Until the advent of DNA testing, fingerprint analysis was the most reliable form of forensic evidence in criminal trials. The reliability of fingerprints is NOT invalidated in any way because an examiner in an individual case identified the wrong finger from the wrong hand as being a match. When humans butcher their job in carrying out a scientific analysis, the fault belongs to the humans, not to the science.
The one US case referred to was posted on FR about four weeks ago. I pointed out then that it will most certainly be slapped down on appeal because it is bad law. Expert witnesses are examined in court to establish both their expertise, and their methodology. If they are deficient in either area, it invalidates their work in that case. IT DOES NOT INVALIDATE THE ENTIRE FIELD.
This is a tabloid-style scare story in The New Yorker. The author was just trying to make a buck by peddling a false scare. The editors should be ashamed of themselves for printing the story.
Congressman Billybob
You need honest judges and honest leadership of the FBI and police departments, and it seems as if honesty and political neutrality are hard to come by, these days. After an amazing series of public failures during the clinton years, and even earlier, the FBI simply cannot be trusted until there is a full-scale housecleaning, which is not yet in the cards.
Several years ago someone very close to me was found sitting in his Ford Bronco dead due to a gunshot wound to his head. This fellow had received what he regarded as a 'death threat' about a month before. The gun, found at the scene, had no finger prints on it at all. It was immediately declared a 'classical suicide'...front entry, rear exit. All blood splatter was on the driver side door, below the arm rest.
Six months later the autoposy report concluded that the bullet entered through the back of his head and exited in the front. The case was then forwarded to the homicide unit.
Meanwhile, the family had requested all police records,through the Freedom of Information Act . Once the detectives found out about the request they became very hostile and they told the family that the case would be closed as a suicide because they had to protect their pensions and make sure that the PD wasn't sued.
One thing, though, I didn't see the article mention which is extremely important is the difference between a "hot" and "cold" hit. In the former case, you have a crime, you have a suspect, and a finger-print check confirms the suspicion. In the latter case, there is a crime, but no suspect, and a fingerprint search identifies someone about whom there had been no suspicion (or even known relation to the case). Assuming fingerprint variations to be random (likely), a partial print that would match 1 in 100,000 people should constitute pretty damning evidence if found in a "hot hit", but not if it just turned up a "cold hit". After all, the population of likely suspects for a "hot hit" is typically less than a dozen people; a hundred or so at most. Even with 100 suspects, the odds that a wrong suspect will be fingered are less than 1 in 1000. By contrast, such a partial print could be virtually guaranteed to turn up some "cold hits" when checked against the entire database. Even a geographically-limitted search against 50,000 fingerprints would have nearly a 50% chance of flagging a wrong person.
Another point which I think needs to be mentioned is that while efforts are often focused on finding matching points, even one clear mismatching point should be enough to disqualify a fingerprint as matching another. When fingerprints are offered for comparison with matching points pre-marked, however, it's unlikely that the examiners are going to search as hard as they should for mis-matched points.
Finally, the existence of inter-mixed partial prints from multiple people can complicate things so much as to make a statistician's head spin. For example, regarding fingerprint features to be letters, consider searching for "SPAGHETTI" in the following list of fragments [* represents 1 or more mising characters]:
S-AGHE BE GHETT LA NG ETTI TO EAPThese fragments contain overlapping parts of three words. Could one of the words be "SPAGHETTI"? Well, the "S-AGHE", "GHETT", and "ETTI" framgents would suggest that. But the fragments could also come from "BETTING", "GHETTO", and "SLAGHEAP". Ascertaining whether a particular word (in this case) or fingerprint (in the real-world case) is involved in a collection of fragments can be very tricky.
"It was obvious the fingerprint was not Shirley's,'' Bayle told me recently. "It wasn't even a close call. She was identified on the left thumb, but that's not the hand the print was from. It's the right forefinger. But how can you admit you are wrong about Shirley's print without opening yourself to doubt about the murder suspect, too?" Bayle posted a comment on Onin.com, a Web site trafficked regularly by the world's fingerprint community. "I have looked at the McKie case,'' he wrote. "The mark is not identical. I have shown this mark to many experts in the UK and they have come to the same conclusions."
I've talked to a lawyer friend about fingerprints in court. He says that in actual trials, the images of the fingerprints are superimposed so that the jury can see the exact match.
Numeric analysis might be used for screening by computer, but the court case will hinge on being able to demonstrate an exact match with the images.
I don't know enough to comment intelligently on this, but if I'm ever on a jury and don't see the exact match, I will be suspicious.