division, because instead of the country being split into large, fairly homogeneous regions, the political split is present everywhere. It would be easier for us to all get along if "red staters" and "blue staters"--yes, I know that's a shallow metaphor but this is a small input box, work with me here :)--could just live in separate states and not bother each other. The fact that people with deeply held but opposite political convictions are cheek by jowl all across the country is part of what makes it so difficult to resolve disagreements.

DC was ~1:10 Romney:Obama which is vastly different than Utah 3:1. Honestly, when you can pick the winner in 2016 without knowing either of the candidates that creates a huge divide. Because, it forces politicians to ignore those states.

I always preferred the cartogram as shown at http://www-personal.umich.edu/~mejn/election/2012/

This seems to be showing completely different information so I'm not sure it's much of a comparison

It would show how many people voted each way, not just how much land mass is covered by states that voted one way or another.

I ment showing this data on that map. The vast land mass and small population in the west makes visuals on a normal map of the US meaningless.

50 Shades of Purple. Thank you for freedom of choice.

ha! I wish I'd thought of that tagline for the map. I might steal that at some future date, if you don't mind :)

The map is really interesting, but its hard to read with those two colours. I'd like to see it with just white-blue or white-red.

Definitely, use it!

Our voting demographics have historically been geographic by coincidence of education, industry, or creed.

In 2012, with almost no information barriers, the demographics are really more based on economics and education, with race, sex, and sexual preference also taking a strong role.

The reality is that there are plenty of idiots to go around. Anyone who pays attention to their own local politics knows that.

Really interesting in the concept, specific color choices aside, it's a great point that general portrayal really bifurcates states along partisan boundaries when they're really pretty close proportionally... Honestly the whole bipartisan system is a joke, the illusion of choice.

Now they should use saturation to indicate which percentage of the voting population actually turned out to the polls.

This probably REALLY cool. #colorblind

What a horrific colour scheme. You want colours that contrast well; a red-green gradient would be much better. A "neutral" state would then be an easy to discern shade of brown.

Translation, for people with a common form of color blindness:

"a brown-brown gradient would be much better. A "neutral" state would then be an easy to discern shade of brown."

Hah, good point. Should've known better. Still, it's not hard to find something better than shades of fuchsia. Blue to orange with white in the middle perhaps?

Reply on news.ycombinator.com