In sifting through the facts about the gun debate, beware of the misleading use of statistics.

Take for example, "fact" number five on the main Vox card on this issue. It contains a chart from Mother Jones, which is presented to make the case — as Mother Jones puts it — that "People with more guns tend to kill more people — with guns." Or as Vox puts it, "More guns, more violence." Both statements are false, but this chart at least seems convincing.

But then, look more carefully at what the chart says — "gun deaths." Not all of those deaths are "violence." This is a bait and switch.

Now look at this chart. Each state is again represented as a dot, but this measures the gun murder rate in the 48 states that submitted timely FBI reports for 2014 (Alabama and Florida did not). The number of gun murders per 100,000 residents is plotted against each state's gun ownership rate as found in a 50-state study for the publication Injury Prevention.

That looks a lot like a shotgun blast, because at least in 2014, the statistical case that high state gun ownership translates to more gun homicides was non-existent. This is what happens when you just look at the relevant numbers instead of cherrypicking whatever numbers tell the story you want. There's no significant statistical correlation here, whether or not you choose to include D.C. (the outlier, way up above the others). And in fact, when you just look at last year's gun homicide rate by state (calculated from FBI and Census data), you see that many of the states that they claim have low "gun death" rates actually have comparatively high rates of gun murders. New Jersey, for example, had a much higher gun murder rate than Idaho and Vermont.

I haven't taken the time yet to calculate rates for every year going back a decade — perhaps for another day — but I have the raw totals and there isn't anything in the numbers that suggests a five-year average would change much.

So now you see why Vox, Mother Jones and others deliberately confuse the issue of gun violence by including gun deaths that don't involve violence: because their cherrypicking makes it seem like people in states with high gun ownership are more likely to shoot other people, when in fact it just isn't so. Perhaps there's another argument to be had about suicide, but it's a very different sort of debate. When most people think about gun control, they're worried about whether it can help stop them from being shot, not about whether it will prevent them from having a gun in case they become incredibly depressed and decide to end it all.

In short, there isn't a good case in the state data for limiting gun ownership on the basis of gun murder rates. But if Mother Jones wants to make the case that gun ownership should be limited for paternalistic reasons because it makes suicide more likely, perhaps they can publish that argument instead, alongside their advocacy for physician-assisted suicide.