With many of my data visualization clients or classes, I like to begin with a simple exercise: draw a line down the middle of a piece of paper. On the left side, write down a list of the things you identify as characterizing a good data visualization; on the right, a list of things characterizing a bad data visualization.
Invariably, similar words and phrases show up. On the good side: “clear,” “simple,” “legible,” “good colors and fonts,” “good labels,” and “accessible.” On the bad side: “biased,” “cluttered,” “vague,” “confusing, “overly complex,” and more.
But one word always catches my eye. Simple. What does that mean? Are the data simple, like dollars or percentages? Or is the graph simple, like a line or bar chart? Is it about the framing of the graph, as in the words in and around the graph are clear?
But simple? Let’s take a closer look at that word. Webster’s includes four entries for the definition of simplicity.
In our data communication work, we are often looking to be direct (4a) and have restraint in ornamentation (4b). But I don’t know if a lack of subtlety (2a) applies to a lot of effective data visualizations. We might ask readers to extend the knowledge presented in the graph to their own experience or expertise.
I’m also not sure an effective data visualization is one that is uncomplicated (1). Take this 2015 line chart from FiveThirtyEight. There are 3,282 lines on the graph—is that uncomplicated? It’s clearly easy to see in the graph how much better Messi (thick, labeled, red line) and Ronaldo (think, labeled, dark gray line) are than the rest of the 3,280 forwards and midfielders over this five-year period. The use of the position of those two lines combined with the different colors, thicknesses, and labels makes those them easy to find in what is an otherwise dense graph.
In short, I’m not sure that simplicity is what we should be striving for in our data work. I wouldn’t call the FiveThirtyEight graph “simple,” but I would say that it is “clear.” It’s clear the author wants us to see those two lines for Messi and Ronaldo. Instead of simplicity, therefore, what if we used the term clarity? Webster’s includes five entries for the word clarity, each of which, I think, does a better job of describing what we are trying to do when creating an effective graph or chart.
We want the graph to be easily understood (a) and for the reader to have a full, detailed, and orderly mental grasp (b). Like, the restraint in orientation entry for simplicity, we want our graphs to have a lack of marks, spots, or blemishes (e). (I’m taking quality of being easily seen through (d) as not applicable to our use case).
If we can strive for clarity in our data communication efforts, our work will be more easily understood and enable our readers to have a full and orderly grasp of the graph and its content. So next time, when someone asks you what makes a good graph, answer that clarity distinguishes the most effective ones.