All the fear that's fit to print

"[Brandon] Mayfield was detained on May 6 and held for two weeks as a material witness in the terrorist attack that killed nearly 200 people and injured about 2,000. He was released with an apology from the FBI.
We now know that Mayfield's fingerprint was not the one on the incriminating bag. Instead, the FBI says the print belongs to Ouhnane Daoud, an Algerian national. Yet, according to the FBI, Mayfield's fingerprints matched the print lifted by Spanish authorities at 15 particular ridge points. This is far more than what is often required in criminal court to assert an absolute match. In the Washington Post last month, Jennifer Mnookin, law professor at the University of Virginia, wrote "experts have declared positive fingerprint matches in court after finding even fewer than eight points."
What the Mayfield case suggests is that fingerprint matches may not be the gold standard of proof that law enforcement has held it out to be. Mnookin said serious concerns about the reliability of fingerprint analysis have been roundly ignored. First, she said, we really don't know whether each fingerprint is unique, since valid statistics don't exist. It could be that many people have fingerprints so similar as to be virtually indistinguishable."