The Zodiac Ciphers and the Cracking of Z340

For fifty-one years, a rectangle of 340 handwritten symbols sat unsolved. It had arrived at the San Francisco Chronicle on 8 November 1969 in an envelope with the paper’s own address block-lettered on the front, seventeen lines of triangles, circles, Greek letters and made-up glyphs, no note attached. The man who sent it had already killed. He had already proved he could write a code that worked. And then he handed the world a puzzle and waited, and the world could not read it, and that silence became one of the most durable obsessions in American crime.

The pull of the thing is easy to feel even now. A killer who was never caught left behind, in his own hand, a message he clearly wanted read. Somewhere in those 340 characters, surely, was the answer everyone wanted: a name. Solve the code, name the man. That was the promise the cipher seemed to make, and it is worth understanding why the promise was so convincing before understanding why it was never quite true.

The killer who wrote his own puzzles

The record here is solid and worth stating plainly, because it is the foundation everything else grows out of. Between December 1968 and October 1969, a man murdered at least five people in and around the San Francisco Bay Area: teenagers Betty Lou Jensen and David Faraday near Vallejo in December 1968; Darlene Ferrin in July 1969; Cecelia Shepard at Lake Berryessa in September 1969; and cab driver Paul Stine in San Francisco in October 1969. Two other victims survived. In letters to Bay Area newspapers he called himself “the Zodiac” and claimed thirty-seven kills, a number no investigator has ever substantiated.

What made him unusual was the writing. Beginning in late July 1969, he mailed taunting letters to the Chronicle, the Vallejo Times-Herald and the San Francisco Examiner, and on 1 August he sent each paper a third of a 408-character cipher, demanding it be printed on the front page or he would “kill a dozen people.” The papers printed it. Within about a week it was solved, and not by the police or the FBI. A schoolteacher named Donald Harden and his wife Bettye, of Salinas, cracked it at their kitchen table by guessing that a vain killer would begin with the word “I” and would probably use the word “kill.” The decoded message opened: “I like killing people because it is so much fun.” It was a homophonic substitution cipher, meaning several different symbols could stand for the same letter, a technique designed to defeat the frequency analysis that breaks simple codes. The Hardens broke it anyway in days.

So the terms were set. The Zodiac could build a cipher that concealed a real, readable message. He had done it once. When the 340-character cipher arrived three months later, the reasonable assumption was that it, too, concealed a message, and that patience and cleverness would eventually yield it up. The 408 had contained a confession of sorts. What might the 340 contain? The Hardens’ success is the reason a generation of solvers believed the answer was always just one insight away.

The fifty-one-year wait

The 408 had also carried a sting in its tail that should have been a warning. Its final eighteen characters, the sequence beginning EBEORIETEMETHHPITI, decoded into nothing sensible even after the rest fell open, and have never been convincingly explained; the likeliest reading is that the killer padded the ending or made errors he did not bother to correct. A cipher that ends in gibberish is a reminder that the man was clever but careless, and that not every mark he made was a message waiting to be read. Solvers who assumed every symbol carried meaning were, in a sense, giving him more credit than he had earned.

It was not one insight away. The 340-character cipher, which codebreakers came to call Z340, resisted everyone. The FBI’s Cryptanalysis and Racketeering Records Unit worked it. The American Cryptogram Association worked it. Hobbyists filled notebooks and, later, hard drives. The problem, it turned out, was that the Zodiac had done something his first cipher had not: he had scrambled the reading order. The 408 could be read straight, left to right, top to bottom, once you knew the symbol-to-letter map. The 340 could not, because the plaintext had been laid into the grid along a diagonal path before being enciphered, and no one knew that.

The breakthrough came from three men who had never met in person. David Oranchak, a software developer in Virginia, had been chipping at the Zodiac ciphers for fourteen years and ran a website and YouTube channel documenting his failures. Sam Blake, an applied mathematician in Melbourne, calculated that there were roughly 650,000 possible ways the message could have been written into the grid and read out. Jarl Van Eycke, a Belgian warehouse worker who wrote cipher-cracking software in his spare time, had built a program called AZdecrypt that could test candidate solutions at speed. Blake narrowed the reading paths; Van Eycke’s software ground through them; Oranchak spotted, in the noisy output of one run, fragments of real English: “TRYING TO CATCH ME,” “GAS CHAMBER,” “PARADICE,” the same idiosyncratic misspelling that had appeared in the 408.

They announced the solution on 11 December 2020, and the FBI confirmed it. The message read, in part: “I hope you are having lots of fun in trying to catch me… I am not afraid of the gas chamber because it will send me to paradice all the sooner.” It was the same voice, boastful and self-pitying, referencing a popular television programme of the day in which a host had discussed the case. Fifty-one years, three continents, and the answer at last came out of the grid.

Where the story wanted to go instead

Here is the fork, the place where the documented history and the folklore of the ciphers part company. Everyone who had waited half a century assumed the message would contain the one thing that mattered: the killer’s identity. The whole romance of the ciphers rested on that assumption. Solve the code, name the man, close the case.

The message named no one. It was another taunt, another performance of invulnerability, with nothing in it that pointed at a single human being. And for a certain kind of believer, that was intolerable, so the story simply moved on. Within weeks the theory appeared that the real code was still hidden inside the solved one, a second layer, an acrostic, a name spelled out by the first letters of the diagonal reading or by the leftover symbols. Two smaller ciphers, the 13-character “My name is —” cipher and the 32-character “map” cipher, had always been treated as the true prize, and they remain unsolved to this day, partly because thirteen characters carry too little information to be cracked with confidence: dozens of names fit, and no method can choose between them. That has never stopped people from announcing that they have solved the 13-cipher and named the killer, usually a relative or a long-dead suspect, on the strength of a fit that a coin-toss could have produced.

This is the same machinery that turns any partial revelation into a doorway rather than an answer, the pattern-hunting instinct explored in red string and corkboard. Give the mind a grid of symbols and the promise that meaning hides inside, and it will find meaning whether or not the sender put it there. The Zodiac’s genuine ciphers primed people to see hidden text everywhere, including in places he left blank.

Who carried it, and why they could not let go

The cast of long-term Zodiac solvers is worth taking seriously, because they were serious people in every practical sense. They were retired engineers, working programmers, cryptography hobbyists, true-crime obsessives, some of whom corresponded for decades and built careful, testable methods. Oranchak himself is the proof that the patient, sourced, method-driven approach could win: he documented every dead end, published his tools, and shared credit. The difference between his fourteen years and the fifty-first false “name solution” is the difference between a method that can be checked and a hunch dressed as a discovery.

Robert Graysmith’s 1986 book Zodiac, later filmed by David Fincher in 2007, did more than any police bulletin to fix the ciphers in the public imagination, presenting them as the master key to a case that had gone cold. That framing was enormously effective and quietly misleading: it taught millions of readers to believe that the code and the killer’s name were the same problem, so that cracking one would deliver the other. The 408 had contained a confession but never a name, and the same was always likely to be true of anything he wrote, because a man who signs his letters with a symbol is advertising his anonymity even as he demands the world’s attention.

What kept the amateur field alive was the structure of the case itself. An unidentified serial murderer is an open wound in the record, and the ciphers offered the tantalising possibility that the killer had, in a fit of vanity, filed his own confession and identity into a form that only patience could unlock. That possibility gave the powerless something to do. Anyone with a pencil and a printout could, in principle, be the one who cracked it and handed the police their man. It is the same democratic promise that drew thousands into the online puzzle-hunt of Cicada 3301: a code in public, a solution theoretically available to anyone, and the intoxicating sense that the next insight might be yours.

What the ciphers were really about

Strip away the hope of a name and what the Zodiac’s ciphers actually document is a man performing his own significance. He killed, and then he curated the killing: named himself after a watch brand, designed a crosshair symbol, demanded front-page placement, threatened to bomb school buses, and wrapped it all in codes that dared the authorities to prove they were cleverer than he was. The cipher was the tool of a man who needed to be the smartest person in the investigation of his own crimes.

The fifty-one-year obsession that followed says as much about the people waiting as about the man who made them wait. A killer who vanishes leaves a story with no ending, and human beings find stories with no ending almost unbearable, which is why the same names circulate for decades and why every unsolved disappearance grows a second life in the hands of amateurs, as with the Somerton Man and his own scrap of code. The ciphers promised that the ending existed, sealed and waiting, if only someone were clever enough. When Z340 finally opened and the ending was not inside, the obsession relocated, to the thirteen characters and the map, to the layers within layers, because the need the ciphers served was never really about cryptography.

The real achievement of December 2020 was quieter than the fantasy it disappointed. Three strangers, working with published methods and open tools across three continents, did something fifty years of hoping had not: they read what the man actually wrote. That it turned out to be a boast rather than a confession does not diminish the work. It clarifies what the whole long puzzle had always been made of, on both sides of the code. A man who wanted to be studied, and a great many people who could not stop studying him, each convinced that meaning was hiding one layer deeper, waiting to be found.