John Puckett was linked to a decades-old rape and murder case based solely on DNA evidence, The Atlantic reports, but the prosecution and defense attorneys had different interpretations of how uncommon a random match would be.
According to one interpretation, there was a one in 1.1 million chance that a random person would match the crime scene DNA evidence, but another interpretation put it at one in three.
The prosecutor proposed telling the jury the first random match probability, and the defense the alternative one. That second interpretation is based on as the "database match probability" statistic, The Atlantic says. This approach takes into consideration the difference between a real random match and a match made from a pool of candidates, like those in a database.
Another way of presenting it to the jury, it adds, would be as an "n*p" statistic, or of the men of the right age in the city at the time of the crime, how many would likely match the evidence? For this case, The Atlantic says two other people living in the area matched the evidence.
"Each of these statistics has a very different interpretation of the significance of the DNA-database match," The Atlantic says. "Yet all are legitimate in one way or another, and there remains a lack of consensus among statisticians as to which one deserves priority within the criminal-justice system."
The European Network of Forensic Science Institutes recommended in 2014 report that additional DNA testing be conducted when only a database match connects someone to a crime. And, the FBI has said that the tables it developed to calculate DNA statistics contain errors. With high quality and quantity DNA, The Atlantic says, those errors seem small, but in other situations, there may be a more dramatic change of a one in a billion match becoming a one in a hundred match.