Michael Connelly Should Stick To Fake Crime

The origin of Michael Connelly’s new true crime podcast, Killer in the Code, was a rather far-fetched tip fielded by Rick Jackson, a retired detective from the Los Angeles Police Department. A self-described “cold case consultant” named Alex Baber told him that he’d cracked two legendary mysteries. The first was the killing of Elizabeth Short, known posthumously as the Black Dahlia, whose bisected corpse was left in South Los Angeles in 1947. The second was the five homicides committed in the Bay Area between December 1968 and October 1969 by an epistle-happy psychotic calling himself the Zodiac. All six homicides, Baber claimed, were committed by one man—and there was proof.

The evidence was, in part, a name found in “Z13,” the cipher the Zodiac mailed to the San Francisco Chronicle in 1970. This was his third cryptographic missive; he’d sent two much longer efforts the previous year. “My name is —,” the killer wrote and, in lieu of his identity, provided a string of thirteen characters. Unlike the earlier ciphers, which have been decoded (the first within days, the second in 2020), Z13 had not been definitively solved. Baber claimed to have cracked the code and to have unearthed the name of the killer: Marvin Merrill.

This is not the first outlandish attempt to solve the killing of Short, or the subsequent Zodiac homicides. History is littered with such endeavors—fingering all sorts of unlucky souls as America’s great uncaught murderer, whether it’s Janice Knowlton pinning the death of Short on her own father or San Francisco Chronicle’s political cartoonist Robert Graysmith using the 1986 bestseller Zodiac to connect the crimes to Arthur Leigh Allen. There was even a prior effort to link these two specific cases, when former LAPD detective Steve Hodel attributed the crimes on his own father.

Where Killer in the Code differs, and is maybe even singular, is that it's not the work of amateurs or attention-seeking family members. The podcast is written and produced by Connelly, a legend in the crime fiction space—a certified grand master who has sold 90 million books. He’s the brains behind Bosch and The Lincoln Lawyer, and is the executive producer of four television series based on his characters. Not a slouch. And neither, as it happens, are his fellow podcasters. Jackson, who spent more than 30 years in homicide, is pals with Connelly and has provided feedback on many of his manuscripts. (“Four people benefited financially from my divorce,” he once said. “My wife, her attorney, my attorney, and Michael Connelly.”) Jackson’s onetime partner, Mitzi Roberts, led LAPD’s Cold Case Unit and inspired Renée Ballard, the star of a string of Connelly novels. And there’s Ed Giorgio, formerly of the National Security Agency—according to the New Yorker, “the only person to have been both the nation’s chief code breaker and its chief code maker.”

High expectations could be excused, given the bona fides of the people involved. Even I, whose opinion of the true crime genre can charitably be described as jaundiced, was intrigued. After all, no less than Pulitzer Prize-winning Christopher Goffard wrote a curtain-raiser for the Los Angeles Times. But having listened to the available episodes and spoken with key members of Connelly’s team, including the author himself, I’m left with one question: What was Connelly thinking?

The man at the center of the podcast—the very real person with very real descendents who will be left to deal with its blast radius—was born Marvin Skipton Margolis to Isadore and Lillian Margolis of Chicago in 1925. After spending his early years in Illinois, he enlisted in the United States Navy in 1943, discharged months after the war’s end. In 1946, he enrolled at University of Southern California’s College of Pharmacy. Margolis then knocked around a bit, and in 1961, as Skipton Merrill, he remarried. This decision to adopt waspish names was not uncommon in post-war America, where Jews routinely faced discrimination.

Merrill had firsthand experience with antisemitism. During his first semester at USC, as recalled in William Mann’s forthcoming Black Dahlia: Murders, Monsters, and Madness in Midcentury Hollywood, “a cross had been burned in front of Zeta Beta Tau, a Jewish fraternity house, on which the letters “KKK” were painted.” This is worth noting, I think, because the podcasters find Merrill’s name change to be darkly suspicious behavior. “The question was,” Baber asks in the first episode, “why did Marvin Margolis become Marvin Merrill?” That the answer might be self-preservation is not, for some reason, considered. A relative of Merrill’s told me it was “absolutely about antisemitism.”

Merrill then returned to California, where he went into real estate and construction, and founded an auto repair shop. In 1966, the Merrills moved to Vista, San Diego—five-hundred miles south of San Francisco. I mention this because geographic proximity is among the pillars of the case made for Merrill’s guilt. Absent forensic evidence such as DNA, fingerprints, and ballistics, it helps to place the accused at least in the vicinity of the crimes. With the Short case, this low bar can be cleared; Merrill, then 21, was living with a roommate in Los Angeles, where he was studying at USC. Short had also lived in their Hollywood Boulevard apartment for 12 days, not long before her death. Merrill and his roommate were, in fact, picked up by the LAPD, interrogated, and eliminated as suspects.

So far, five episodes into the podcast, Connelly’s sleuths have not been able to place Merrill even in the Bay Area—covering nine counties and 7,000 square miles—at the time of the killings and on the timestamped dates during which the Zodiac mailed his infamous letters. What we do know, thanks to newspaper stories, is that the Merrills had, as of 1971, resided at the same address on Mara Vista Drive in San Diego for five years, a period which covers the entirety of the crime spree. Furthermore, a paper trail of clippings place him in Vista during three Zodiac murders.

In January 1969, Merrill testified at a hearing in Vista weeks after David Faraday and Betty Lou Jensen were shot and killed on Lake Herman Road in Benicia. On July 4, when Michael Mageau and Darlene Ferrin were shot at Blue Rock Springs Park in Vallejo, a classified ad placed by Merrill, advertising for a business partner, began running in the Vista Press. And then in mid-December, Merrill was documented searching for the lost family dog a few days before the Zodiac sent taunting Christmas wishes—postmarked San Francisco—to Melvin Belli.

Of course, none of this means that Merrill was definitively in Vista during the crimes. But the makers of Killer in the Code have not provided evidence to the contrary. During the third episode, Connelly asks: “Can we place Marvin Merrill in geographic proximity to the [Zodiac] killings?” Answering his own question, Connelly cites unspecified records showing that Merrill “kept a home near San Diego through the late '60s and into the '70s, but records also place him in San Jose in 1969 where he rented a house under the name Skip Merrill, another alias[.]”

In other words: Not convincingly.

In interviews with me, neither Baber nor Jackson were willing to consider the possibility that Merrill isn’t, you know, the LeBron James of murder. They swear there are documents placing Merrill in the Bay Area on the necessary dates, but each would claim the other guy had them. "Trust me, Elon,” bragged Baber. “I don't say something unless I have the receipts." When I pressed him to produce said receipts, Baber could not. When I asked Jackson if he, a former detective in the LAPD, was able to verify Merrill’s presence in the area, he said he hadn’t, “and we may never be able to.”

Of course, Baber and company don’t feel the need to prove Merrill mailed the ciphers since, according to them, they can prove he names himself as the killer in Z13. As Connelly says of Baber in the first episode, “He devised a computer program that used artificial intelligence to weed through 71 million possible names.” This sounds like an impressive feat from a man the Killer in the Code website describes as “an autodidact polymath” who has logged “more than 13,000 hours of privatized, self-directed study, integrating forensic science, criminal psychology, and investigative methodology to advance a comprehensive understanding of homicide investigation.” The trouble is, even if one ignores that Baber keyed in on a man whose association with the Short case was well known, and we allow for the possibility that the person who had written "I will not give you my name” in a prior cipher was suddenly willing to risk the gas chamber, “Marvin Merrill” is, as Baber himself acknowledges, just one of thousands of possible name permutations.

To his credit, Connelly sought independent verification of Baber’s solution. Giorgio, the ex-NSA codebreaker, and his own team checked Baber’s work, and deemed it sound. “We found a plausible explanation for the proposed solution, using methods described in foundational cryptographic texts,” they wrote in their published findings. Giorgio also looked for a keyword—a word chosen by the cipher-maker that is often related to the solution. Finding it would, therefore, confirm the validity of Baber's solution.

Giorgio’s team found it: ELIZABETH. This was just the confirmation Connelly and company were looking for. That key, Giorgio told me, “identifies the Zodiac suspect and also provides new evidence linking it to Elizabeth Short.” But does it? It’s true that this keyword, in combination with an idiosyncratic scheme devised by Giorgio and Baber, produces the plaintext of MARVIN MERRILL. But, as a cryptanalyst told me, one can use the keyword MONASTICISM, adhering to those same rules, and still produce MARVIN MERRILL as the cipher’s solution.

This is indicative of the glaring problems at the heart of the whole process. For one, Z13 is widely believed to be too short to determine a unique permutation key. Unlike the first two Zodiac ciphers, which were hundreds of characters long, the brevity of Z13 doesn’t allow for an unambiguous solution. It’s not impossible for the cipher to be cracked, but a solution would require external data.

In order to learn what it would take to confirm a solution to such a slim cipher, I called a couple of cryptanalysts. Bill Briere, who has delved deeply into the Zodiac ciphers, says that the killer would likely need to be first definitively identified by other means, such as DNA. Once that happens, perhaps a search of his residence turns up “all the notes he used to encrypt with, and now we have his key in hand.” But perhaps the killer himself would have to personally assist with the solution. Even this doesn’t guarantee success. David Oranchak, a member of the team that cracked Zodiac’s second cipher—known as Z340—points to the example of serial killer Dennis Rader, who had been sending law enforcement letters and puzzles, including a fractionated cipher. The FBI’s cryptanalysis unit was stumped, on account of how short it was. Once Rader was caught, he was asked what the message meant, but, even after he revealed the plaintext and explained the method, the FBI still couldn’t replicate what he’d done. Rader had made mistakes. “Despite being unable to decipher his own ciphertext,” a Bureau cryptanalyst later wrote, “Rader seemed to view this as an obvious system with a message that should have been read immediately.”

The takeaway, I think, is that the podcasters have acted with certainty where they ought to embrace caution, and worse, gone about the process backwards. With no concrete supporting evidence, they reached a conclusion—that MARVIN MERRILL was the solution to Z13—and then backfilled an explanation.