Tasers are less reliable than their maker has claimed. The results can be deadly

Phil Grenon is exactly the kind of person the Taser was designed to save.

At 76, he'd long struggled with mental illness, and his condition had recently deteriorated. One day, he'd started shouting threats to his neighbors through the walls of his Burlington, Vermont, apartment, and they called the police.

Grenon had no history of violence — he was a devoted father and grandfather — but he'd lately been having paranoid delusions. So when two patrol officers showed up at his door early on the evening of March 21, 2016, Grenon confronted them with a knife in each hand.

The Taser was created for precisely this scenario: when police need to protect themselves but don't need lethal force. The Taser employs electricity to lock up a person's muscles for a few seconds, long enough for an officer to disarm and handcuff a suspect, usually without inflicting severe injury. In the past two decades, Tasers have become a ubiquitous law-enforcement tool. Most patrol officers in the United States carry them, and every year tens of thousands of Americans are shot with them.

The weapons are produced by Axon Enterprise Inc., which has a monopoly on the American market. The company has long promoted Tasers to police as a reliable and effective alternative to guns. In fact, Axon's slogan is "Protect life," and the company keeps a running tally on its website of the hundreds of thousands of lives it says it's saved. Over the years, Axon has claimed that Tasers are between 80 and 97 percent effective at subduing a suspect in the field.

But a year-long investigation by APM Reports shows that police rate Tasers as considerably less effective. Data from some of the largest police departments in the nation reveals that officers rate their Tasers as effective as little as 55 percent of the time, or just a little better than a coin flip. When Tasers fail to subdue someone, the results can be life-threatening — for police, and especially for the public.

APM Reports found more than 250 fatal police shootings nationwide between 2015 and 2017 that occurred after a Taser failed to incapacitate a suspect. In 106 of them, the suspect became more violent after receiving the electrical shock, according to a review of case files and media reports, suggesting the Taser may have made a bad situation worse.

APM Reports also found evidence that two of Axon's newer models may be less effective than older ones. Data from three of the nation's largest cities — New York, Los Angeles and Houston — showed that officers rated newer-model Tasers as less effective than older ones.

It's not clear why the newer models were rated as less effective, though two modifications were noteworthy.

Beginning in 2009, Axon reduced the power of its weapons, a change the company said would be safer for suspects. In the Taser model that Houston is using, Axon also altered the angle at which darts leave the weapon. The change meant officers needed to be farther from suspects for the weapon to work reliably — a tough requirement, because data from some cities shows police most often fire Tasers within 7 feet of a suspect.

All told, the company has sold more than 600,000 of the Taser models that police rate as less effective than older versions. Most of them likely remain in circulation.

Many police officers, and even some police chiefs, seem unaware of how often Tasers fail to subdue suspects, and most departments spend little time investigating the reasons why. The only public hints are often a little-noticed phrase that appears again and again in news stories about fatal police shootings across the country: "The Taser failed."

But at least one big-city department already knew. Since 2015, the Los Angeles Police Department's own data showed that its Tasers were less effective than the previous model, subduing suspects little more than half the time. Yet the LAPD officials neglected to investigate the problem — and then they bought thousands more Tasers.

For more than three years, the LAPD has required virtually all its patrol officers to carry those newer Tasers and use them in volatile, life-threatening encounters, even though its officers were consistently giving them lower marks for effectiveness.

Axon canceled a scheduled interview with APM Reports, but in a written response, the company raised concerns about the accuracy of police department databases tracking the effectiveness of its Tasers.

"[Tasers] are the most studied less lethal tool on an officer's belt," the Axon statement read. "These studies, along with nearly 4 million field deployments over 25 years, establish they are the most safe and effective less-lethal use of force tool available to law enforcement."

Nevertheless, Axon — a publicly traded company — has taken in hundreds of millions in taxpayer dollars selling weapons to protect life that work considerably less often than the company has claimed.

'Put down that knife'

When Phil Grenon came to his door holding two knives that evening in March 2016, the two Burlington officers, David Bowers and Durwin Ellerman, backed up. Ellerman pulled his gun. Bowers drew a Taser. For two tense minutes — recorded on the officers' body cameras — Grenon stood silently while the officers begged him to drop the knives. Then he finally spoke. "I'm a lawyer," he said. "I'm a psychiatrist."

"Well, tell me more about that, but put down that knife," Bowers replied calmly, his Taser still trained on Grenon.

"I just did, you stupid son of a bitch," Grenon screamed back. "Leave me alone!"

Grenon was not a doctor or a lawyer. After graduating from the University of Vermont in 1967, he'd thought about law school and even took the LSAT, but he ended up getting a master's degree in education. Grenon taught at the community college level before his mental illness made that impossible.

Mental illness ran in his family. His mother was committed to a state mental hospital when he was about 9, and he and his four brothers were sent to an orphanage. There, he would later tell friends, he was physically and sexually abused.

After Phil stopped working, Sally, his wife at the time, supported the family working as a nurse while Phil became a stay-at-home dad to their daughter, Niki.

When Sally and Phil divorced in 1998, Niki helped him find a subsidized apartment in an old brick building in Burlington, where he'd lived ever since. Phil found friends and made a life for himself. He liked to talk politics over coffee with a high school buddy who had become his state senator. He wrote letters to the editor of the Burlington Free Press, defending the dignity of the mentally ill.

"Many, many have been unable to fulfill their dreams because they have been stricken with this dreaded and misunderstood disease," he wrote in 1999. "For the injustices of a painful bio-chemical imbalance in the brain, and a strong social rejection, one would rather be dead or have cancer."

Grenon could be gruff, even rude, to the neighbors he didn't like. Everybody knew he had mental illness. But around the winter of 2015, it got worse. He was convinced someone was out to get him. He scribbled a rambling letter predicting "vagrant[s] ... dressed in police uniforms" would come to his door. He vowed to "kill them before they kill me." But even as he increasingly struggled with paranoia, he still called Niki and his grandkids every Sunday.

Standing in his doorway, Grenon never said another word to the officers. As Grenon stepped forward to slam the door, Bowers squeezed the trigger of his Taser.

For the weapon to work, a lot has to go right. First, an officer must hit the target. Tasers simultaneously shoot two barbed darts attached to thin, electrified wires. Both darts have to hit the target to deliver a debilitating jolt of electricity. Each dart must strike within an inch or so of the skin — or better yet, penetrate it — to create a complete electrical circuit. If someone is wearing a heavy coat or loose clothing, the electricity may not arc into the body enough to lock up muscles.

There's also a chance that a person convulsing under the Taser's power will manage to remove one of the darts — or break the wires that lead back to the Taser, ending the flow of electricity.

Where the darts hit matters, too. They must be at least a foot apart from each other when they hit someone for the electricity to flow through enough muscle to reliably incapacitate the person.

To Bowers and Ellerman, it looked like at least one dart missed or got snagged on Grenon's door when it closed. Whatever the reason, the weapon had no effect. It wouldn't be the last time a Taser failed to subdue him that night.

'The Steve Jobs of law enforcement'

Some 400,000 American patrol officers carry Tasers on their hips, and the man who put them there is Rick Smith.

The 48-year-old founder and CEO of Axon has built his company into one of the top suppliers of technology to law enforcement. Axon is outfitting police across the country with body cameras, surveillance drones and virtual-reality simulators. And Smith even has a vision of using artificial intelligence to write police reports. He's jokingly referred to himself as the "Steve Jobs of law enforcement."

But the foundation of the company has always been the Taser. The weapon was invented in the early 1970s by Jack Cover, a physicist living in Southern California. Cover named his creation "Taser" as a loose acronym for "Tom Swift and His Electric Rifle," a young adult science fiction novel he'd read as a boy. He sold his patents to a company named Tasertron. Early Tasers were too big to fit on a cop's belt, and regulatory barriers made them difficult to sell to consumers.

In September 1993, Smith, a 23-year-old fresh out of business school, founded the company that would become Axon. He was particularly interested in electrical weapons. He contacted Cover, who, as luck would have it, had been nurturing an idea for a new kind of Taser, one that used compressed nitrogen gas instead of gunpowder to propel its darts.

Smith and Cover built what they called the Air Taser, and Smith's company began selling it. But the fledgling business nearly went bankrupt because the patents held by its competitor, Tasertron, prevented it from selling weapons to U.S. police departments. Smith held on thanks to a cash infusion from his father.

Tasertron's patent expired in 1998, which allowed Smith to sell his weapons to police. But to convince cops to make the switch, he needed to solve a big problem: His weapons weren't powerful enough. With the company's last million dollars, he "dialed up" the electrical charge in every Taser pulse and crammed more muscle-contracting pulses into every second.

He also modified the weapon so it looked more like a gun and could fit neatly into a holster. The resulting models, the M26 and its smaller successor, the X26, were hot sellers with police departments.

Smith changed the name of the company to Taser International and took it public. By 2003, it was dominating the market and bought up what was left of Tasertron for just $1 million. From then on, Smith's Taser International had the market to itself. Last year, the company now called Axon reported $420 million in sales, up 22 percent. The company took in $253 million of that from Tasers.

How often they work

Tasers are popular with police departments because they can prevent shootings while also protecting officers. Unlike a night stick, a Taser can be used at a safe distance, and unlike pepper spray, there's no blowback. Every year, tens of thousands of people, some of whom might have otherwise been shot by the police, are taken into custody without lasting injury thanks to a Taser.

Axon has made varying claims over the years about how reliably its Tasers incapacitate suspects. In earnings calls and marketing materials, company officials have asserted that Tasers are effective 86 percent, 94 percent, and 97 percent of the time in the field. The company has even claimed success rates of 99 or 100 percent in testing and demonstrations. Axon no longer makes such precise assertions of effectiveness in its marketing materials. Still, as recently as 2015, Smith said in an interview that the weapons subdued people "80 to 95 percent" of the time in the field.

But the APM Reports investigation found that police rate Tasers as less effective at bringing people down than the company has claimed.

APM Reports sought data on Taser usage from police departments in the nation's 20 largest cities and received usable data from 12 of them. The departments show a wide range of effectiveness, in part due to varying definitions and measures.

For instance, the LAPD counted every trigger pull as a Taser usage. Other departments, such as the New York Police Department, only track each officer's Taser, not trigger pulls. Consider an incident in which an officer shocks a suspect with a Taser three times, and the first two attempts fail to subdue the suspect, but the third one does. In L.A., the effectiveness rate would be 33 percent (three trigger pulls, one incapacitation); in New York, the effectiveness rate for the same incident would be 100 percent (one officer with a Taser, one incapacitation).

The department with the highest rated effectiveness — El Paso, Texas — corresponds to the lowest end of Axon's claims: 80 percent. But 7 of the 12 departments had effectiveness rates below 70 percent. In Los Angeles (57 percent) and Indianapolis (55 percent), a Taser failed to subdue someone at least four out of every 10 times.

In its statement to APM Reports, Axon said that data from police departments doesn't accurately reflect Taser effectiveness because it may not include instances when a suspect was subdued after an officer merely displayed or threatened to fire a Taser. The company argues that just the sight of the weapon can be a significant deterrent to a suspect, incidents that should count as effective use.

In most cases, the data that APM Reports obtained from those 12 major police departments included only instances in which Tasers were fired.

And none of the departments — over years of engagements and more than 30,000 uses — saw effectiveness rates near 95 percent, the top of the range claimed by Smith in 2015.

Each time a Taser fails to incapacitate someone, lives are potentially at risk, as the company has acknowledged. "Quality is crucial in our devices, because when an officer needs our device to work, it's got to work every time or somebody's going to get injured or killed," the company's then-vice president of training and education, Rick Guilbault, said in a 2011 marketing video.

And Rick Smith knows it, too. In October, he took the stage at Axon's 25th anniversary party, held at a House of Blues in Orlando. The party coincided with the International Association of Chiefs of Police's annual conference, and the place was packed with law enforcement officers. They gave him a rousing cheer. Smith told the officers that he understood how high the stakes are when police use a Taser.

"We know as our technology has gotten better you've come to rely on it more and more, and it's really painful for you and for us when it doesn't work, when it doesn't get the job done," he said. "And that's what keeps us up at night."

The standoff

Burlington Police Chief Brandon del Pozo was on the shooting range of the Vermont Police Academy, 60 miles south of Burlington, when he got the call that a mentally ill man armed with knives was in a standoff with his officers.

Del Pozo was 41 years old at the time and only seven months into his job. Ivy League educated and media savvy, he came to Burlington after 18 years at the NYPD.

Del Pozo had commanded two precincts in New York and seen his share of police shootings. He could tell the situation with Grenon could end badly, and he quickly drove to the scene to try to save Grenon's life.

"I was happy to see when I got there that the scene was under control," del Pozo said. "They were taking their time, and they were trying to get him to talk so they could negotiate."

Grenon was alone in his apartment. He couldn't hurt anyone except possibly himself.

The officers used a new tactic that del Pozo had brought to Burlington from the NYPD. They tied a rope around the doorknob and anchored it, so Grenon couldn't burst into the hallway and provoke the cops into shooting him.

The door wasn't going to open until the police decided to open it. They had time on their side, so they waited.

They tried to reach his psychiatrist, but she was out of the country. They called his phone more than a dozen times and left messages offering to help him.

"Your daughter is worried about you," Officer Mike Henry said in one of the voicemails. "We can't leave until we talk to you."

Grenon remained silent. It wasn't clear whether he was still alive, and the cops wanted to see what was going on in the apartment.

The police department didn't own a drill or a saw, so del Pozo went home and got his tools. They cut holes in the walls and inserted a camera.

"What we saw was nothing," del Pozo said. "We just saw empty rooms."

Almost four hours after Grenon had slammed the door, the officers entered the apartment. Grenon was hiding in the shower, still holding the knives. He didn't attack the officers. Didn't say a word. Didn't even move.

Plan A was to smoke him out with a device called a PepperBall, which is a glorified paintball gun that shoots rubbery plastic balls filled with a chemical irritant similar to pepper spray. Officer Henry crept up to the door of the bathroom and fired eight capsules. They struck the shower wall above Grenon's head. As they burst, the noxious powder inside rained down. But he only let out one tiny cough.

As the dust spread through the apartment, it seemed to affect everybody but Grenon. All the officers in the room were wracked with coughing fits. "Note to self," Sgt. James Trieb said to no one in particular, "never use Pepperball again inside of close quarters."

Trieb and del Pozo decided it was time to try the Taser again.

The plan was to disarm Grenon with a Taser and pin him to the wall with a plastic shield, allowing officers to put him in handcuffs and take him to the hospital. The cops lined up at the bathroom door. Officer Ellerman stood at the front of the line. In one hand, he held a shield. In the other, a Taser.

What goes wrong?

When a Taser doesn't bring down a suspect, it's often hard to know exactly why it failed.

That's partly because police departments typically don't investigate the cause. But it's also because there are so many factors that can influence how well a Taser performs, from where the darts hit, to what the suspect was wearing. Axon has long acknowledged two key variables in this complex equation: power level and distance.

Since its early days, the company has understood the relationship between the level of electricity coursing through a Taser's wires and its ability to incapacitate a suspect.

Increasing the electrical charge made the M26 and X26 big sellers and popular with police. But then, in 2009, the company changed course. Axon reduced the power in its next generation of Tasers, including popular models called the X2 and the X26P. They had roughly half the electrical charge of the X26.

The company said in its marketing materials that the X2 and X26P would have a "significantly improved safety margin." The decision to reduce power came when the company was simultaneously fighting dozens of product liability lawsuits, alleging Tasers caused death or serious injury. Axon also added new warnings to its products as part of a more cautious "risk management" strategy.

The lawsuits peaked in 2011, when the company was fighting 55 of them. After releasing the X2 in 2011 and the X26P in 2013, Axon's legal exposure has steadily declined. As of this spring, the company is facing just eight active product liability suits.

But two new lawsuits have recently emerged, claiming that the lower-powered Tasers don't put out enough juice to protect police.

There's one in New Orleans, from the family of an officer who was shot and killed after his lower-powered X26P Taser was allegedly ineffective. And there's another from a Houston police officer who says she was injured in a fight after her Taser X2 failed to subdue a suspect. The company has vigorously contested the allegations in the suits.

Axon claims its tests prove that the X2 and X26P are just as effective as their more powerful predecessors, but there is just one publicly available study supporting that claim.

Axon medical director Dr. Jeff Ho presented a paper comparing the effectiveness of the different models in 2012. The study found the newer weapons were just as effective as the old ones at preventing volunteers from completing a simulated attack with a rubber knife. But that study involved only four people, who each received just two Taser shocks.

Ho, who's a part-time sheriff's deputy in Meeker County, Minnesota, and an emergency room physician at Hennepin Health, a hospital based in Minneapolis, did conduct a few other studies comparing the effectiveness of Taser models, according to a report he prepared in response to the Houston lawsuit. The report says those studies also showed the lower-powered Tasers were just as good as the higher-powered ones. In all, they involved 150 test subjects.

Axon's marketing materials have claimed the newer models were actually "more effective" than their predecessors, though Ho's findings did not support this claim. Rather, Ho found that the different models "have very similar incapacitation characteristics when compared to each other."

Data from some of the largest police departments in the U.S. conflicts with Dr. Ho's conclusion that the X2 and X26P work just as well as their more powerful predecessors.

Police in New York, Los Angeles and Houston reported lower levels of effectiveness when using the X2 or X26P. While each city tracks effectiveness differently, the declines in effectiveness in New York, L.A. and Houston were remarkably similar. In each city, the lower-powered weapons were 6 to 7 percentage points less effective than previous models.

Given the size of the datasets, each city saw a statistically significant correlation between the lower-powered Tasers and the decline in effectiveness. Combined, the datasets for the three cities covered nearly 14,900 Taser uses.

APM Reports conducted an analysis of the data to determine what other factors — such as offense type or the rank of the officer involved — might account for the drop in effectiveness. Even controlling for those other potential factors, the analysis found that the model of Taser remained an important predictor of effectiveness.

Reporters also collected data from other large U.S. police departments. Other cities didn't have usable data because they either changed methodologies for tracking effectiveness or used Tasers too infrequently to have a large enough sample size. But none of the cities reviewed saw effectiveness increase after switching to the X2 or X26P.

J. Patrick Reilly, an electrical engineer who spent most of his career doing scientific research at Johns Hopkins University's Applied Physics Laboratory and has studied Tasers, said that reducing the power could have made the weapons less effective. "I think it's a reasonable bet that as you reduce this charge, you were going to reduce the probability of making the subject fall down," he said.

There's another key factor in whether a Taser is likely to make someone fall down — distance.

Tasers are typically designed to work best at a specific range from the target. If officers are too far away, they'll likely miss the shot. But if officers are too close, the Taser is less likely to halt someone. That's because the darts, when they hit the target, won't be far enough apart to lock up someone's muscles.

Over 25 years, Axon has changed its recommended spread between the darts. Early on, the company said that the darts needed to hit only 4 inches apart to incapacitate someone. It later asserted that the darts needed to be 9 to 12 inches apart. And it now recommends at least a 12-inch spread between the darts for electricity to flow through enough muscle to reliably bring someone down.

When a Taser is fired, the darts spread apart from each other as they fly through the air toward a suspect. That means the range at which a Taser can effectively be used largely depends on how quickly those darts spread apart and how long it takes them to reach the desired 12-inch spread (see graphic).

Axon's earlier models were designed to work best at longer range. Most of its models, dating to 1994, had darts spread apart such that they'd be reliably effective at 7 feet or more.

When the company released the Taser X2 in 2011, it narrowed the angle at which the darts spread apart. That meant officers had to be even farther away — at least 9 feet — for the X2 to reliably bring someone down.

However, none of those distances apparently reflect the reality on the street, where the violent encounters that send officers reaching for their Tasers often happen much closer.

APM Reports obtained databases from two large departments — New York and Fort Worth — that track the distances at which officers fired their Tasers. Both departments found that about 75 percent of Taser discharges happen at 7 feet or less.

The data suggests the possibility that virtually all Tasers currently in circulation are typically not used at the ranges where they are most effective.

And the re-designed Taser X2 may have exacerbated that potential problem, even though it was promoted as being more effective than previous models.

The best Taser yet

In October 2018, Axon released its first new Taser in five years, claiming it would be the most effective ever.

In his speech at the anniversary party in Orlando, Smith promised the new Taser 7 would be "stronger, faster and smarter than any that has come before it."

Axon says there's nothing wrong with the lower-powered models. But the company made a number of alterations in the new Taser 7, designed to address longstanding problems that police experienced with earlier models — including better darts and improved laser sights.

At a training session outside Fort Worth last year, the first question on the mind of Sgt. Karl Johnson was: What about "the power issues?"

Johnson, a Taser instructor from Saginaw, Texas, told an Axon executive that the last time the company came out with a new weapon, "the volume got turned down on the effectiveness of the device no matter where the probes were deployed."

Johnson didn't mention it, but a few years earlier an officer from his department had shot and killed a man named Michael Dale Brown after a Taser X26P failed to subdue him.

Technically, Axon didn't turn up the electrical output of the Taser 7, but it focused the energy in shorter, more concentrated and more frequent bursts. The company claims delivering electricity in that condensed manner will make the device more effective.

But perhaps the most dramatic change is that the Taser 7 is the first device Axon has ever designed to be reliably effective when a police officer is face-to-face with a suspect, as close as 4 feet.

To make the weapon work better at such close range, Axon had to widen the angle at which the darts spread apart when they're fired. The Taser 7 fires darts at a 12-degree angle, an increase from seven or eight degrees in previous models.

That 12-degree angle is not a new idea, however. It is the original angle used in Tasers dating to the 1970s and made by Tasertron through the early 2000s, according to James McNulty, who was an executive with Tasertron. The realization that a 12-degree angle might work better in real-life situations isn't new, either.

In 2000, a Canadian police sergeant published a study of Taser effectiveness and wrote, "Based upon the fact that the wider the dart spread, the better the takedown, Tasertron's 12-degree separation would have a better Taser effect over a larger body surface especially within the 2.5-12 foot range where most Taser applications take place."

Axon was certainly aware of that study. It is included in an index of Taser research the company touts on its website. Yet it never used a 12-degree angle in its weapons until just last year.

The company continues to sell the older models with a narrower angle, and hundreds of thousands of them remain in circulation. Axon didn't answer questions about why it hasn't produced cartridges with darts that fire at a 12-degree angle for the X2 and X26P, which would allow those weapons to be more effective at closer distances.

In its statement to APM Reports, Axon said that it's constantly trying to improve its weapons based on feedback from officers. "The TASER 7 is the result of Axon's commitment to develop new, innovative products and improve its existing products," the company wrote. "Some of those developments sought to address common reasons why a [Taser] may not cause [muscle incapacitation]."

'We did not expect him to move that fast'

Hours into the standoff, Phil Grenon was still hiding in his shower, and the police were preparing to storm the bathroom. The key to their new plan was the Taser. They would stun him, and he'd drop the knives. Then they could pin him down with shields, put him in handcuffs and get him safely to the hospital.

Sgt. Trieb took a broom he found in the apartment, reached over Officer Ellerman's shoulder with it and swept back the shower curtain.

Again, Grenon said nothing. He stood there clutching his knives and turned his body toward the officers. Ellerman pulled the trigger on his Taser.

"The plan stops working the moment they fire the Taser," Chief del Pozo later explained.

The darts appeared to hit Grenon this time, and he let out a scream that could be heard on the street below.

Grenon looked down at his sweater, where the Taser darts had lodged. And then, to the astonishment of the officers watching, he simply brushed them away. "He pulled the fucking barbs out of himself," Ellerman later told investigators. As soon as Grenon removed one of the barbed darts, he broke the circuit, and electricity stopped flowing.

Pulling out the darts of a Taser is something Axon co-founders Rick and Tom Smith have portrayed in the past as unlikely because the person being shocked is temporarily paralyzed.

"What's to stop a perpetrator from breaking those wires off?" cable TV host Leo Laporte asked Rick Smith in 2002.

"The 50,000 volts that's going through his body," Smith replied with a grin.

"Have you ever seen a test subject able to yank these [darts] out?" Nightline co-anchor Bill Weir asked Tom Smith in 2011.

"No," Smith answered quickly. "You can't control motor function."

But the Axon training materials the Burlington Police Department used in 2016 did mention the possibility that someone being tased could retain muscle control, "particularly in arms and legs." That was clearly the case with Grenon.

"The Tasers hurt him enough to make him really angry and to aggravate his episode, and yet did not hurt him enough to incapacitate him," del Pozo said.

Grenon was no longer cowering in silence. With a howl, he stepped out of the shower, knives swinging. The officers backed up into the bedroom, and Grenon chased after them.

"We did not expect him to move that fast," Ellerman said.

Another officer, Chase Vivori, tried his Taser too. "It looked like a good hit, I thought would have had an effect, but it didn't," Vivori told investigators.

There was hardly any time to tell whether Vivori's Taser would be any more effective than the others, because a moment later, Officer Bowers fired six bullets from his G22 handgun in the space of about two seconds. Grenon fell to the floor, bullet holes in his chest, thigh, groin and abdomen. He died soon after.

He also had six smaller marks on his body, the kind Tasers leave behind.

"By the time we were done with this encounter, unfortunately, the room was just a crisscross mess of Taser wires," del Pozo said.

The shootings

Grenon's story is like hundreds of others all over the country. Police end up shooting someone after their Tasers prove ineffective. APM Reports found more than 250 similar cases over just a three-year period.

In a wooded area in California, a suspected burglar named Joseph Melvin was hiding from the cops. When Officer Michael Dietrick tried to arrest him, Melvin fought back. So Dietrick fired his Taser.

"It didn't deter or slow him down in the fight at all," Dietrick told investigators. "If anything, I feel like it just ramped it up."

Melvin got ahold of Dietrick's flashlight and started beating him over the head with it. Dietrick drew his gun and killed Melvin.

In a suburb of Miami, a mentally ill man named Cornelius Brown walked into a convenience store swinging a broom handle. When an officer confronted him, he ran away. Other cops showed up and began firing Tasers at him. Five officers discharged Tasers that night. None were effective, and two officers finally shot and killed Brown.

And in a suburban housing development north of Seattle, a veteran suffering from PTSD and drug abuse called 911. Juan Salinas said he wanted to "kill cops," but really he wanted them to kill him. When officers arrived, Salinas was stalking the streets, covered in blood. In one hand he held a knife. The other was wrapped in an American flag. An officer fired a Taser, but it wasn't effective, because one dart either missed or got snagged in the flag. Salinas finally ran toward the cops, and one of them shot him in the abdomen.

The stories all follow the same disturbing pattern. They start with police using a Taser. It's ineffective. Officers resort to firearms, and someone ends up dead. In more than 100 cases, a suspect appeared to become more aggressive after a Taser failed to bring him or her down.

Reading the investigative reports and news coverage, it's hard to escape a chilling conclusion: Had the Tasers performed the way the police hoped, these people would probably still be alive.

In some cases, it's obvious why the Taser didn't work, because one or both of the electrified darts missed their target.

But with many of the shootings, it's much murkier. The darts hit. They just don't do much. And the investigators spend little time trying to figure out why. They tend to focus on the bullets that proved fatal, not the Tasers that proved ineffective. The use of a Taser is usually treated only as evidence that officers did everything they could to avoid deadly force.

The Phil Grenon shooting was no different. The Vermont State Police didn't investigate why the Taser failed during its review of the incident, records show.

Less than two months after the shooting, Chittenden County State's Attorney T.J. Donovan (now Vermont's Attorney General) ruled the shooting justified, and the Burlington Police Department released the videos recorded by the cameras the officers wore on their uniforms that night.

It wasn't until the next year, on the anniversary of Grenon's death, that his niece, Sarah Grenon, could bring herself to watch.

"I don't even know why I watched it," she said. "I guess just to maybe find out what went so wrong."

That's when she saw how close Grenon was to Ellerman when the officer fired the Taser. Ellerman was in the bathroom doorway. Grenon was in the shower. "They were face-to-face," she said.

Grenon appeared to be 3 to 4 feet away from Ellerman, based on measurements APM Reports conducted of his old apartment.

Sarah started researching Tasers. She discovered that the model the Burlington police were using, the X2, is reliably effective only at a distance of 9 feet or more. It would have been difficult to achieve that kind of distance in Grenon's tiny bathroom.

"If I know that, they should have known that," Sarah said. "Why didn't they know that?"

APM Reports obtained the Taser X2 training PowerPoint that Axon supplied to police departments such as Burlington in 2016. The training presentation states that people can sometimes fight through the shock of a Taser or pull the darts out of themselves, especially when using the X2 at close range.

But the 222-slide "X2 User Course" never explicitly states that officers shouldn't use the weapon at those ranges. It asserts that the "preferred range" of the weapon is "7 to 15 feet from target." On a slide titled "Deployment Distance Considerations," it states that using the X2 from zero to 7 feet away can result in greater accuracy, but less "muscle mass affected." The presentation advises officers firing at such close range to "split the belt line," meaning land one dart above the waist and one below the waist, which is exactly where Ellerman told investigators he aimed.

An autopsy confirmed Grenon's body had marks from Taser darts above and below his waist. Yet despite the officers following the Axon training for firing at close range, the only apparent effect the Tasers had on Grenon was to enrage him.

Axon's new Taser 7 is designed to improve performance at close range, but the change came too late for Phil Grenon.

A department that knew

The decline in Taser effectiveness is especially evident in Los Angeles. The LAPD was an early adopter of the Taser, and nearly every one of its patrol officers now carries one, though the department's own research has shown that Tasers are far less effective than the company has claimed.

In March 2016, the LAPD released a report showing a decline in effectiveness at the same time that officers started carrying the new X26P.

Three years later, the department has yet to investigate the reasons for the decline in Taser effectiveness.

After the 2016 report's release, the Los Angeles Times found that ineffective Tasers were a recurring element in a number of the city's police shootings. Then-Chief Charlie Beck went on local television to defend the weapons. He described Tasers as just one of several force options, all of which are crucial to officers, but not foolproof. "None of them are 100 percent effective, and I think that's important to note," Beck told KTLA5. "[W]e have to have realistic expectations."

Tasers were the most widely used weapon that year, outpacing chemical sprays, batons or bean bag shotguns.

The drop in effectiveness came while the LAPD was in the midst of a Taser buying spree. Between 2014 and 2015, the department purchased more than 3,100 units. And in 2015, officials ordered virtually every patrol officer to carry an X26P. By 2015, when officers began widely using those new X26P Tasers, the weapons were proving to be less reliable.

The consequences of those failures were, at times, deadly. There were 21 fatal police shootings by LAPD in 2015 and in at least five of those incidents LAPD officers had tried an X26P before resorting to a gun.

In April 2016, as the LAPD was deciding how to respond to questions about Taser effectiveness, then-Assistant Chief Michel Moore questioned the significance of the department's own stats. In an email obtained by APM Reports, he chalked up the inquiries to "a rumor" that the new X26P was less effective.

Internally, Moore, who's now the chief of police, called for additional research. But Moore's confidence in Tasers remained steadfast, internal correspondence shows, and he wanted more of them. After the LA Times editorial board chimed in the following week cautioning the department not to count on Tasers as a "magic solution" for reducing police shootings, Moore directed a staffer to "Please prepare a rebuttal to support the added devices."

In summer 2016, the department officials made a few changes meant to bolster Taser effectiveness: They purchased new cartridges with a range up to 25 feet and had longer barbs they hoped would more easily penetrate heavy clothing. The department later went on to revise some of its policies on when Tasers should be used, a change that officials say is responsible for a significant decline in the number of Taser uses by LAPD officers in the past year.

But a records request turned up no evidence of LAPD research on why its officers were rating the X26P Tasers as less reliable.

LAPD officials say the department did study why the department's overall effectiveness rate (about 57 percent) was so much lower than other major departments. The answer was that LAPD tracked Taser data in a more detailed way, counting every trigger pull as a Taser usage.

Not only did the LAPD choose not to investigate the decline in reliability, the department doubled down on the weapons. Less than four months after releasing its initial report, on June 24, 2016, the department agreed to buy 4,400 more Tasers.

The aftermath

Officer David Bowers was just 23 when he shot and killed Phil Grenon. He'd been with the Burlington Police Department less than two years.

When he pulled the trigger, he estimated Grenon was only 4 or 5 feet away from him, slashing at officers with a knife. Bowers was terrified, both for his life and for the other cops in the room. He opened fire, he told investigators, because he knew no one else was in the position to do it in time. The awful responsibility fell to him.

As Grenon lay dying on the floor, the chemical irritant from the Pepperballs the officers had used earlier still hung in the air. Bowers watched as his fellow officers turned over Grenon's body to give him first aid.

Bowers saw one of his bullet holes. All of a sudden, he couldn't breathe. He walked out of the apartment.

Bowers wasn't physically hurt, but the police chief sent him to the hospital, just to be safe.

He wanted to talk to his parents about what happened, but he figured he shouldn't go into the details with the investigation going on. He was worried they'd somehow be dragged into it. The only people he felt safe confiding in were his lawyer and his union rep.

The next night, he couldn't sleep. In the morning, he grabbed his phone and sent a text message to his ex-girlfriend. He didn't want to involve her, but there was something he couldn't get out of his head.

"It was very eerie to me how he just didn't say a word," he told the investigators.

There was something else that bothered him. He couldn't believe it had been so easy for Grenon to overcome the effects of the Taser.

Bowers had just gone through Taser training a few weeks earlier. He'd seen people get shocked, and it always seemed to work perfectly.

"Watching this guy being tased, and walking toward us swinging a knife at us, shocked me," he said.

Burlington police don't use Tasers often. It's the biggest municipal police department in Vermont, which isn't saying much. It has about 100 officers, and the year Grenon was shot, department records show only seven officers discharged their Tasers. Three of those were during the incident with Grenon.

None of the officers who fired Tasers that day had used the devices in the field during the previous six years — if ever.

Chief del Pozo had never used a Taser in the line of duty, either, though he'd carried one for much of his career as a supervisor in the New York Police Department. But his general impression before that day was that the devices were highly effective.

"I've learned a lot about Tasers since the Phil Grenon incident, some of which surprised me," del Pozo said.

He learned the X2 Tasers the department had bought at the end of 2015 put out less electricity than the ones the department had before. And he learned those Tasers fail to subdue suspects more often than he ever would have expected.

"The Taser is this complicated piece of machinery with electricity, and its success is contingent on a lot of different factors of human physiology and luck," del Pozo said. "It's the most complicated thing a cop has on his or her belt."

So, in the wake of the Phil Grenon shooting, the Burlington police department went looking for simpler solutions.

It spent about $250,000 to buy a big truck and outfit it with every piece of equipment that could possibly help in case of standoffs, mental health calls, and hostage negotiations: shields, power tools, lights and communication equipment.

There are a couple of items on the Emergency Response Vehicle del Pozo wished the department had back in 2016. One is called a Y-bar. It's an 8-foot-long steel pole with a semi-circle at one end, about the size of a man's chest. If the cops had one, del Pozo explained, they could have simply pinned Phil to the shower wall at a safe distance. That way, he couldn't have threatened the officers.

The rig also carries a couple old-fashioned chrome-plated fire extinguishers, filled with pressurized water.

"If you spray that at someone's face, they cannot advance toward you," del Pozo said. "They have to look away, or put their hand up in front of their eyes. That and a metal bar shaped like a Y can mean the difference between having to shoot someone or not."

There are no Tasers on the Emergency Response Vehicle, but Burlington police officers still carry them on their belts. Del Pozo says Tasers can be useful as a last alternative to using a gun, and he wants his officers to have as many options as possible. But the Phil Grenon shooting has changed the way he thinks about Tasers.

"Knowing what I know now, if all things are being equal, and there's a man with a knife in a bathroom down the street from this police headquarters, we would not make the same plan. We would not say the best way to end this after hours and hours, is to send in a team that will rely on a Taser," del Pozo said. "If you're using [a Taser] to conclude a stable situation, you better have a back-up plan, because there's a good chance it's not going to work."