The video is grainy, but the incident seems painfully clear. A man lunges for an officer’s gun. The officer shoves him away. As the man falls back to the ground and lies there, the officer opens fire—first two rounds, then a third.

It’s over in six seconds.

The officer, a Special Forces veteran, lowers his sidearm. The camera swoops in for questioning.

“He was attacking you from the ground?” an interrogator asks.

“No, he was not on the ground when he started to make his move back to me,” the officer replies.

“He was still standing—he made his move; I shot him.”

“You did not shoot him while he was just lying there?”

“Not that I recall,” he answers.

The video, however, shows otherwise.

“That’s murder,” Duane Dieter says later.

Dieter is a trainer who’s coached everyone from police to military special forces to spooks from “the agency” (“We all know who that means,” he says). He was sharing a video from Close Quarters Defense, or CQD, his training compound on Maryland’s Eastern Shore about 90 miles from the nation’s capital. The shooting was a live-action simulation, featuring real guns with paintfilled bullets. And it illustrated a phenomenon disrupting police and cities across the nation: black men shot by police in what appear to be the most inexplicable of circumstances.

There are many examples. Charles Kinsey, the North Miami therapist shot after complying with an officer’s orders to lie down in the street; Samuel DuBose, the driver fatally shot during a traffic stop near the University of Cincinnati; Laquan McDonald, the teenager gunned down as he walked away from officers in Chicago. And the videos of the incidents that emerged contradict the sworn accounts given by the officers involved.

There’s more to the story than untruthful officers and racial bias, Dieter points out. There’s another factor that might answer the question of why these events unfolded the way they did, and one he believes is given too little attention: fear.

“The stress of what it feels like to be that close to death or that close to injury, it can create this feeling of chaos, and that can create a sense of rage,” he says.

Experts have long known that peril can blind people to their surroundings and allow them to mistake a common object, like a wallet, for a weapon, or a hesitation as a sign of an impending attack. They also know that such fears can be fueled by racial bias, something studies have found affects both whites and blacks.

Dieter says he’s found one solution, a way to teach de-escalation not through talk or classrooms, but by exposing officers to all levels of stress, from mundane tasks like giving directions to the most intense encounters imaginable, like ambushes and shootouts.

It’s a method he developed himself, and it’s very different from the programs rolled out by the Justice Department this summer and the training used by the nation’s roughly 18,000 police departments and other private training companies.

In the past 30 years, about 5,000 officers have gone through Dieter’s program, he says, the cost—$100 to $300 per officer per day for courses  that can last from three days to several weeks—is typically paid for by their departments or through federal grants.

There’s no way to tell how well Dieter’s program works, and there are questions, too. If an officer’s reactions are driven by bias—implicit associations, the fear of black men as dangerous—does stress-based training solve the problem? And if so, can it be an antidote to what one police chiefs’ group recently called the “historical mistreatment” of minorities by police? Is training even the answer?

“Training isn’t going to fix when we send people to do these things,” says Phillip Atiba Goff, president of the Center for Policing Equity at the John Jay College of Criminal Justice. “Treating me like a threat is different than treating me like someone who is threatened. You’re setting people up to fail if you train them that a situation that you’re in is always life and death.”

But it’s also, perhaps, a wake-up call. When shown videos of the simulations later, some officers who underwent the training—longtime veterans, training officers, and department leaders—broke down crying. Plunged into simulations mimicking the situations in North Miami or St. Paul or South Carolina, they often discover they react just like the officers they’ve condemned: they shot the unarmed suspects, too.

The entrance to CQD is unmarked, its classrooms, gun ranges, and tactical houses hidden by pastoral woods and farmland.
Since opening in 1981, tens of thousands of police officers and Special Forces operators have gone through the program, as well as women taking self-defense classes and executives assigned to offices in volatile countries overseas who want training on how to rescue family members taken hostage. Tiger Woods, after the death of his father in 2006, reportedly flew in by helicopter1 and spent time leaping walls and kicking down doors alongside Special Forces operators. 

Commenters in anonymous forums online have hailed Dieter’s techniques and derided them. Although he’d trained Navy Special Forces since 1989, Dieter lost that contract in 2011 when the Navy Special Warfare group made a controversial switch to techniques focused on mixed-martial arts, a move that senior naval officers say was a nod to the evolving nature of combat techniques.

With the ambush-style shootings of two police officers in Iowa on November 2, the number of officers killed this year rose to 115, up from 100 at the same time in 2015, a year that ultimately saw 123 officer deaths in the line of duty.

Similar ambushes of cops in Dallas and Baton Rouge in July and in Brooklyn at the end of 2014 have fueled fear in law enforcement circles of a violent backlash against police, perhaps driven by tense community relations and accusations of bias.

Line-of-duty deaths, however, remain at historic lows: 41 officers died last year as a result of felonies, such as shootings or assaults, according to FBI data2. Another 45 died in accidents. Statistically speaking, the job is far less perilous than, say, logging, which regularly tops the Labor Department’s list of most dangerous jobs in America.

Police academies and private training centers, which officers often attend on the taxpayers’ dime,  commonly urge cops to open fire if they ever feel imperiled. De-escalation may be a virtuous goal, but hesitation can kill—and academies reinforce the message by showing video after video of officers getting maimed or killed.

“All of that can unintentionally confuse the message,” says Seth Stoughton, a former police officer and assistant professor at the University of South Carolina School of Law. “If instructors are talking about accurately assessing risk, but everything around the officer is telling them, ‘Everything is always really risky,’ that can send a mixed message.”

Dieter screened a similar slate of videos during a visit this spring. It showed clips of officers getting stabbed or shot to death, of officers beating civilians, plus live demonstrations by trainers of how quickly things can go wrong. Dieter said his point, however, was not to emphasize policing’s inherent risks but to illustrate the countless ways cops’ training can fail and also ignites explosions of violence.

Dieter points to two incidents that allegedly began with physical altercations not caught on video: a white California Highway Patrol officer who was filmed punching a black woman at least 10 times as he straddled her on the side of a highway in July 2014, and a white South Carolina cop recorded gunning down a black man as he ran away in April 2015. (The first officer was cleared, the second, Michael Slager, is on trial for murder.)

“He remembers what the person did [before the altercation], not what he’s doing now,” Dieter says. “We have hundreds and hundreds of hours of this craziness.”

This is where courses like “The Bulletproof Warrior,”3 the course Officer Jeronimo Yanez who gunned down unarmed Philando astile during a traffic stop in Minnesota this summer, attended in 2014. These courses so often get the idea of being a “warrior” wrong, he says. The focus shouldn’t be on the capacity for violence or the idea of battle. It should be about restraint: “the effort it takes to control yourself,” Dieter says, as well as confidence in your own ability to protect yourself and that there’s no need to panic, one of the results his course aims to cultivate.

Dieter sees his Close Quarters Defense program as a kind of fix to the warfighter mentality. His main training tool is a simple black hood, kind of a cross between a windsock and surplus gear from Guantanamo. One hangs in each training room in the main building; the rooms are outfitted to look like city streets, back alleys, black-walled interrogation chambers, and rudimentary wood- framed homes. A trainee stands near the middle of the room, and the hood is lowered to his head. When the simulations begin— lights set low, speakers rumbling—the hood is suddenly yanked away, and the trainee is confronted by any number of men standing in the room, near or far, armed or not, asking for directions or yelling obscenities. It is up to the trainee to figure out how to react.

"We have to ask the community, how do you want to be policed, and incorporate that into the protocols of policing." Khalil Muhammad, Professor of History, Race, and Public Policy at Harvard University's John F. Kennedy School of Government

Most people flunk their first time through, many spectacularly. Grizzled Special Forces operators and rocksteady officers shoot civilians or forget a straightforward goal like “escape,” instead panicking and tackling a partner or the nearest “suspect.” Some even try to hide by cowering behind an instructor.

But the experience doesn’t start with peak stress. Dieter puts students through mundane tasks, like giving directions or helping a girl with a twisted ankle, before ratcheting up the intensity over the course of a program that can last a few days or a whole summer. The ultimate goal is to be able to recognize a threat in as little as 0.2 seconds and let that guide all the actions that follow.

“The idea is, when they get out to their workplace, they don’t think of anything else but what they see: if it’s a threat, it’s a threat; if it’s not, it’s not; and if it’s not a threat one second, the next it can be, and vice versa,” Dieter says. “That’s why we’re trying to build the tactical officer, so he’s sharp and understands what’s best for protecting him and the suspect and the community.”

The officer at a South Carolina gas station who asked a man filling his car to retrieve his license then shot him as he reached into the vehicle and the case of 12-year-old Tamir Rice who was killed by a police officer two seconds after arriving at the scene show how the police officers’ actions can be sparked perhaps by racial bias, fed by an inability to distinguish an actual threat, then inflamed by panic and irrational fear.

It’s still no excuse: “The military or police or any person authorized to enforce the law should be held to a higher standard of discipline,” Dieter says. “Other people can get killed by acting wrongly.”

“Of course stress alters perception,” says professor Maria Haberfeld, a former Israeli police officer and codirector of the NYPD Police Studies Program at the John Jay College of Criminal Justice. “Your frame of mind is completely different.”

And that stress can go into overdrive when what you’ve been assured should subdue a suspect—a Taser shot, a punch, or a blast of pepper spray—doesn’t work the way the coaches promised during training.

“If you say, ʻOK, I know this weapon will work most of the time, and it will take real special circumstances for it to not work,’ and then it doesn’t, you think, ʻThis must be those very special circumstances! This guy must be very amped up or especially dangerous,’” says Jonathan Wender, a former cop and a sociologist at the University of Washington who studied police behavior for the Defense Department. “Panic produces dysfunctional responses.”

Add implicit bias, and the scenario quickly deteriorates. Whites and blacks alike more quickly associate African-Americans with risk or danger, studies have found. Officers also more readily stop African-American men and more quickly resort to violence. And that bias becomes even more pronounced under pressure when the mind relies on evolutionary shorthand (stereotyping) to make snap decisions.

Add sleep deprivation, as common to cops working 12-hour shifts as heartburn and bad precinct coffee, and that can make the brain even more dependent on its neural shortcuts.

“Stereotypes are a cognitive mechanism to deal with a lot of information,” says Itiel Dror, a cognitive scientist at University College London who has advised police departments and the Air Force on how to prevent errors. “You do want police officers to learn from experience—the problem is that it can cause bias.” The implicit bias training programs rolled out this year by the Justice Department aims to attack this issue.

Dieter says his training, which forces cops to focus on threats alone, gets around bias entirely.

But no matter the approach, training by itself isn’t sufficient.

“We have to ask the community, how do you want to be policed, and incorporate that into the protocols of policing,” says Khalil Muhammad, a professor of history, race, and public policy at Harvard University’s John F. Kennedy School of Government. “If we don’t, we’re going to see this strange universe of reforms producing oppression and failure.”

Prior attempts at reform, like instituting mandatory minimums for certain crimes, were in part an effort to make the justice system more equal by limiting the potential influence of bias in sentencing decisions. But black communities ultimately bore the brunt of the result: mass incarceration.

“If we train them better, if we train them not to see race, then they’ll do a better job, the harvest of all that is in fact mass incarceration,” Muhammad says.

Bias training, use-of-force training, teaching restraint, and de-escalation—all that is positive, he adds. But it doesn’t necessarily reconcile with the “perspectives of people in the communities that they’re going to put that new training on—unless there’s more collaboration to see the humanity of the people first and foremost.”