An archive of serial murder has to decide what serial murder is, and then it has to decide what to do about the cases that fail the test but that everybody expects to find. Thirteen entries here are carried on those terms. Each one says in its own first paragraph that it does not belong.
The definition this roster uses is the ordinary one: separate killings, by one person, with an interval between them. The interval is the part that does the work, because it is what separates a person who killed and then went back to ordinary life from a person who did not stop.
Four entries fail on the interval alone. Richard Speck killed eight student nurses in one house on one night. John Linley Frazier killed five people in one attack in one place. Nasra Al-Enezi killed a great many people in a single act on a single day. Jeremy Bamber was convicted of shooting five members of his family at White House Farm in one night; he maintains his innocence and a Criminal Cases Review Commission application lodged in 2021 remains under review. All four are mass murder, and all four are among the most searched names in true crime.
Charles Starkweather is the closest call in the archive. He killed once in December and then ten more times over eight days at the end of January, moving continuously and never returning to ordinary life in between. That is a spree, the interval exists but the return to normal does not. His entry says so, and keeps him, because the distinction is genuinely fine and a reader who comes looking for him deserves the explanation rather than a blank page.
Phoolan Devi and Hiroko Nagata fail on something else. Devi is held responsible for a number of men killed in a single incident, in a case that is really about caste violence in Uttar Pradesh. Nagata's fourteen deaths were a political purge inside the United Red Army, a leadership group killing its own members over a few weeks. Neither is one person killing strangers over time. Both are here because people look for them here.
Then there are the four entries that exist to say a story is wrong.
Vera Renczi appears in almost every encyclopedia of female serial killers, with a photograph, a count of thirty-five, and a set of vivid details about zinc coffins in a cellar. No trial record, no arrest record, no contemporary Romanian press account, and no verifiable person has ever been produced. Her entry's conclusion is that she probably never existed.
Carl Tanzler killed nobody, was never accused of killing anybody, and is on these lists constantly. What he actually did, taking a dead woman from her tomb and living with the body for seven years, is macabre enough that the misfiling has stuck.
Dean Faiello killed one person by practicing medicine he was not qualified to practice, and pleaded guilty to manslaughter. Thomas Quick confessed to more than thirty murders, was convicted of eight, and had every conviction quashed between 2010 and 2013. He is in an archive of killers as a record of a miscarriage of justice.
The last two are the borderline ones, kept because the reader should be able to see the reasoning. Lowell Amos has one conviction and four attributed deaths, his mother and three wives, all of whom died in circumstances later reinvestigated. Eli Stutzman has one conviction and five attributed, and no court has connected them.
There is an argument for cutting all thirteen. A roster of serial killers that contains a man who killed nobody is, on its face, not doing its job.
The argument against is what happens if you do. Somebody arrives looking for Speck, or Starkweather, or Renczi. If the entry is not here, they go and read one of the encyclopedias that has Renczi down for thirty-five murders and prints the photograph, or one of the lists that has Tanzler filed as a killer. The misinformation does not stop existing because this archive declined to address it.
So the rule is disclose rather than exclude. Carry the case, put the classification note first, and let the reader see exactly where the line is and which side of it this one falls on. Thirteen entries out of 1,018 is a small price for not pretending the question is simpler than it is.