On June 30th, 1999, officers in St. Louis, Missouri, found the body of Ricky McCormick, 41, in a field. He had been murdered. There were no clues, with two exceptions: two notes written in code in McCormick's pants pockets.
The murder remains unsolved. Over a decade later, the FBI's Cryptanalysis and Racketeering Records Unit (CRRU) has essentially thrown up its hands. "We are really good at what we do," said CRRU chief Dan Olson in a solicitation on the FBI's website, "but we could use some help with this one."
Crowdsourcing cryptography—it's the latest crime-solving innovation from the FBI. Since setting up a website devoted to the problem last week, responses have been pouring in.
McCormick's code uses what the FBI calls a "maddening variety of letters, numbers, dashes, and parentheses." McCormick was a high school dropout—but literate—and his family has said that he used coded notes since childhood, though no one in his family knows how to decipher his code. Twelve years of analyzing have gone nowhere. "Standard routes of cryptanalysis seem to have hit brick walls," said Olson.
The public might be able to help in one of a few ways. First, someone out there might have experience with a similar kind of code. Even if there isn't further evidence out there, said Olson, "Maybe someone with a fresh set of eyes might come up with a brilliant new idea."
The Boston Globe reported that a cryptography expert at MIT didn't hold out much hope for the code to be cracked. The data sample—two notes—was simply too small for statstical models to be of help. "Current techniques we've developed require some amount of data to do it," said Regina Barzilay. "The goal would be to require less and less data." New Scientist talked to several amateur codebreakers, though, and found them hard at work. The hive of minds on the online group Kryptos, dedicated to cracking the code on the CIA sculpture of the same name, has set to work.
The move by the FBI seems strange. Do they really think the public can do a better job? If so, why don't they crowdsource all their cryptographic projects? And why wait 12 years on this case? We add two theories of our own—for the FBI's motives: 1. Occasionally public exhortations for investigative help actually yield leads. Murderers are said to revisit the scene of their crime—and that sometimes extends to visiting websites set up by investigators. The public solicitation could yield new data—even if it's not data on the code itself. 2. It's a public relations maneuver. The FBI page is labeled as Part 2 of a series on breaking codes to stop crime, and contains a very basic sidebar on the rudiments of code-breaking. In that, it almost resembles another page on the FBI site—the kids page, which has features like "About Our Dogs" and "Play Our Games."