As horrific as mass shootings are, they are outliers. The vast majority of gun deaths in America are either suicides or one-on-one shootings. Mass shootings represent a small fraction of deaths by firearm.

Still, the premeditated shooting of multiple people is the kind of crime that preys on people's sense of security, so it makes sense to drill down and take a closer look at those events. The advocacy group Everytown for Gun Safety has inventoried all mass shootings between 2009 and 2016, and its data are revealing.

Its statistics show that mass shootings are, in the main, linked to domestic violence. That's apparent in the locations of the incidents...

  • Home — 133
  • School — 6
  • Work — 6
  • Multiple — 17

... as well as in the relationships of the victims to the shooter:
Even when you count larger-scale mass shootings, 54.49 percent of all incidents include victims who are the shooter's current or former partner, or another family member.

"Mass shooting," defined

It's important to point out that these statistics are driven by which definition of "mass shootings" is used. Everytown defines it as "any incident in which four or more people, not including the shooter, are killed with a firearm." That leaves out criteria used by some other groups, which narrow "mass shooting" to mean only indiscriminate attacks in public places.

Under Everytown's broader, "four-or-more-killed" definition, mass shootings take on a decidedly domestic character.

"We didn't actually come in with any preconceived idea of what we would find," says Sarah Tofte, Everytown's director of research and implementation. "And it's almost like we backed into this idea that, 'Oh my gosh — the vast majority of these are connected to domestic or family violence.' "

Most shocking, Tofte says, is the finding that 1 in 4 mass shooting fatalities is a child.

"We've never really broken that out before, and that's an essential part of this story," she says.

Grim details

By combing through court records and news accounts for every mass shooting since 2009, Everytown has gathered other details not generally available from government statistics. NPR analysis of the Everytown data shows:

The "red flag" category is of special interest to Everytown — and especially controversial for gun rights organizations — because it's at the heart of the gun control movement's recent strategy to promote state laws allowing judges to remove firearms from people deemed to be a threat to others. These "gun violence restraining order" laws vary in the criteria they use to take away someone's guns, so it's not clear that all of the "red flag" situations tallied here could have led to judicial intervention. Everytown uses the term "red flag" for any of a number of warning signs, ranging from threats of violence to certain firearms offenses. Twelve of the 64 "red flag" cases since 2009 consisted solely of problems stemming from substance abuse.

Only 16 of 156 mass shootings involved assault rifles

Finally, Everytown's numbers offer some new data points for the perennial national argument about semiautomatic long guns, sometimes called "assault weapons." Critics say these powerful rifles make mass shootings even deadlier; in fact, when you break down these data, the picture is more complicated:

This counterintuitive result may be explained by the fact that handguns tend to be used at close quarters, while long guns are more likely to be used at a distance, and with less precision. So in these mass shootings, someone shot with a handgun is more likely to die.

But there's no doubt that assault rifles produce more victims, overall, both dead and wounded. Despite the lower accuracy, mass shootings with assault rifles average 32 victims per incident. Shootings with handguns average 10 victims.

Copyright 2017 NPR. To see more, visit http://www.npr.org/.