Wednesday, October 3, 2012

Dear Mr. Playfair, your chart is too complex for our readers

Graphic by William Playfair

Lately, I've found myself thinking a bit about innovation in infographics and data visualization, as you've probably noticed after my recent posts about connected scatter plots and homages to Charles Joseph Minard.

In the past I've had my share of conversations with colleagues and managers about graphics that are deemed "too complex", "too unusual", or "too dense" for "our readers." In fact, Chapter 3 of The Functional Art starts with a story about an infographic published by Época magazine —where I worked between 2010 and 2011— that sparked a lively and very informative discussion about what "too much" means. Here's the infographic.



Época's editors are arguably the smartest team I've ever worked with, so I took their opinions very seriously. Many of them thought that this infographic was "too complex." I didn't. It is true that it contains a lot of information, but it's not particularly busy, and even if a slopegraph is not a common graphic form in news publications, it should be pretty easy to read by anyone with a minimum understanding of line graphs.

The challenge, I believe, is that traditional journalists usually think that we should dumb down visual presentations. Graphics, they say, are for people "who don't read". Information designers have been fighting this war since Tufte's The Visual Display of Quantitative Information. My rebuttal is simple: If you write for adults, you should create infographics and visualizations for adults as well, not for children or teenagers. Don't get rid of complexity and depth if they are crucial to understanding all the nuances of a story.

All this is debatable, of course, so I used part of Chapter 3 to discuss the balance between tradition (sticking to rock-solid graphic forms) and innovation (trying new things every now and then). It's true that I have a profound distrust for visualizations that are needlessly intricate and that waste a lot of space and ink (or pixels) to convey trivial messages. But adopting new graphic forms and experimenting with them is something that all infographics teams should do on a regular basis. The main reason is that any graphic form is too complex and unusual until it is not complex and unusual anymore because readers have grown used to seeing it. Think about how many scatter plots newspapers, magazines, and news websites published just a decade ago. Almost none.

I usually joke with managers in different countries saying that if William Playfair had worked as a news designer, he would never have systematized the use of statistical graphs. After all, before The Commercial and Political Atlas (1786), data were displayed just in tables. Playfair took some calculated risks and he succeeded.

So dare to try new things. Do it carefully, obviously, as you don't want to scare readers away. Use common sense. Colin Ware has put it beautifully in the recently released third edition of Information Visualization (more about it soon): "Consider adopting novel design solutions when the estimated payoff is greater than the cost of learning to use them."