Why Pie Charts are (Usually) a Bad Idea (Part 2)

This is part 2 on this topic. Click here for Part 1.

I concede, a pie can be an acceptable way to get a message across as long as the message is simple. There are other limitations as well.

The Pie revisited

I made an honest attempt to create a decent pie chart here, and it didn’t come out too bad except for the horrible line wrap in the “Our Company” label. I tried to resize the label’s bounding box but for the life of me I couldn’t figure out how.

Ignoring the formatting issue, notice none of the three major problems with pie charts exist here:

  1. Relative values are easy to gauge. There is no question “Our Company” is bigger than “Brand X”.
  2. No misrepresentation. The whole pie represents the entire entity space, i.e., the whole market.
  3. No extraneous features are here, save maybe the leader line that spares us from correlating colors with entities.

All the charts on this page have the same dimensions. Admittedly, I did choose a size that cramps the pie chart a bit to prove my next point. Even a well-formed pie has a problem that cannot be overcome by any amount of design skill:

Pie charts are not efficient.

We are only comparing two values (yes–two, not three. Do you see why?) This could be accomplished in much less space.

Like these for example:

The column chart on the left displays the same data as the pie but has the added bonus of making it easier to see the quantitative difference in market shares. There is also room for additional companies.

The line chart on the right adds a time dimension to the data and manages to remain perfectly legible in the same physical space. Look how much more information is contained in this chart, as compared to the pie. This is what we call a “high data-to-ink ratio”, and we could make it even higher, but this will do for now.
The time series would be impossible with a pie. Multiple pies could be created, and that would be an extraordinarily bad way to show a time series, but it has been done:

In this series of pies a few features can be made out but in order to recognize them the eye must wander left to right and back again several times to compare the sizes of the slices, the sizes of the colors, etc. All the information you could possibly get out of the series of pies and more is quickly assimilated in a short glance at the line chart above.

Advertisements

One Response to Why Pie Charts are (Usually) a Bad Idea (Part 2)

  1. Pingback: Why Pie Charts are (Usually) a Bad Idea (Part 1) « andyholaday

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: