The visual display of quantitative information
edward r. tufte
To create excellent graphs, we must define excellence and build a theory for achieving it. Tufte develops these foundational principles. But The Visual Display of Quantitative Information gives little guidance for applying them.
Graphs communicate ideas. Excellent graphs are clear, precise, efficient. They “give the reader the greatest number of ideas in the shortest time with the least ink in the smallest space.” They uphold integrity by representing the underlying data accurately in the correct context.
To achieve this excellence, graphs must follow certain principles. They maximize “data-ink”—the amount of the graphic devoted to data. They minimize “chartjunk”—decoration that distracts from the main ideas. They pursue data density—the amount of data per graphic area. And they follow aesthetic guidelines.
The Visual Display of Quantitative Information is as academic as the title suggests. It develops a theory for good charts, defends it, and uses it to demolish bad examples.
The book distills charts into some pithy principles but has little step-by-step guidance. The overall takeaway: within reason, remove ink. Keep ink that communicates data. Downplay ink that supports the data. Remove everything else. I’m oversimplifying. Tufte gives explanations and provides meaningful sub-principles. But using these sub-principles to fix my broken chart takes some interpretation.
The path to dataviz nerddom starts here. Tufte—a dataviz founder—coins “data-ink”, “chartjunk”, “small multiples”, and other terms within. But if you just want to make better charts, consider a straightforward reference instead.
Feel free to comment on Twitter