Blog series: User-Centered Data Visualization. Part 3 – What is this???

Tobias Komischke / Monday, June 18, 2012

Recently, TIME magazine published this info graphic about top-earning movies. It’s kinda interesting. I spent much more time looking at this than I normally would have. So by that definition, the graphic already reached its goal of engaging its audience. I guess what drew my curiosity was that it’s an own species of thing – something like a mutated pie or donut chart.

Source: TIME magazine, 4/16/2012

Let’s see: what are its properties and what are properties of a purebred donut chart?

Donut charts have a circle shape –> check
Donut charts have segments –> check
In donut charts the size of segments map to value they represent –> check
In donut charts the largest segment starts at the noon position, the others are then lined up in descending order clockwise –> check. There’s even a visual indicator that shows that clearly.
In donut charts all segments together add up to 100% or a total number of the whole –> no check. Here’s the issue. There’s no whole that the segments would be part of. The segments are just an ordered list of movies that earned more than $1 billion.

Now, granted: this info graphic does not have the ambition to be super correct in a scientific way, but it’s still confusing to see the information shown in this way. After all, you could argue that if you don’t want to apply the standard donut characteristics you may just as well show the segments in ascending order. Is it an overstatement to say that TIME forced the data into the wrong chart type?

How else could the same information be provided?

Maybe not as nice but bar charts are “better” than pie charts. “Better” as in: less time required to understand them and higher degree of accurate interpretation.

Check out the other parts of the User-Centered Data Visualization series: