SERIAL KILLERS WHO DID THE LAST THING YOU WOULD EXPECT

In 1978 a registered sex offender on parole in California was cast as Bachelor Number One on a network dating show, went on air, and won the episode. Rodney Alcala had killed at least twice before the taping. He killed again after it. The woman who picked him declined to go on the date, having found him unpleasant off camera.

That is the sort of detail this archive collects, and it is worth being clear about why. It is not because it is strange. It is because a broadcast tape exists. Most of what is written about these men comes from the men themselves, in confessions given years later to detectives who wanted a case closed, and this archive has an entire article about how badly that goes. The details below are different in kind. Somebody else made the record: a television network, a court clerk, a traffic officer, a woman walking her dog. None of the six below rests on a killer's own word for it.

The Alcala tape is the plainest example. He was on parole and on the sex offender register at the time, which is a matter of record, and none of it prevented him being put in front of a camera. Nobody involved in the booking had any reason to know what he was, and that is the point. The screening that existed asked whether he was entertaining. It did not ask anything else, and it had no mechanism for asking. He was convicted of seven murders in the end and is suspected of many more, and the tape sits in the middle of that period rather than before it.

In Santa Cruz County in 1973 the district attorney's office was prosecuting Herbert Mullin and Edmund Kemper at the same time. Two separate men, thirteen deaths and ten, in one small coastal county, in the same year, before the same court. Mullin's jury rejected insanity in August 1973 and convicted him of two counts of first degree and eight of second degree murder. He was refused parole repeatedly and died in prison on August 18, 2022, aged seventy-five.

This archive has written at length about the Baton Rouge investigation, where Derrick Todd Lee and Sean Vincent Gillis were killing in the same city in the same years and some of Gillis's victims were briefly attributed to Lee. Santa Cruz is the earlier and smaller version of the same problem. Nothing about the caseload of a county prosecutor is designed for two at once, and the reason both cases are remembered separately is that both men were caught, which was not inevitable. Lee was found when a two-detective department went back to two old files. Gillis was found when a laboratory identified a tyre and somebody drove round to the ninety houses that had bought one. Neither was found by the apparatus built to find him.

David Berkowitz was caught by a parking ticket. Cacilia Davis, walking her dog near the scene of the Moskowitz shooting, saw a man take a summons off a car parked by a hydrant minutes before the shots. The ticket traced to an address in Yonkers. Two Yonkers patrolmen, Peter Intervallo and Thomas Chamberlain, had independently been building a file on the same man for reasons of their own.

Set that against the shape of the New York investigation as a whole, which was the largest manhunt in the city's history to that point. What ended it was a summons issued by an officer doing something entirely routine and a member of the public who noticed a stranger at a car. The paperwork existed before anyone knew it mattered. Cacilia Davis was not a witness to a shooting. She was a woman who had seen a man behave oddly around a parked car, and the reason her account mattered is that the summons he took off the windscreen had a number on it.

Edmund Kemper's contribution is stranger and more consequential. He talked at length to the FBI agents Robert Ressler and John Douglas in the 1970s, and those interviews became foundational material for the Behavioral Science Unit's work on sexual homicide. A body of professional knowledge used to catch people was built substantially out of long conversations with a man in a California prison.

The roster's own note on his entry is worth repeating, because it applies to everything he is quoted saying: he is unusually articulate about what he did, which is the reason he is quoted so often and the reason the quoting needs care. An account given by the perpetrator is still an account given by the perpetrator, however fluent. The Behavioural Science Unit knew that at the time and worked with it anyway, having nothing better.

Bundy is the counter-example that makes the distinction sharp. On death row he talked a great deal. Ressler, who interviewed him too, found him manipulative and evasive. Special Agent William Hagmaier and the Washington detective Bob Keppel got further, drawing out confessions and locations, and Bundy went as far as offering his own reading of the Green River Killer, who was at that point unidentified. In the days before his execution he admitted thirty murders across six states.

Thirty is the number he gave. It is not a number a court established, and the difference between those two things is the argument of twelve confessions nobody could stand up. What survives from Bundy that nobody can dispute is a docket, a set of dental impressions and a pseudonym in a book. What survives from the talking is a figure he chose.

One more, and it is not about a killer at all. Ted Bundy's first known attack, on January 4, 1974, was on Karen Sparks, an eighteen-year-old student in the University District of Seattle. She survived with permanent brain injury and lived for decades afterwards under the name Joni Lenz. The pseudonym was given to her by Ann Rule, who wrote about the case.

That is the only entry on this list where the documented detail protects somebody. A survivor got a name that was not hers so she could live behind it, and it held for decades, in a case that has been written about more than almost any other in this archive. Everything else here is a record somebody kept without knowing why it would matter. That one was a record somebody withheld on purpose.

There is a version of this subject that treats the whole thing as trivia, and there is a reason to be careful with it. A game show appearance is memorable because it is incongruous, and incongruity is entertaining. But the reason to keep the Alcala tape is not that it is entertaining. It is that a state that had him on a register, on parole and under supervision could not stop a television production company putting him on air, and the record of that failure is a videotape rather than anybody's later account of it.