10 Comments
User's avatar
Schneeaffe's avatar

That seems like a bad example to me. Its really difficult to tell the ratio of the mixed places, or any kind of colour in the thinly populated areas. 2020 looks noticably redder than 2000. I think it would be better to start with the graduated colour map and add the density somehow - maybe vary saturation?

Expand full comment
Bob Rogers's avatar

I like dot maps and I think this is a pretty good example of their power. The lack of color in the thinly populated areas is a feature and not a bug. It shows how little vast parts of the country matter to the outcome.

The 2020 map has more red across the Midwest, but less on the west coast.

I think the best result would probably come from scaling the dot total by total vote. So if 1=1000 is the best representation of 2020 (producing 15,000 dots across the map), then using 1=665 for 2000 (which would also result in 15,000 dots) would keep the relative densities uniform and avoid the issue Peter points out about the cities blocking up.

Expand full comment
Peter Banks's avatar

Why don’t you try it and ping me! I would love to see it.

Expand full comment
Schneeaffe's avatar

How? I have experience only with the statistics kind of graphs.

Expand full comment
Peter Banks's avatar

Chatty is super helpful with this sorta stuff tbh!

Expand full comment
Brendon's avatar

Couldn’t share pics in a comment under a post, so I had to restack with my thoughts

Expand full comment
Reid's avatar

I wonder if maybe the graduated color map with counties’ opacities representing population would be good.

Expand full comment
Peter Banks's avatar

That sounds potentially very cool! You should do it with chatty it would probably be pretty straightforward

Expand full comment
Ruv Draba's avatar

Nice! So glad you produced these, Peter. 👍🏻👍🏻

Expand full comment
Peter Banks's avatar

Yeah they are a ton of fun!

Expand full comment