I’ve long held that design can be evaluated on fairly objective grounds: there is wrong, right and shades in between.

For example, what constitutes a wrong is an element or a motif that does not add to the cohesiveness of the communication. I have a hard time departing from an established visual language. If a link is blue, it makes sense for all links to be blue. If the color yellow is used to express “you are here”, it dilutes the communication to use it in another fashion. (This is an over-simplification and you can lean on things like context and established convention to bend the rules, but the concept remains).

In other words, what I’ve been doing is reducing design components to their atomic unit, attaching meaning to them, and using these basic building blocks to define the visual language.

However, what I consider good design is usually a bit more textured than the above methodology can justify. A better model for evaluating design is propositional density, which I’ll let Moritz Stefaner describe using the Fed Ex logo as an example:

Let us start with the notion of a proposition: in this context, a proposition is simply an elementary, atomic statement about the object at hand. “The FedEx logotype is purple” and “The FedEx logotype is set in a sans-serif font” are propositions, and because they describe salient, perceptible properties of the design, they are referred to as surface propositions.

Now, the FedEx logo became famous for a perceptual trick: The white space between the E and the x creates an arrow. This arrow induces, by its semiotic reading, a number of additional associations and readings of the design: “FedEx is on the go”, “FedEx is forward-thinking”, etc. Note that these propositions, unlike the surface propositions, are much harder to enumerate as they depend on the meaning that the observer ascribes to the arrow. These are called deep propositions as they describe underlying and often hidden meanings of the design. You can think of an iceberg, where the surface propositions are over the water – easy to see and clear cut – but the much larger part is under water.

These two concepts can be combined into a simple formula to calculate propositional density: # deep propositions / # surface propositions.

Generally speaking, good design usually has a high propositional density. On the other hand, if your propositional density is below one, you probably have superfluous, merely decorative elements in your design, which do not add to the deep reading.

Attempting to calculate this is neither practical or precise, but the recognition of deep propositions allows us to escape a rigorous system like the one I described and dabble in design elements that I would otherwise discard as decoration.

I have a love / hate relationship with A / B tests. Leaning too hard on them to make design decisions can make for very anemic process. It encourages an incremental, guess-and-check approach that feels like a task better suited for an automaton. Even when isolating one variable, the results mainly speak to “what” had the effect on behavior, rather than the “why”. I’d rather be solving problems and taking bigger strokes. But you simply can’t argue with its place in the toolbelt, especially when seeing some of the results on ABTests.com