Tufte Talk: Chapter 3

"Graphical competence demands three quit different skills: the substantive, statistical, and artistic."

I am in this picture, and I do not like it.

So chapter three follows the topic of chapter two by asking, "Why do people that make bad charts continue to get work making bad charts?" And if this chapter was on low volume when this book was published (1983), it has likely led to a blizzard of thesis statements on behalf of anthropologists and statisticians alike in the last 10 years. Or maybe even in just the last three and a half.

I find it haunting that this book calls out a culture of "assuming the audience is bored or confused by graphics" as culprit for the disintegration of data visualization integrity over two decades before we would have to explain to belligerent citizens what 'flatten the curve' actually means.

Now here we are. Neck deep in both chickens and eggs. I'm growing less concerned with which came first.

What can be easily taken away from this chapter - which is either a Nostradomus style prophecy or the user manual for information dystopias - is put simply in the quote above. Tufte talks about how the modern world hands data graphics over to artists that shrug their shoulders at statistical or skeptical thinking, instead focusing on aesthetics and entertainment over the truth. This is then justified by two further heaps of bullshit: that graphics are boring and people are stupid. I've read enough X-Men comics to know what happens when you fear and distrust an entire demographic. You're the badguy.

I think the part that punched the hardest was this quote:

"If the statistics are boring, then you've got the wrong numbers. Finding the right numbers requires as much specialized skill - statistical skill - and hard work as creating a beautiful design or covering a complex news story."

He then goes on to highlight an exceptional example from The New York Times where the graphic was sub-par compared to most high school textbooks and was, somehow, coupled with an article that makes even Tufte's pseudo-sesquipedalian lingo sound like an attempt to verbally bully me off of the schoolyard. Somehow we use the same pages to demand our audience is thirsty for the verbose, but too overwhelmed by a squiggly line on one quarter of the Coordinate Plane?

But, let's be honest. What am I really trying to do here?

I'm trying to cover up the fact that, at least on the surface, I've got oodles of artsy, a good, solid foundation for the substantial, and... Well, I've watched more than a few YouTube videos about statistics (Thanks, Adriene Hill!)

There are days where I really, truly worry that the world will look at my resume and think, "What is this guy?" I've been sliced down the middle into two stereotypes: art nerd and code geek. What does the world need with someone drinking SQL tutorials in equal measure with Adobe Illustrator Pen techniques?

Thank goodness Tufte clears it up. You need me because I'm NOT one of those two sub-categories. I've been, over and over again, the one guy on the team that is willing to ask questions as if I didn't already know (Substantive!) and I've struggled to make every project I turn in easy and pleasing to look at (Aesthetics, check!), and now, two years into it, I'm adding the third piece of Triforce to my skillset: the numbers, my friend. The numbers.

At first this chapter felt like getting called out. It's easy to say, "So you've designed a milk carton, and now you want to do dashboards?... yeah, we need data scientists for that." But that's also a misconception. You need someone that connects the information to the insight. You need someone that doesn't just code the right query, ask the right questions, and make the graphic hit home.

You need all three.

Take a graphic with boring numbers (the wrong numbers) and make it engaging? You're skewing the message, you're bullshitting. Take a graphic that is accurate and rigorously double-checked and present it without clarity or audience awareness? You're loosing the message, you're playing to the false impressions that Tufte is calling out.

My biggest take away from this chapter is pretty personal. I can't get away with just smugly saying, "Hang on a sec, there, code-jockey. Let me show you what a real dashboard looks like," and then throw some brand colors and custom icons into the mix. I can't even call it good at just asking the right questions. I have to be ready to do the math.

What once felt like stretching myself too thin, like I was serving too many masters, is starting to feel more and more like the only way to really be able to do this right.

"The conditions under which many data graphics are produced - the lack of substantive and quantitative skills of the illustrators, dislike of quantitative evidence, and contempt for the intelligence of the audience - guarantee graphic mediocrity. These conditions engender graphics that (1) lie; (2) employ only the simplest designs, often understandardized time-series based on a small handful of data points; and (3) miss the real news actually in the data."