34 Years of Mass Shootings in One Chart

July 8, 2016: This chart has been updated to include the killing of five police officers in Dallas.

No one disputes that the massacre at an Orlando nightclub early Sunday morning, which left 49 victims dead and another 53 injured, is the worst mass shooting in U.S. history.

While the attack was singular in its scale, the shooting was also part of a familiar story. It was the third public shooting that left at least three victims dead in 2016, following shootings in Hesston, Kans. and Kalamazoo, Mich. in February. Barely six months earlier 14 people were killed in the rampage in San Bernardino, Calif.

There is no one authoritative source on how many people have died or been wounded in such incidents. After reviewing different sources of data, we’ve relied on the record maintained (and generously open-sourced) by the magazine Mother Jones to chart recent mass shootings and their casualties.

It’s worth noting that mass shooting account for a very small percentage of gun-related homicides in the U.S. each year, leading some to argue that the focus on these horrific events draws attention away from the larger menace of smaller death tolls writ large.

It’s also possible to find wildly divergent figures for mass shooting casualties. The FBI’s historical data on mass shootings, for instance, was found to be both inaccurate and highly incomplete, according to a 2013 investigation by USA Today. The FBI acknowledged to the newspaper that its data was flawed.

The Gun Violence Archive reports 139 mass shootings in 2016 alone as of this writing. The main reason for the discrepancy is that they define a mass shooting as any in which four people are shot, but not necessarily killed, while the traditional definition is that four people must be killed in a public space, not including the gunman. (In 2013, a federal mandate lowered that threshold to three fatalities.) Mark Follman, who leads the Mother Jones effort, explains the discrepancies in depth in this New York Times column, and the Washington Post visualized the differences in counts between several sources.

Our graphic does not include the perpetrator among the victims in cases where he or she died at the scene. We independently verified that each of the victim counts in the Mother Jones dataset did not include the shooter, which is the cause for several small discrepancies between this graphic and the original dataset.