There are few, if any, better examples of Simpson's Paradox than the broad numerical arguments used in favor of gun control. Assuming that we begin by separating suicide and homicide*, the US is a nation that is largely free of gun violence. There are extremely focal areas in the US where the gun violence rate makes Fallujah in 2005 or the Honduran jungle look placid by comparison. The people who legally own guns and the people who are murdering one another at rates that surpass active war zones are startlingly separate groups.
Were this fact based upon difficult to obtain information or laborious review of large amounts of data, it would be excusable for people to be ignorant of it. HOWEVER, everyone, regardless of their political bent, agrees with it. They agree with it to the extent that they uniformly make the largest purchase of their lives, their home, based largely on this single data point. No one who can afford to live elsewhere lives in a high crime area (Chicago actually tried to pay cops/firefighters/EMTs to live in high crime areas in order to provide positive role models....there were no takers).
The issue, however, is even more focal than that. Some of the most interesting new research in the sociology of gun violence points out how extremely tight webs of individuals, mostly inter-related via gangs, produce all or nearly all of the gun violence. Violent sociopaths whose communities a.) do not/cannot trust law enforcement and b.) do not have the means or standing to use the legal system to solve disputes suffer far fewer restrictions upon their behavior than similar individuals in wealthier communities with better relationships with the police and the legal system. Those individuals drive a culture of violence which is nothing short of horrifying.
I think that if the NRA really wants to do something about gun control, it should take on inner city violence. Nobody else has made inroads, and a change in that number not only weakens the Left's best argument for gun control, it also does something positive for a group of people who have very few, if any, champions.
*Based upon international comparisons, our suicide rate is completely unremarkable, even when ignoring outliers like Japan, where the police rule any death a suicide if they can't easily badger someone into confessing to it. Combining suicide and homicide gun deaths together as gun violence only makes sense if one begins by assuming that gun control is the necessary solution, which is essentially begging the question when done in a debate over the utility of gun control. You can't use a suicide hotline to prevent murder and you can't use a mandatory minimum sentence to stop a depressed individual intent upon suicide. The problems are vastly different, but inappropriately combining the numbers makes graphs which clearly show how "more guns=more gun deaths" possible. The comparison becomes laughable in light of even a cursory glance at the overwhelming culturual factors which influence suicide, but that needs to be pointed out more often.