Nuke this bubble chart

This unfortunate chartjunk appeared in NYT Magazine this weekend.  Once again, bubbles prove to degrade, not enhance, our ability to interpret the data.

How to explain the overlapping circles?  The solid versus empty bubbles?  Those with numbers inside, and to the left or to the right?  Those bubbles showing a precise number and those that show a range?  Pakistan ranking below India?

The chart fails our self-sufficiency test: the chart does not lose any power if we remove all the bubbles because every piece of data has been printed on it.

A two-sided dot chart may be appropriate here, shown next.  The relative scale of Russia and U.S. warheads to those of other nuclear powers is starkly revealed.

Comments

[John S.]

I think your dot chart would be more informative if you used a log scale. The U.S. and Russian stockpiles are so large that any number less than about 500 looks like zero.

[Mike Anderson]

This would actually be a good candidate for the dreaded stacked bar chart; it's count data with subdivisions. And as John S says, the 2+ order of magnitude range in the data cries out for a log scale -- or a v-e-r-y long, skinny graph.

[John F. Opie]

What makes the chart even worse is that the distinction made is specious: we actually don't know how many warheads Israel, Pakistan et al may or may not have, and the distinction between stockpiled and operational warheads is not meaningful when the country involved doesn't talk about their active systems. For all we know, Israel and Pakistan may well have warheads on missiles, or Indian and Israel, or whatever combination you might make up.

The only point where the difference between operational and stockpiled warheads makes any sense is in a detailed analysis of secondary and tertiary threat stances, as well as a discussion on the difficulties of achieving nuclear reduction goals.

And the comment on log scaling is absolutely correct: otherwise there really isn't much of a graph here.

And there really isn't much of a graph here: it's only a table...

[Kaiser]

I agree with John O in questioning whether such a chart is necessary to begin with.

As for log scale, I try to avoid them if I can. The reason is that it has both benefits and costs. The benefit of inducing separation at the small end of the scale is obvious.

The dreaded cost, for me, is that it makes visual estimation of distance impossible. The same physical distance now means very different things at either end of the scale. The lay person will likely misjudge this.

I'm not a fan of the dot chart. It becomes harder to read the more nukes a nation has stockpiled. A good chart is not stockpile-dependent.