|Posted on Tuesday, May 09, 2006 - 08:47 pm: ||
I would like to respond to a couple of items if interest within your post, and potentially have you explain them a little bit more...
"half a fingerprint that was not a 100% match"
Question - how much of a match was it? Did someone report or mention that it was "close" or "similar" or that there wasn't quite "enough" to make a match? Or was it simply not identified?
"it was not in the mans area"
Do you mean the print was found at a location other than near the victim, or in a location not able to be accessed by the victim? Other?
"according to his families testing nature"
I would like to understand what this means a little better before I fully answer.
But what I will say is that identification can occur manually on 1/2 a fingerprint. If comparisons between that "latent" (hidden, or crime-scene) fingerprint and the "known" fingerprints of suspects or victims were conducted at the time, it is sometimes a very different thing than being run on AFIS. That brings up your last point...
It is true that AFIS systems have improved over the years. How much improvement occurred between 1997 to 2001 on the particular system in question? I don't know the state or the vendor, but it is likely very little. But the most important thing to remember with AFIS systems is that running a print again in a slightly different way (or by different examiners) can and will improve "hit" rates. Sometimes examiners see things a little more or a little less like the system sees things, and simply re-running an impression with a few more characteristics, or a few less characteristics, can make the difference in hitting or missing the match.
Finally, the answer to your last question, "can you really get an accurate reading from less than half a fingerprint and no other evidence whatsoever?", I reply with a resounding "yes". It happens reliably across the world thousands of times a day.
|Posted on Tuesday, May 09, 2006 - 08:12 pm: ||
I have a question that maybe someone can help me with. I was reading a case in which a man was convicted for a crime that happens year prior in which a house was robbed and the man killed. The only evidence found on the man was half a fingerprint that was not a 100% match. It was not in the mans area and according to his families testing nature. Now what i want to know is from all the new fingerprinting systems out there being developed from 1997-2001 how many points of comparison need to match and can you really get an accurate reading from less than half a fingerprint and no other evidence whatsoever ?