My mom, still right about everything. In a previous post, I explained her immediate skepticism about the Brandon Mayfield arrest. From the NYT today:
…the F.B.I. at one point told federal prosecutors that Spanish officials were “satisfied” with their conclusion.
But in interviews this week, Spanish officials vehemently denied ever backing up that assessment, saying they had told American law enforcement officials from the start, after their own tests, that the match was negative. The Spanish officials said their American counterparts relentlessly pressed their case anyway, explaining away stark proof of a flawed link — including what the Spanish described as tell-tale forensic signs — and seemingly refusing to accept the notion that they were mistaken….
Carlos Corrales, a commissioner of the Spanish National Police’s science division, said he was also struck by the F.B.I.’s intense focus on Mr. Mayfield. “It seemed as though they had something against him,” Mr. Corrales said, “and they wanted to involve us.”
The FBI continues to maintain it was just a random mistake by an examiner who didn’t even know Mayfield’s name, much less his religion, that initially led them to focus on Mayfield. I continue to maintain that’s total BS. Finally, does this fingerprint examiner have a family? Because I bet they would really, really like more time to be spent with them.
Statistical Update: This 2001 Washington Post article lists some widely varying estimates as to how many Muslims live in the US. The highest number was produced by a group of Muslim organizations and has been the subject of some doubt (numbers in millions).
Mosque Study Project: 6 to 7
2001 Britannica Book of the Year: 4.1
National Opinion Research Center: 1.5 to 3.4
CUNY Religious Identification Survey: 2.8
Reading the article, the methodology of the Mosque Study Project was obviously pretty bad. The total U.S population, according to the census bureau, is 293, 425, 566. So it seems as if probably more than 1% but substantially fewer than 2% of Americans are Muslims. I think that in the original article the FBI said the computer provided them with 50 close matches, from which Mayfield’s print was chosen as the best by an examiner (again, allegedly, without reference to his personal details).
{ 11 comments }
q 06.06.04 at 7:58 am
Although motives are difficult, some statistical analysis is helpful sometimes. How many people does the FBI have fingerprints? 1 million? How many of those people are Muslim? 50,000? How many Muslim converts: 5,000? How many high profile Muslim converts: 500? If it was a genuine mistake, the chances are 1 in 20 the person picked up would be not Muslim: 95%. If 1 person in 2,000 is a high profile Muslim convert, then the chances of a mistake selecting them are slight. Or were they only checking Muslim fingerprints? What is strange about the case is that if the fingerprint doesn’t match now, why did it match then? What has changed? Does anyone have a photo of the two fingerprints so we can compare?
Belle Waring 06.06.04 at 8:53 am
I don’t know of any pix, but if you read the article the Spanish investigators detail what a poor match it was: wrong number of loops, ridges turning up instead of down, that sort of thing. They gave it 7 points of similarity, I think, while the FBI claimed 15, but actually there is no one accepted standard for the number of points which need to match. This is especially true of latent partial prints. They claimed to have run it through a database of 44 million prints IIRC; they had Mayfield’s because of Army service. Given how few Muslims there are of any stripe in the states I think the odds that a random match from those 44 million would be Muslim are so slim as to defy credulity. I guess stranger things have happened.
John Quiggin 06.06.04 at 8:57 am
From the NYT
I don’t think this is actually crucial. The vital question is the second one posed by q. How likely it is that a mis-identification would pick up a Muslim with some sort of dubious connection (convert status isn’t critical, but is probably one of the factors that would count against someone)?
I’d guess about 2 per cent of the US population is Muslim and of those maybe 1 in 10 would have some sort of link that might give rise to suspicion, giving a 1 in 500 chance of this thing happening entirely by accident.
My guess is that the examination process spat out quite a few partial matches (say 50 to 100) and that Mayfield was the only one who produced any red flags when they cross-checked.
As the story points out, this episode is going to be very damaging for the FBI’s use of fingerprint evidence, and will almost certainly force them to tighten up criteria and rely less on “expert” judgement.
Tom T. 06.06.04 at 7:54 pm
The 9/11 commission is preparing to hand the FBI its ass for having failed to connect various dots prior to Sept. 11. Under that circumstance, if a list of 50-100 pops up a Muslim convert with at least some small ties to one of the Portland Six, it’s no surprise that the FBI is going up jump on that person. The only way of assuring no false positives is to assume, like Belle’s mother, that any identification is wrong, and the FBI’s incentives don’t lean that way these days.
Think of it this way, Belle: if you’re looking for a person with ties to an al-Qaeda bombing, what are the odds that that person is not going to be a Muslim? Certainly, the person that Spain is now naming as belonging to that fingerprint is a Muslim.
Matt Weiner 06.06.04 at 8:58 pm
The only way of assuring no false positives is to assume, like Belle’s mother, that any identification is wrong
Although, in this case, apparently double-checking Mayfield’s fingerprints against the one found in Spain would have sufficed.
Raymond 06.07.04 at 1:31 am
Lets not forget that Portland is a hotbed of muslim militancy. Also, this guy is a muslim convert (likely a bit of a cipher/psychotic).
I’m not comfortable with what the government did, but I’m also not comfortable with this idiot living in america.
Jeremy Osner 06.07.04 at 2:58 am
Raymond sez: The guy is a convert to Islam — and hence is likely psychotic or tending towards. I am not comfortable with this idiot living in America.
I say: Raymond — how do you feel about converts to Lutheranism? Judaism? Agnosticism? I am not comfortable with an idiot of your sort living in America.
q 06.07.04 at 1:14 pm
tom t: _if a list of 50-100 pops up a Muslim convert with at least some small ties to one of the Portland Six, it’s no surprise that the FBI is going up jump on that person._ Quite agree with you – just seems a bit amazing that out of 44 million checks, assuming 100 hits, with 500 high profile Muslims on the books (1 in 88,000) that one of them appeared at all. Using my bad maths I calculate the chance that a selection of 100 random people from a database of 44 million wont include someone from a subset of 500 people is 99.886% which is approximately 1 in 1000. If they have done 500 similar fingerprint checks this would reduce the odds dramatically, and it suddenly would not be so incredible.
In order to know any more, we need to understand the frequency and methodology of the fingerprinting operation.
Raymond 06.07.04 at 11:53 pm
Muslim converts are creepy and often zealots. There’s way too many better religions out there. We have ufo religions and such; surely no sane person converts to the religion of hatred, violence and oppression.
Raymond 06.07.04 at 11:53 pm
Muslim converts are creepy and often zealots. There’s way too many better religions out there. We have ufo religions and such; surely no sane person converts to the religion of hatred, violence and oppression.
Tom T. 06.08.04 at 12:13 am
Just for the record, my comment above should not be read to imply that I agree with Raymond’s views, which I expressly disavow.
Comments on this entry are closed.