Can A Computer Really Track Down Serial Killers Hiding Among Us?


Serial killers are, officially, vanishingly rare. Contrary to what pop culture tells us, you’re more likely to get in a car wreck, shot in a robbery, or hit by lightning than fall afoul of a murderer. But that doesn’t mean they aren’t out there, and a computer algorithm digging through decades of murders may be the key to finding and stopping them.

The New Yorker has published a fascinating profile of Thomas Hargrove, a data-driven journalist who has been archiving and studying America’s murders (750,000+ cases) since 1976 for the Murder Accountability Project (MAP). Most of these cases are not committed by serial killers. They are the sad, ugly stories any journalist knows by heart; the abusive spouse nobody stopped, the petty argument that flared up into a tragedy, the cheater caught or attempting to cover their tracks. Yet Hargrove does uncover patterns:

In August of 2010, Hargrove noticed a pattern of murders in Lake County, Indiana, which includes the city of Gary. Between 1980 and 2008, fifteen women had been strangled. …Four years later, the police in Hammond, a town next to Gary, got a call about a disturbance at a Motel 6, where they found a dead woman in a bathtub. Her name was Afrikka Hardy, and she was nineteen years old. “They make an arrest of a guy named Darren Vann, and, as so often happens in these cases, he says, ‘You got me,’ ” Hargrove said. “Over several days, he takes police to abandoned buildings where they recover the bodies of six women, all of them strangled, just like the pattern we were seeing in the algorithm.”

Vann is going to trial in 2018 and is facing the death penalty. And it’s a victory for Hargrove, who’s convinced that data applied to crime statistics will uncover more serial killers. Hargrove, in fact, is convinced that there are at least two thousand active serial killers at large in America at the moment, according to his data. In particular, he’s concerned about a killer or killers that may have been active in the Atlanta area for at least forty years.

But are there really that many serial killers out there? That’s a tougher question than you might at first think. Keep in mind, algorithms only reflect the data we feed into them. In an analysis of an algorithm designed to help judges decide if criminals were more likely to reoffend, Pro Publica uncovered that the algorithm, despite efforts to remove race from the equation, consistently ranked non-white offenders more likely to commit future crimes, even if the white offender had a long record and the non-white offender had never committed a crime before. In fact, a teenage girl who idly picked up a bike from a driveway was seen as a greater risk than a hardened criminal.

And Hargrove admits that teasing out potential serial killers is largely a matter of guesswork, something he spent years building, working backwards from real cases. We are built to see patterns, as a species, and if you hunt for something specifically in the data, you may well find it even if it’s not there.

Still, as the New Yorker profile points out, MAP is a useful tool for understanding a crime we all fear, to take what’s vague and amorphous in our minds, defined more by fiction than fact, and put it in stark black and white, in numbers and points we can understand. And it helps us better understand the real form of crime in our communities. So even if Hargrove never finds another serial killer, he’s doing us all a public service.

(Via The New Yorker)