I've always hated the data-ink ratio metric. It's totally invalid.
The proof of this is right there in one of the examples in The Visual Display of Quantitative Information.
He shows an example where the axes of a scatter chart are shortened to the range of the data in each direction. This is a data-ink win-win! You get more information (the range of the data, at a glance), and it doesn't matter how little importance that is because you also have less ink!
But of course it actually looks terrible. To think otherwise, you have to be so busy thinking about metrics that you ignore your eyes. The axes no longer overlap in the corner, so the whole thing looks like a soup of visual fragments rather than a single cohesive thing.
Imagine you had several of these on a page side by side. It would be far harder to visually focus on one for a moment, because the fragments of different subfigures would be one big visual soup.
The solution is to put boxes around them (they don't have to be big thick black boxes, just something to visually separate the different things). But of course Tufte hates boxes. They add ink for no information content.
It's just bad science. Or brains don't process black pixels, one at a time, spending mental energy on each one. They see different visual fragments, having been heavily pre-processed by the CNN in our optic nerves before reaching our frontal contexts. The thing you actually need to minimise is visual complexity. Often, adding ink (for no information) helps with that.
The proof of this is right there in one of the examples in The Visual Display of Quantitative Information.
He shows an example where the axes of a scatter chart are shortened to the range of the data in each direction. This is a data-ink win-win! You get more information (the range of the data, at a glance), and it doesn't matter how little importance that is because you also have less ink!
But of course it actually looks terrible. To think otherwise, you have to be so busy thinking about metrics that you ignore your eyes. The axes no longer overlap in the corner, so the whole thing looks like a soup of visual fragments rather than a single cohesive thing.
Imagine you had several of these on a page side by side. It would be far harder to visually focus on one for a moment, because the fragments of different subfigures would be one big visual soup.
The solution is to put boxes around them (they don't have to be big thick black boxes, just something to visually separate the different things). But of course Tufte hates boxes. They add ink for no information content.
It's just bad science. Or brains don't process black pixels, one at a time, spending mental energy on each one. They see different visual fragments, having been heavily pre-processed by the CNN in our optic nerves before reaching our frontal contexts. The thing you actually need to minimise is visual complexity. Often, adding ink (for no information) helps with that.