Light entertainment: Behold the 10 percent change!
Free NYU session on data visualization

Involuntary head tilt

A data graphic's first mission is communications. Looking cute should not come before. This one is a big fail by MIT Technology Review (link).


What makes the designer want to tilt the reader's head?

This chart is unreadable. It also fails the self-sufficiency test. All 13 data points are printed onto the chart. You really don't need the axis, and the gridlines.

A further design flaw is the use of signposts. Our eyes are drawn to the hexagons containing the brand icons but the data is at the other end of the signpost, where it is planted on the surface!

Here is a sketch of something not as cute:


Notice that I expressed time as years, and undid the log-transform on the axis of users. The mobile-related entities are labelled red. The dots could be replaced by the hexagonal brand icons.


The other two charts on the same page have their own issues. Health warning: your head may tilt.



Feed You can follow this conversation by subscribing to the comment feed for this post.


The number of users is not a log scale on the original figure

Andrew Gelman


Your graph is, from the statistical perspective, a zillion times better than what was published. On the other hand, unusual images can get attention. Recall the famous/notorious clock plot from Florence Nightingale. This is why I've moved to the idea of accepting both styles. Maybe Technology Review could feature their arty graph, but then when the reader clicks on it, they go straight to your informative graph. And then another click could go to a spreadsheet with the raw data (and as much metadata as needed).

P.S. On the details, I think your graph could be better. I'd like to see labels on all the points; one option would be to put the y-axis on a log scale so it will all be more readable. Also, then the straight lines on log scale will correspond to exponential growth, which might be more realistic than linear for most of the data. You could also play around with having the x-axis be absolute time rather than relative time, so that we could see when each platform started.

Chris Moore

I don't think that the tiltedness in itself is a big problem. That is, I think that human beings are well enough able to parse the isometric layout for it to be a legit way to lay out the information, and it's easy enough on the eye. If your graph were laid out in the isometric layout, then there'd be no problem at all.

I'm more troubled about the bottom graphs. The line graph would be absolutely fine if the labels weren't arranged to be perpendicular to the page. As it’s done here, you’re encouraged to interpret the lines as coming out of the graph as well, and then you’ve lost your ability to interpret the data.

The stacked column chart is…a bit odd. Again, there’s nothing wrong with presenting the data this way, in my view, but there are a few simple and obvious things that could be done to make it easier to follow.

Anyway, my point is this – if the chart is essentially good, then the isometry shouldn’t be a problem. There are a few easy-to-fix data visualisation errors and some basic execution errors here that make the whole thing a bit of a chore to look at. But I’ve got to say that I would enjoy looking at this if the execution were better.


Andy: You're right. I was confused by how far apart the lower points
were on the chart (but of course, the locations of the icons don't
mean anything).

Andrew: There is something to your tradeoff thesis but I wouldn't use
it the way you are proposing. I agree it is great to offer up an
aggregated chart and then allow readers to click through for details.
But not two charts showing essentially the same thing.

Yes, the absolute time chart is interesting, so is plotting the
trajectory rather than just the peak of each curve. I suspect, though,
that the designer started with the full trajectories and decided to
simplify to just the peak. (Those charts are widely available.) So the
line to the origin is just a guide to show the average growth rate.
Exponential growth is likely a better approximation to the truth.

Chris: Interesting that you like isometric but yes, the bottom two
charts are even more unnecessarily tilted than the top one.

The comments to this entry are closed.