Originally published at: https://boingboing.net/2019/08/29/why-hurricane-maps-do-a-terrib.html
…
I agree, needs more pulsating blue / red colors with captions of “get the f#ck out” tossed in.
If you have lived through one you know.
It’s not bad data viz, it’s stupid people. I know exactly what this map means and I don’t even live anywhere subject to hurricane risk.
By definition, half of the population are stupider than the average. And many are stupider than that.
When it comes to disaster warnings? If your data viz doesn’t make perfect sense to stupid people, it is bad data viz.
Maybe, but it’s much easier to re-design the data viz than it is to cure people of stupidity.
BTW It’s not just hurricane forecasts. Try asking people what flood maps mean…
…and then there are spaghetti plots:
Note the advice at the bottom:
“If anything on this graphic causes confusion, ignore the entire product”
Well there you go, I guess I’m just stupid. Until reading the article I didn’t fully appreciate the information the graphics tried to represent.
I’ve lived most of my life on the Gulf coast and reading hurricane maps seems like an everyday thing. That some people thought the cone size indicated storm size would have never occured to me.
I lived near Houston and it didn’t take me long to figure them out. I’m going to be an ahole here and say that what this study shows is that there will always be a percentage of people who don’t get it. But if you got a better idea, let’s see it.
Not necessarily. You can’t idiot proof everything. Here the data is really simple: a map, a projected path, and an expanding probability cone. Adding data like color coding would make the data viz more complicated and less likely to be understood. Removing data would make it meaningless (no map, no path, etc., would remove its intrinsic basic value). The problem here is cartographic, mathematical, and scientific illiteracy.
My hope and assumption is that the Venn diagram of low information voters and low information potential hurricane victims has significant overlap. Just sayin’ . . .
DataViz is critical - go read Tufte’s chapter on the Challenger disaster. All the info was there, but it was so poorly presented the danger wasn’t communicated and the shuttle blew up.
BTW: just today WeatherUnderground updated their 10-day forecast page to show many more data streams and they did a rather good job of it. (unusual these days!)
so you’re against improving the maps so that more people can easily understand them?
me too.
heck, i think we should print the place names in russian, with north at the bottom of the graphic, and only use colors that are difficult for the color blind to see. i mean, right now they’re not obscure enough. we can do better.
Don’t attribute to me arguments I did not make. I am all for improving the maps if possible. But it is not possible. The map is already very simple, yet people do not understand it. If you add information and complexity, it will be similarly hard for people to understand. The commonality is that people don’t understand maps and graphics and information well. The problem is not the map - it’s people. But please do improve it if you can.
First off, it’s a terrific lesson in the importance of good databiz.
Secondly, the preceding sentence is a terrific lesson in the importance of proofreading.
And expanding what-now cone?
Seriously. If you think that the average person in America can explain back to you clearly what an expanding probability cone is, then you haven’t spent a lot of time around people.
I’ve lived though every one to hit Florida since Hurricane David. I’m getting ready for this one, as it’s pointed right at me.
No one I know has ever confused the widening of the cone due to uncertainty for an increase in power. That being said, hurricanes make people panicked, stupid and desperate.