In Defense of Pie Charts

In the world of data analysis, there are few things more reviled than the pie chart. Among "serious" data people, it is at best trivial and naive, and at worst downright evil.

I do not agree with this. The pie chart is simple, but that is its beauty. It does exactly one thing and it does it well: it shows you how much different parts contribute to a whole. This isn't the only question you ever have about your data, but when it's the question you do have, the pie chart is perfect. That is not evil and it is not naive. It is data visualization doing what it should: taking something large and abstract and saying something simple about it that your brain can easily internalize.

For example, this pie chart tells me very clearly that there is one big thing and lots of tiny things:

This one shows me there is one big thing that's about half, and three others with a significant part:

This one shows me there are five things that are about the same:

Something that it doesn't show me is exactly which one is a little bit bigger than the other. To do that, we need a bar chart. This is the classic argument against a pie chart: with the bar chart, I get to see this extra piece of information of which one is slightly larger or smaller.
But I would argue that in the vast majority of cases, that's not actually what you care about. The information that the five things are about the same is almost certainly the useful and meaningful information in this chart. If you're obsessing about which one is a little bit bigger than the other (if one is 100 but the other is 101), then in almost all cases, you are missing the point. This is exactly the reason people hate non-zero-based graphs so much. You focus on small differences (which could even be measurement error) rather than the larger picture.

If you really need to know exactly what the numbers are, you just look at the numbers. The point of the visualization is to give you an intuition about the relative sizes of things. And here, the pie chart is dramatically better.

Now, I will agree that there are cases in which do you care about the small differences, just as there are cases where a non-zero-based graph is appropriate. But that's not a reason to abandon the pie chart entirely.

In fact, the pie chart has a particularly nice property that it is very clearly not good at the things it is not good at. Visualizations to be despised are ones that look like they are showing you one thing, but are actually showing you something else. The pie chart does not do this — it doesn't show you very much, but what it shows you doesn't lie.

So join me today in celebrating a much-maligned hero, the pie chart. Or if you really still can't bring yourself to do that, cut out the middle of it to make it a donut chart.