I was happy to attend Edward Tufte’s excellent class Presenting Data and Information on December 6 in San Francisco. The New York Times calls him “The Leonardo da Vinci of data.” Says Wired about the course: “One visionary day….the insights of this class lead to new levels of understanding both for creators and viewers of visual displays.”
One thing I loved was his notion that you know you have a good design when you’ve got people all having different experiences because of it – like it’s the cacaphony that’s the success, instead of passive silence. To him, the sign of a good data graphic is when everyone is thinking different things, asking questions and interacting, evoking different experiences and cognitive style. Saying things like “What’s missing is…” Once again this makes me feel like the diversity itself is the key, and if you let people “be who they are” by whatever you do, everyone wins.
I also loved that he brought Galileo to the mix, and lingered on the importance of Galileo’s visual evidence of the sun spots – which ultimated rocked our world in the most profound way. Before Galileo: Wordy Authority; After Galileo: Evidence. Tufte makes the analogy here to the importance of good visual design as evidence itself. For the first time, the earth moves — with visible certainty. Tufte also passed around an original print of a 1570s book (I didn’t catch which one, because I was dumbly fumbling with my camera) that you could say demonstrates early 3D prototyping. So how far have we really come in several centuries?
Below are some of the other highlights I got out of it.
On the Human Eye-Brain System
He spent some time dwelling on the human eye-brain system while lamenting what we have done to “dumb-down” the vastly superior capabilities of this system. “We can process tons of information visually, so how come our graphics look like cartoons?”
- Find a supergraphic — an intriguing texture of content.
- Get the highest resolution monitor you can find – yet it will still be in essence an insult to the level of resolution your brain is capable of processing.
- “Nearly every pixel on a serious computer user’s screen is devoted to content.”
- He went into a beautiful muse about the “ocean of streams and stories,” fluid and changeable, that we are able to process.
On Software Design
- Design the surface first, then the application. This is user-centered design.
- “Interface should not be the byproduct of someone’s application.” “Start with what the user sees.”
- “Documentation is a fundamental quality control measure.” Just its mere presence indicates care, craft, and credibility.
- “Good design cannot salvage failed content.”
On Web Design
- You want a documents-based interface vs. a marketing-based interface.
- Design mimics hierarchy.
- The top-ten news sites have 250-400 links on their opening screen! Scroll and link. Google News — good design for a site; flat, two layers. Pull it up to the front.
- Good design is 90% content (not 50%).
- Have a navigation bar and an invisible grid that helps the type fall in an orderly way.
- Can’t be fancier till we get greater resolution on our computer screens (compared with paper).
- Don’t devote your homepage to the marketing experience — the people have already arrived! Fill with content.
- Think about causality. What caused things? What was the reason? What’s the evidence?
- In the diagram, annotate everything.
- Create a “linking line” – an important gesture but it shouldn’t be visually active. Maybe it’s a grey line – it’s “not yelling” – it just does its job; make the “smallest effective difference.”
- “Don’t get it original — get it right.” Think of who’s good at tables, and then go there (the newspaper?) and take them and adapt them. Think first: what is the reason for my presentation? Then think: where can I find an excellent idea to copy? Talent gets it original; genius steals (ts eliott).
- High-resolution data.
- Annotate data lines — not legends. Legends take people away from the data.
- The best designs are invisible. You want people thinking about the content, rather than the poverty of information or your amateur design efforts. Minimize the “design admiration/figuring it out” time.
- “People are thinking about the content, not the bullets.”
- From sunspots and Galileo, Tufte vectored in to small multiples, and how these were “about the best design around.” Why do they have credibility? Because it implies “I’m showing you all my data.” The biggest threat to credibility is cherry-picking. “Am I seeing the results of evidence, or evidence selection?”
- From here, he vectored into the concept of the 45% slope (easiest to intuit — ceiling/floor is not intuitive) and how you want to see “lumpy” in line charts.
- Thus are spawned sparklines – word-sized inline graphics. In this rather futuristic view, graphics are everywhere and no longer “a special occasion,” having the same (high) resolution as ordinary typography.
- Some MS Office plugins will generate sparklines, but he seemed to wish there were high-quality open-source solutions.
On Email and Newspapers
- Newspapers will stay alive because of their high resolution (and with sparklines).
- Email — He thinks it must be just about “all over” — looking at the amount of “marketing experience” it has become.
- “People don’t become stupid just because they come hear you talk.”
- “The reason for the presentation is the *thing contained*, not the container.”
- On multimedia: maybe we shouldn’t do it all at once. If you need to animate, there should be content reasons to animate.
- The bullet list has reduced our thoughts to the level of grunts. A bullet is a grunt – sentences have been replaced with grunts. Sentences are causal: “Something is doing something to something else.” A sentence is smarter than a grunt.
- Use PowerPoint solely as a full-screen projector.
- Give people handouts in advance of your presentation – provide a technical report – 11×17, folded in half, written in Word (not PowerPoint), and permanent. It summarizes: 1) what is the problem? 2) who cares? and 3) what is the solution? Leave these handouts behind so that people take physical evidence with them.
- Reading happens 3-4x faster than talking.
- Welcome people looking at your things.
- Rehearsal improves performance.
- Show up early.
- Never apologize in the introduction.
- Finish early.
And then he did.