Foundational dataviz example

I'm really excited this week.

After more than 5 months of hard work, my Matplotlib Journey project is finally complete! If you want to learn how to make great charts with Python, take a look. I'm very proud of what we've created and the feedback is awesome.

The last module focuses on making maps in Python. As always, we combined Python techniques with dataviz theory. It was the perfect opportunity to dive into a dataviz story that I believe is foundational.

Let me share it with you and sharpen your dataviz skills at the same time!

Impeach this

In October 2019, Trump posted a choropleth election map with those words.

The map looked striking: a sea of red counties that seemed to scream overwhelming support.

It racked up 162k likes.

Take a moment to look at it. How accurate do you think it is? Could it be misleading?

Land doesn't vote. People do.

Luckily, many pointed out how misleading this map was.

Karim Douieb created an alternative: a bubble map where each bubble represents a county, sized by population.

This version tells a very different story, probably much more accurate.

This story highlights a critical flaw of choropleth maps:

Large areas dominate the visual, even if they represent very few people.

This can create a false impression of dominance when what usually matters is the number of people, not the amount of land.

Another perspective

There's another issue mentioned by someone on my LinkedIn post. A choropleth map like this uses a red/blue binary. A county with 99% Trump votes looks the same as one with 50.05%.

Here's an alternative using a color gradient to show shades of support:

source

Different, right? What we would really need is probably a combination of both: bubbles + gradient!

Why this matters

Even though this story dates back years, I believe it's a must-know for anyone working with data visualization.

It's rare for a dataviz topic to go mainstream and spark such widespread discussion about how charts can mislead.

In fact, this moment arguably kickstarted the improvements we've seen in election maps ever since!

Have a wonderful week!
Yan