Framed: Astonishing True Stories of Wrongful Convictions

By John Grisham & Jim McCloskey

11 min read

Why does it matter? Because every confession, expert witness, and conviction you trust may have been manufactured.

Most of us believe the system, battered and imperfect as it is, bends toward truth. That a confession means something happened. That a forensic expert saw what they claim to have seen. That by the time a jury delivers its verdict, the machinery of justice has done its honest work. Framed asks you to hold that belief carefully — and then watch it come apart, piece by piece, across ten real cases where the machinery didn't malfunction. It was aimed. Aimed at men who walked into police stations to help. At a Black janitor because the white ranger told investigators, bluntly, that one of them had to be the suspect and he elected the Black one. At soldiers celebrating a bachelor party forty-eight miles from where a man was shot. The interrogation, the expert witness, the smoking forensic detail — each was a tool not for finding guilt but for manufacturing it. And when the truth finally surfaced, the system's first instinct was to protect the verdict.

Innocent People Confess — and the Police Know How to Make Them

At 6:30 in the evening of July 7, 1997, a twenty-five-year-old former Boy Scout named Dan Williams walked into the Norfolk police building to answer some routine questions. He was cooperative, polite, and had nothing to hide — he had been asleep in his apartment next door when his neighbor Michelle Bosko was raped and murdered, and his wife could confirm it. What he did not know was that a detective had already decided he was the killer, based on nothing more than a nervous neighbor's muttered suspicion.

Over the next eleven hours, police told Williams he had failed a polygraph he had actually passed. They told him witnesses had placed him at the scene. They suggested maybe he'd been sleepwalking, or had blacked out, and that they were there to help him piece it together. Each time he laid his exhausted head on the table, they ordered him to lift it back up. By 3:00 a.m. he had started to question his own memory. By 4:51 a hardened interrogator named Glenn Ford arrived to finish the job. Williams later tried to describe the moment he broke: he knew what he was saying wasn't true, but he needed the questioning to stop, so he gave them the story they wanted. When the tape recorder finally switched on — eleven hours in — he confessed to a murder he didn't commit, including details the detectives had fed him across the night.

Williams confessed not because he was guilty but because the interrogation was designed to produce that outcome regardless of guilt. Virginia law permitted police to lie freely about polygraph results, witnesses, and evidence. It did not require them to record interrogations. And here is the inversion that reorders everything you thought you knew: between eighty and ninety percent of innocent people waive their right to remain silent, because they trust the police and want to prove their innocence. Guilty people are far more likely to ask for a lawyer. The system is built in a way that specifically disadvantages the innocent.

The physical evidence at Michelle Bosko's apartment pointed to a single assailant. Her killer's DNA was eventually matched to a violent criminal named Omar Ballard, who confessed and acted alone. None of that stopped the prosecution from holding seven sailors — all excluded by DNA — on capital murder charges. The confession had been obtained. That was enough.

Junk Science Wore a Lab Coat and the Jury Believed Every Word

Twelve people with no training in forensic science are asked to evaluate a forensic scientist. The only tool they have is whether the person sounds like they know what they're talking about. What the book reveals is that credentials are often a costume, and the science behind them is sometimes no science at all.

Consider what happened to Levon Brooks, a Mississippi bartender convicted of murdering a three-year-old girl he had never met, while working at a nightclub eight miles away. The prosecution's decisive witness was a Hattiesburg dentist named Michael West, who had somehow expanded his practice to include bite mark identification, ballistics, arson, tool marks, and wound patterns. West told the jury that tiny marks on the victim's wrist had been made by Brooks's teeth. His conclusion: 'indeed and without a doubt' — a phrase he coined himself, stepping around the standard requirement that experts testify within 'a reasonable degree of scientific certainty.' To sharpen the comparison, West had visited Brooks in jail before trial and pressed a wad of Silly Putty — the children's toy — against his teeth, because standard plaster dental molds weren't giving him sufficient detail. He then testified the analysis was conclusive. Brooks was sentenced to life.

Years later, a California lawyer set up a sting. He sent West a photograph of bite marks from a completely unrelated case, along with a dental mold belonging to a friend who had nothing to do with it, and paid West $750. Two months later, West returned a twenty-minute video methodically explaining why the friend's teeth had made the marks. It was fabricated from nothing, and West matched it anyway. Defense lawyers in Mississippi spent a decade trying to use that video to disqualify him. The courts kept letting him testify.

The system's vulnerability is specific: juries cannot evaluate expertise, so they evaluate performance. West was polished, folksy, and fluent in technical language. He listed thirteen states where he'd been qualified as an expert, which meant, to each new judge, that some previous judge must have already vetted him. Legitimacy laundered itself down the chain. A 2001 study found that trained bite mark analysts matched marks to the correct teeth accurately only one-third of the time. The National Academy of Sciences eventually declared the entire field scientifically unsupported. By then, Brooks and a second man, Kennedy Brewer, had spent a combined thirty years imprisoned for murders committed by a man whose DNA was already in the police evidence room.

The Playbook: How a Hunch Becomes a Conviction

Think of wrongful conviction less as a mystery — who actually did it — and more as a recipe. The ingredients are almost always the same, assembled in the same order, producing the same result. What the cases in this book reveal is not a collection of individual tragedies caused by rogue officers and bad luck. They reveal a machine with interchangeable parts, running the same sequence every time.

The sequence starts with a hunch. In Norfolk, a detective spoke to a nervous neighbor who said a sailor named Dan Williams seemed 'kind of obsessed' with the victim. That offhand suspicion became the investigation's spine. Everything after it — eleven hours of sleep deprivation, fabricated polygraph results, suggested amnesia — served the hunch rather than tested it. When DNA excluded Williams, the detectives didn't reconsider; they expanded the theory. An accomplice, then two, then an 'Eight Man Gang' of strangers who had supposedly never met but spontaneously organized a gang rape together. Seven innocent men sat in jail on capital murder charges, all excluded by DNA, while the actual killer had already confessed. The hunch, once formed, was load-bearing. The entire prosecution rested on it.

Then comes the next ingredient: manufactured certainty. Texas Ranger John Wesley Styles told Clarence Brandley, 'Since you're the nigger, you're elected,' before the investigation had even begun. The actual killer's hair was found at the scene. The evidence box disappeared. Each piece of false corroboration looks legitimate in isolation — a confident analyst, a jailhouse informant, a witness identification, a confession. Each appears to be independent confirmation. What the book shows is that they are often manufactured from the same source: the original hunch, pressuring compliant people downstream until the file looks full.

The recipe works because no single ingredient announces itself as poison. So what happens when the truth arrives anyway — when the DNA comes back wrong, when the hair evidence resurfaces, when someone finally talks? That is where section four begins.

When the Real Killer Confesses and the State Still Won't Budge

What happens when definitive proof of innocence arrives and the state has it in hand? You might assume the machinery reverses — that some prosecutor or judge sees the new information and corrects course. What the book shows is something colder: the system's first instinct is not to fix the error but to explain it away, because admitting error means admitting that investigators lied, prosecutors hid evidence, and judges signed off on a fiction. That accounting is too expensive. So the state doubles down.

In Louisiana in 1976, a teenager named Jerry Francis walked into the sheriff's office and confessed to the Hasty Mart murder with the kind of specificity that cannot be coached. He described the position of the victim's body, the cigarettes knocked off the counter, the cash taken and the change left behind in a bag, the getaway car's color — details the police had never made public. He named his accomplice, Preston Demouchet, and the driver, Malcolm Roy, and told the investigators: the six Black men you've already charged didn't do this. We did. Sheriff Wattigny's response was to threaten Francis with jail for lying. When Roy also confessed, the sheriff dragged Roy's wife down to the station so she could berate him in front of the deputies. Then, under sustained pressure, both men recanted — and the police manufactured a bribery story to explain it away, claiming the defendants' families had paid the real killers to take the fall. The story required believing that a teenager's conscience-stricken confession, filled with accurate scene details, was a paid performance. Prosecutor Dracos Burke accepted that story and proceeded to trial anyway. The murder weapon — a .32 revolver recovered from a canal where Francis led police after confessing to a separate crime — was matched by ballistics to the Hasty Mart killing. Burke's solution was a gun-swap narrative: Demouchet had conveniently loaned the gun to defendant David Alexander days before the murder, and Alexander returned it afterward. No one investigated whether this was possible. Two innocent men were sentenced to ninety-nine years of hard labor.

Some of the people perpetuating this pattern genuinely cannot see past the original hunch. But the effect is identical either way: exculpatory evidence gets reframed as a threat to the case rather than a correction to it. Every confession by the real killer becomes a bribery plot. Every scientific report becomes an opinion that failed to merit a stay. The system has no mechanism for absorbing the possibility that it was wrong from the start, because being wrong from the start means everyone downstream — the detective, the prosecutor, the judge, the jury — participated in destroying an innocent person's life.

The Suppressed Tape, the Missing Box, the Buried Report

When a St. Louis County prosecutor filed his evidence boxes after Ellen Reasonover's 1983 conviction, two audio cassettes went in with everything else. Each had a label. The first was a secretly recorded conversation between Reasonover and her co-accused, Stanley White, taped while they sat in adjacent jail cells the night police arrested them. The second was a call police had arranged between Reasonover and Rose Jolliff, the informant whose jailhouse-confession testimony had just sent Reasonover to prison. Both tapes had been requested by the defense before trial, as part of standard discovery. Neither was handed over. They sat in the prosecutor's files for thirteen years.

When a federal judge finally forced the prosecutor's office to open its complete file in 1999, and Reasonover's legal team listened to the recordings, they heard something that made the state's entire case collapse. On the tape with Stanley White — two people who had no idea they were being recorded — Reasonover and White spent an hour furious, bewildered, and emphatic. They said they were innocent more than twenty times. Reasonover described seeing the crime scene photographs: she called the victim a 'little white boy' who looked 'pitiful' on the floor, adding that 'whoever killed him, that was cold.' The victim, James Buckley, was six feet eight inches tall and nearly two hundred pounds. The only way to describe a man that size as a 'little white boy' is if you had never actually seen him.

The second tape, recorded five days after Jolliff had first told investigators about Reasonover's supposed confession, captured Reasonover telling Jolliff directly: 'Why they trying to stick me with something I didn't do?' She was not speaking carefully for an audience. She was venting to the person she assumed had heard her, and she was describing innocence.

The prosecutor, Steven Goldman, told the federal court he had no recollection of either tape. The second one was found in his files with a handwritten label reading: Rose/Ellen, 1/12/83, Copy. Federal Judge Jean Hamilton's ruling was seventy-five pages long — three times the typical length — and she filed it within a month of the hearing. She wrote that no reasonable juror, hearing that first tape, would have believed a word of Jolliff's testimony.

What the Reasonover case makes impossible to ignore is the label. Not the wrongful conviction — those happen through negligence, through bad luck, through witnesses who lie. What happened here is different. Someone wrote Rose/Ellen, 1/12/83, Copy on a tape, put it in a box, and kept it there for thirteen years while a woman served a life sentence. The tape was not lost. It was filed.

The System Wasn't Built to Admit It's Wrong — But a Few People Were

The system is not designed to exonerate. It is designed to convict, and every institution downstream — the appellate courts, the parole boards, the governor's office — inherits the same incentive as the trial court: admitting error means admitting that investigators fabricated, prosecutors buried evidence, and judges presided over a fraud. Exonerations almost never emerge from the system correcting itself. They come from individuals who chose, at real personal cost, to act against it.

Bill Srack was a Republican suburban construction manager from Houston who had never given much thought to civil rights or the racial dynamics of small-town Texas justice. He sat through the first Clarence Brandley trial in 1980 with no particular agenda, and at the end of it he simply was not convinced. He found the janitors' testimony thin, he was troubled that the medical examiner had thrown away the semen swabs, and he thought the prosecution hadn't made its case. He said so in the jury room. What followed was eleven hours of being screamed at, called a racial slur repeatedly by three fellow jurors, eating breakfast alone at the sequestration hotel while everyone else sat together, and sitting by himself in the coffee room listening to the insults drift in through the wall. He held his vote. Brandley got a hung jury instead of a death sentence, and the cracks in the prosecution's case that eventually freed him had time to widen.

The retaliation came later. When Srack retired from corporate life and applied for a county government job, one of his three interviewers was the former prosecutor from Brandley's trial, now elevated to a judgeship. The prosecutor spent the interview questioning whether a man who had stubbornly defied eleven jurors could work with other people. Srack did not get the job.

Mary Johnson was the court reporter at Brandley's trial. She testified honestly against her boss, Judge Martin, and was fired and blacklisted from the Conroe courthouse. Jim McCloskey was a Princeton divinity student who became an investigator, spent forty years working wrongful conviction cases, and achieved seventy exonerations — one at a time, against a system with every institutional incentive to ignore him.

That is the cost the system extracts from the people who slow it down. The machinery of wrongful conviction is enormous and institutional. The machinery that reverses it is almost always one person, absorbing the consequences.

What a Lone Juror and a Divinity Student Have in Common

A retired juror who sat alone at breakfast for a week. A journalist who flew from London because something didn't add up. Jim McCloskey, who gave his life to pulling strangers off death row one at a time, knowing the row would never empty. The machine that puts innocent people in prison is built from policy, incentive, and law. The machine that gets them out is built from those people — fragile, replaceable, not guaranteed to appear. Which means the real question this book leaves you with isn't whether the system is broken. It's whether you are the kind of person who would have held that vote.