Case Study 33.1 — The Central Park Five: Five Confessions, No Forensic Link, and the Confession That Wasn't Theirs

A real, publicly documented case. The facts below are drawn from the public record of the 1989 attack, the 1990 trials, the videotaped statements, the 2002 confession and DNA match of Matias Reyes, and the court's vacatur of the convictions that same year. We treat a brutal assault soberly and confine ourselves to documented public facts. Nothing here minimizes what the surviving victim endured; the point of the case is that the wrong people were convicted of it, by their own coerced words, while the man who actually did it remained free to attack again.

Why this case anchors the chapter

If you wanted one case that contained every claim in this chapter at once, it is this one. Five teenagers — four 14 to 16 years old — were interrogated for many hours, gave statements admitting involvement in a crime none of them had committed, were convicted largely on those statements despite the absence of any physical evidence tying them to it, served years in prison, and were exonerated more than a decade later when the actual perpetrator confessed and his DNA matched the crime-scene evidence the five never could. It is the coerced-compliant false confession (§33.3) at scale: not one vulnerable suspect breaking under pressure, but five, on the same night, in adjacent rooms, each producing a statement that contradicted the others and the physical facts — and each persuasive enough, on tape, to convince two juries. It is also the cleanest possible demonstration of the chapter's central evidentiary point: a confession is a claim to be tested against the physical record (§33.5), and when it is finally tested against DNA, five confident-sounding admissions weigh nothing against one true match.

Background: the crime and the interrogations

On the night of April 19, 1989, a 28-year-old woman jogging in New York City's Central Park was raped and beaten nearly to death; she survived but had no memory of the attack. The same night, a large group of teenagers had been in the park, and several were involved in unrelated assaults and harassment of other people there. In the investigation that followed, five of them — Antron McCray, Kevin Richardson, Yusef Salaam, Raymond Santana, and Korey Wise — were singled out, taken into custody, and interrogated about the rape.

The interrogations are the heart of the case. They were long — the youths were questioned over many hours, some reportedly for the better part of a day or more — and they had the features §33.4 catalogs as the danger signs. The suspects were young: four were juveniles, the youngest 14. They were questioned under exhaustion, fear, and isolation. And critically, the process that produced the admissions was largely unrecorded — only the final statements were captured on videotape, after the hours of questioning that shaped them. Each of the five eventually gave a statement implicating himself and the others in the attack. Four of the five statements were videotaped; those tapes, of frightened teenagers describing a crime, became the prosecution's most powerful evidence.

A note on what "they confessed" actually meant here. The statements were not free-standing narratives that happened to match the crime. They were obtained at the end of long, guilt-presumptive questioning, and — exactly as §33.5 predicts — their incriminating detail did not hold up as independent guilty knowledge. This is the part the chapter teaches you to look for, and the part two juries did not.

The forensic evidence — and the evidence that wasn't there

The most important forensic fact about this case is an absence, and it was known at the time of trial:

• No physical evidence tied any of the five to the rape. Semen recovered from the victim and the scene did not match any of the defendants. Hair comparisons offered against them were, at best, the weak and later-discredited microscopic kind this book treats with great caution (Chapter 19) — not an individualizing link, and nothing that placed any of the five at the assault as its perpetrators. The biological evidence that could speak to the source of the attacker's DNA pointed to none of them.
• The confessions did the work the physical evidence could not. With no forensic link, the prosecution's case rested overwhelmingly on the videotaped statements. This is §33.1's lesson in its purest form: a confession is so persuasive that it can carry a conviction over the absence of any physical corroboration — jurors who saw frightened teenagers say they were involved discounted the fact that not a single piece of biology connected them to the crime.
• The statements contradicted each other and the facts. The five accounts were inconsistent with one another about who did what, where, when, and how — the kind of disagreement that, in a true multi-perpetrator account, is hard to explain, and that contamination and compliance explain easily: each suspect, pressed to confirm a narrative, supplied details to satisfy the interrogation rather than from memory of an event he had not witnessed. The accounts also did not fit the physical scene cleanly. A confession's power is supposed to come from its corroborated detail (§33.5); here the detail neither corroborated across the five nor matched the evidence.

The contamination question (§33.5), applied. Ask the diagnostic the chapter gives you: did any of the five volunteer a nonpublic fact the police did not have, which was then independently confirmed? The record of this case is the reason the diagnostic matters. The incriminating details in the statements were facts available to the interrogators, in a heavily investigated, highly public case — not self-generated knowledge that could only have come from being there. And because the hours of questioning that produced the statements were not recorded, no one could later reconstruct who first said each detail. The one thing that would have authenticated the confessions — volunteered, confirmed, previously unknown facts — is exactly what they lacked.

The turn: a confession that was actually true

The five were tried in 1990 in two groups and convicted; they received juvenile or youthful-offender sentences, except for Korey Wise, who was 16, tried as an adult, and sent to adult prison. They served their terms — several years each; Wise served the longest, roughly thirteen years.

Then, in 2002, a man named Matias Reyes, already in prison for other crimes, came forward and confessed that he alone had raped the Central Park jogger. His was the kind of confession the chapter tells you to take seriously, because it had the feature the five statements lacked: it could be tested against the physical record, and it passed. Reyes provided information consistent with being the perpetrator, and — decisively — his DNA matched the semen recovered from the victim and the scene, the same biological evidence that had excluded all five teenagers in 1989. The DNA had always pointed to a single attacker who was none of the convicted five; in 2002 it acquired a name.

On the strength of the DNA match and the reinvestigation that followed, the Manhattan District Attorney's office moved to vacate the convictions of all five, and in December 2002 a court did so. The men, no longer boys, were exonerated of the attack that had defined their lives. They are now known as the Exonerated Five.

Read the case against the chapter

Almost every section of this chapter is present in the facts at once:

• Why an innocent person confesses (§33.1). Five did — under length, fear, youth, and the engineered pressure of guilt-presumptive interrogation. The case is the standing refutation of "no innocent person would confess": five innocent people did, on one night, and convinced two juries.
• Interrogation, not interview (§33.2). The questioning was accusatory and guilt-presumptive; the suspects' denials were not the end of the inquiry but obstacles to be overcome. A process with no mechanism for recognizing innocence did not recognize it.
• Coerced-compliant confessions (§33.3). The five did not come to believe they were guilty; they complied, under pressure, with a demand, expecting — as coerced-compliant confessors do — that the truth would later sort itself out. It did not, until DNA forced it.
• Risk factors, stacked (§33.4). Youth (four juveniles, the youngest 14), long interrogations, exhaustion, and isolation — the dispositional and situational risk factors compounded, in precisely the combination §33.4 warns produces false confessions.
• Contamination and the unrecorded hours (§33.5). Only the final statements were taped; the hours that shaped them were not. The confessions' "knowledge" could not be authenticated as crime-derived rather than interrogation-derived, and not one self-generated, confirmed, previously unknown fact appears among them.
• A confession is a claim to be tested (§33.5, the chapter's thesis). Five confident-sounding admissions were contradicted by the physical record — the DNA that excluded all five — and finally overturned by a sixth confession that the same DNA confirmed. The science got the last word, exactly as it does for Renner in the cold case.

Outcome and significance

The convictions of all five were vacated in 2002; the men later reached a substantial civil settlement with New York City. The case became one of the most cited false-confession cases in the United States, taught in law schools and police academies, and a major driver of the reform argument in §33.6 — above all the case for recording entire custodial interrogations (so the manufacturing, not just the product, is on tape) and for special protections for juveniles, whose extreme vulnerability the case made impossible to ignore.

There is a further, sobering dimension that belongs in any honest account. Matias Reyes was not a stranger to the system in 1989; he went on to commit further rapes — including a murder — after the night of the Central Park attack and before he was finally caught. The five wrong confessions did not merely imprison five innocent people; they helped close the investigation around the wrong suspects and left the actual perpetrator free to attack again. This is the cost the chapter names at both ends of the error (§33.1, and the book's Theme 1): a false confession convicts the innocent and protects the guilty.

The lesson

The lesson is the chapter's thesis, made unmistakable by numbers: the persuasiveness of a confession is not evidence of its truth. Five confessions — more corroboration, on its face, than most cases ever have — were worth nothing against the physical record, because all five were produced by the process the chapter dissects and none contained the one thing that authenticates a confession. The case also delivers the chapter's hardest correction to the lay intuition: it is not only that an innocent person can confess, but that five can, on the same night, in good faith on the part of frightened teenagers who wanted the questioning to stop — and that the system will believe them over its own forensic evidence unless something as decisive as DNA forces a second look. That is exclusion over proof (Theme 1) and the CSI effect cutting both ways (Theme 4) in a single case: juries over-trusted the most persuasive evidence (the confessions) and under-weighted the most reliable (the exclusionary DNA), and only when the DNA was given a name did the error come undone.

Discussion questions

1. There was no physical evidence tying any of the five to the rape, and the semen excluded all of them in 1989. Using §33.1, explain how confessions could nonetheless carry a conviction over that absence — and what that says about the relative weight juries give a confession versus an absence of forensic corroboration.

2. Apply the §33.5 contamination diagnostic to this case. What single feature would a confession need to count as genuine guilty knowledge, and why does the fact that the pre-statement hours were unrecorded make that feature impossible to verify here?

3. The five statements contradicted each other and the physical facts. In a true multi-perpetrator crime, inconsistency among accounts is hard to explain; under compliance and contamination it is easy. Explain why, using §33.3 and §33.5.

4. Four of the five were juveniles, the youngest 14. Using §33.4, explain why youth so sharply raises false-confession risk, and connect it to the §33.6 reform of special protections (counsel / an appropriate adult) for juvenile suspects.

5. Matias Reyes's confession was treated as true; the five teenagers' confessions were false. Using the chapter's standard — a confession is a claim to be tested against the physical record — explain what made the difference, and why the DNA match is the decisive fact rather than the confession's emotional credibility.

6. The wrong confessions left the real perpetrator free to commit further crimes. Connect this to the book's Theme 1 (forensic evidence excludes) and to §33.1's framing of the cost of a false confession at both ends — the innocent convicted and the guilty free. How should that double cost inform how much weight an investigator gives a confession before the physical evidence is in?