Subscribe to DSC Newsletter

10 Resources to Help You Stop Doing Pie Charts

Update: A few commentators were looking for alternatives to the pie chart. The alternative recommended by almost everybody in this list is the simple bar chart. Here is a perspective from Stephen Few:

But what if we could display this same information in a graph that is easy to read; one that adds useful meaning by allowing us to compare the magnitudes of the values without labeling them? Now the values can be compared with relative ease and precision, relying solely on the graph, without labeling the values. What value does this bar graph offer, compared to a table? In little more than a glance it paints a picture of the relationships between six companies regarding market share. Not only is their relative rank apparent, but the differences in value from one company to the next is readily available to our eyes. Could we construct this same picture in our heads from a table of the same values? Perhaps, but it would take a great deal of effort and time. Why bother when a graph can do the work for you and tell the story in a way that speaks directly to the high-bandwidth, parallel imaging processor in your brain, which operates much faster than the part of your brain that handles text, which is needed to process tables?

Pie Chart Vs Bar Chart Example

In our analytics careers, we would have made at least one pie chart to show the output of our data & then had a slightly nagging uncomfortable feeling about it. Here are 10 resources by experts & practitioners to help you stop making pie charts & make that uncomfortable feeling go away.

1/ Edward Tufte, the god of visualization in his landmark book “The Visual Display of Quantitative Information

A table is nearly always better than a dumb pie chart; the  only worse design than a pie chart is several of them, for   then the viewer is asked to compare quantities located in  spatial disarray both within and between charts [...] Given   their low density and failure to order numbers along a visual dimension, pie charts should never be used.

h/tip

2/ Stephen Few, another luminary on data visualization wrote an entire note on not using pie charts:

Pie charts are not without their strengths. The primary strength of a pie chart is the fact that the message “part-to-whole relationship” is built right into it in an obvious way. Children learn fractions by looking at pies sliced in various ways and decoding the ratio (quarter, half, three quarters, etc.) of each slice. A bar graph doesn’t have this obvious purpose built into its design. Not as directly, anyway, but it can be built into bar graphs in a way that prompts people to think in terms of a whole and its parts. This can be accomplished in part by using a percentage scale. It is easy and natural to think in terms of various percentages in relation to the whole of 100%. Seeing a bar extend to 25% along a quantitative scale conveys a part-to whole relationship only slightly less effectively than a pie chart with a quarter slice, especially if the bar graph’s title declares that it displays the parts of some total (for example, “Regional Breakdown of Total Revenue”). Despite the obvious nature of a pie charts message, bar graphs provide a much better means to compare the magnitudes of each part. Pie charts only make it easy to judge the magnitude of a slice when it is close to 0%, 25%, 50%, 75%, or 100%. Any percentages other than these are difficult to discern in a pie chart, but can be accurately discerned in a bar graph, thanks to the quantitative scale.

3/ Story Telling with Data: My main beef with pie charts is this: our eyes aren't good at attributing quantitative value to two dimensional spaces. In English: pie charts are really hard for people to read! When segments are close in size, it'd difficult (if not impossible) to tell which is bigger. When they aren't close in size, the best you can do is determine that one is bigger than the other, but you can't judge by how much.

4/ Richard Hollins highlights 3 main problems with using pie charts

  1. Problem 1: pie charts are bad at communicating the basics

  2. Problem 2: pie charts are worse at showing trends

  3. Problem 3: pie charts are terrible at showing trends in absolute numbers

5/ Oracle, the market of umpteen BI tools lists a top 10 countdown on the reasons not to use pie charts

  1. Number 10 - Pie Charts Just Don't Work When Comparing Data

  2. Number 9 - You Have A Better Option: The Sorted Horizontal Bar Chart

  3. Number 8 - The Pie Chart is Always Round

  4. Number 7 - Some Genius Will Make It 3D

  5. Number 6 - Legends and Labels are Hard to Align and Read

  6. Number 5 - Nobody Has Ever Made a Critical Decision Using a Pie Chart

  7. Number 4 - It Doesn't Scale Well to More Than 2 Items

  8. Number 3 - A Pie Chart Causes Distortions and Errors

  9. Number 2 - Everyone Else Uses Them: Debunking the "Urban Legend" of Pie Charts

  10. Number 1 - Pie Charts Make You Look Stupid and Lazy

6/ Bernard Marr, a consultant on analytics & data: From a design point of view, a pie chart takes up far too much space to convey a set of data compared to other options. In addition, the labels don’t line up, so the result becomes cluttered and hard to read — strike three against pie charts, as they often make the data more complicated than before. Pie charts also don’t work well at various sizes. A small pie chart is all but useless, whereas a small bar graph or line graph can still easily show differences in data.

7/ Evolytics - An analytics firm : Your brain processes heights of bars more easily than it process areas in a pie. Using a bar chart also provides you more real estate to display the values, either as dollar amounts and/or percents of the whole (as you were trying to convey by using a pie chart). To take this one final step further, if you only care about comparing how each category performed against the others for one period in time, you can isolate the bar chart to show just the year in question.

8/ Business Insider: The reality is, humans aren't very good at comparing slices of a circle when it comes to size. It's the reason you probably found trigonometry and radians a lot more difficult than you found basic rectangle geometry.

9/ Steve Fenton makes three arguments against pie charts:

  1. ARGUMENT ONE – COMPARING INDIVIDUAL VALUES

  2. ARGUMENT TWO – SIZE MATTERS

  3. ARGUMENT THREE – COMPARING COMBINED VALUES

10/ Tom Wilson, a data geek gives five reasons pie charts are not effective:

  1. Pie charts rely on colors to differentiate the data and the color differences are not easily apparent

  2. Pie charts labels are much hard to link to the data

  3. Pie charts suck at the small percentages.

  4. Pie chart are not easy to spot small differences

  5. Pie charts use a lot more space than bar charts & have more white space

Views: 14194

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Chandrasekhara S. ("C.S.") Ganti on August 26, 2015 at 6:17am

 Pls. also note the following, in the ester era of PC World -- we used SYSTAT a  very unique single license --- not any renewals fees etc had excellent modules -- pedagogical walk through for graphics, I survived on the PC SYSTAT for almost a decade plus at my work, and the graphics were excellent - Any questions answered by Stattisticians - for algorithms, methods / explanation. and point to source of a paper from journal etc.

Leland Wilkinson

Adjunct Professor of Computer Science, University of Illinois at Chicago
VP, Data Visualization, SKYTREE

leland at skytree dot net



Projects:

SYSTAT
The Grammar of Graphics
nViZn
Piano

Comment by Ron Segal on August 25, 2015 at 11:03am

Switching this around a bit, pie charts can be useful where the emphasis is on conveying an impression of proportions of a whole, e.g. how much of the pie has been eaten, where precision isn't needed (and assuming perceptually sound colours or patterns can be selected). Bar charts are good for conveying relative quantities, less so fractions of a whole, which requires additional annotation.  They can be used where more precision is needed. So for me at least the question remains, is there a better graphic than pie charts for quickly grasping proportion of a whole?

Comment by Carol Marquez on August 25, 2015 at 8:54am

The first rule in advertising is to keep it simple. If you need lots of numbers and narrative the purpose of the graphic is circumvented. A more obscure reason that pie charts fall short is that individuals with color blindness may not see the data at all. I painfully learned this when creating a dashboard with pie charts for a new SVP and overcompensated with patterns. It was a complete mess and was so busy, no one could see or understand the data. (I will definitely check out the Edward Tufte book) Bar Charts are probably the most straight forward, especially in comparative situations.

Comment by surabhi srivastava on August 20, 2015 at 8:27pm

I thought i will get the alternative method to pie charts.But here it all the about disadvantages which we are already inline with

Comment by Ron Segal on August 20, 2015 at 10:05am

A wee bit disappointed as I was expecting advice on compelling alternatives to pie charts rather that what appears to be largely subjective criticism of a hackneyed technique that is nevertheless often perfectly capable of getting across the message.

Comment by Rebecca Barber, PhD on August 20, 2015 at 7:39am

And yet, I STILL cannot convince people to avoid pie charts because execs keep asking for them anyway...

Follow Us

Videos

  • Add Videos
  • View All

Resources

© 2017   Data Science Central   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service