I think the essence of practical, working data visualization is about saying a lot without many words. I say "practical" and "working" to differentiate between what we might do for reporting or briefing uses and what we might do for more aesthetic reasons. Of course, there is lots of territory between these poles.
There are lots of examples of great applied data visualization, done over the years (and a lot of dreck), but it is surprising how many years this means. It makes me wonder how far we have actually come. This thought arose in the context of my sneaking feeling that “data viz” is getting a bit self-referential.
I started thinking this when I saw Joseph Heller's outline for Catch-22, which is in a sense somewhat the opposite of data visualization, that is, it was planning a lot of words. But it still represents a visually assisted understanding of something complex.
Of course, writers and filmmakers have been doing this kind of visualization for a long time, but we data people sometimes seem to think it's all new. History is always helpful.
Here's my favourite.
Arguably the greatest data, or information, visualization of all time, was done by Charles Joseph Minard in 1869. It depicts the steadily worsening straits of Napoleon's 1812 Russian Campaign. The number of survivors dwindle over time and at crucial junctures. It's informative and visceral at once. This reminds me that you are should not just be displaying data, you should be trying to say something, explain something.
But look back even farther, to this example of beautiful functionality by Michelangelo. It’s a grocery list given to an illiterate servant. Does it get any better?
I wonder if we are getting better, or are we just doing more clever stuff?
So now I’ll have to ask myself: Does your work illuminate something the way Minard’s did Napoleon's folly? Does it quickly and cleanly do the job like Michelangelo's grocery list? We’ll see.