Fingerprints on Trial

[R. Gambler]

In a shuffled deck, one would only have to receive one hand of 5 cards to get a Royal Flush; not liklely, but possible. In being dealt cards, the first hand of 5 cards could be be a royal flush, or after 2,598,950 hands one might not ever have received a royal flush. The point was supposed to be that to quote a probability of a latent print being duplicated, one would be admitting that there is a possibility that such a fingerprint might exist or be communicating that possibility exists. Does the possibility of duplication of an area of friction skin exist? We know that the DNA is not duplicated (except in twins) and that even twin's fingerprints are not duplicated (because in variables in environmental factors/growth/time). Addendum: David Farihurst was correct; the probability for a Royal Flush is 4 in 2,598,960 or 1 in 649,740. A mere transcription error and I stand corrected.

[Neville]

POINT WELL MADE

[Tazman]

Hi Gambler

Here's the deal. (That was a pun.) A Royal Flush is five complete cards. That is a "hand." Consider another hand -- the one growing on the end of your arm. That hand has five fingers. Take the five completely rolled fingerprints from that hand. What are the chances of another person having identical five rolled fingerprints in one hand? Pretty darned low, wouldn't you say?

To take the metaphor to a lower level, what are the odds of drawing one of the cards in a Royal Flush? One out of 2.6? But what about only a fragment of a card? What are the odds of a corner of a card matching? Or the color?

To carry on with the metaphor, then, what are the odds of one bifurcation matching? How about four points? The infamous "50K v 50K" study calculated the odds of four points matching between different fingerprints was 1 in 10,000,000,000,000,000,000,000,000,000 (1 in 10 to the 27th power). If the "50K v 50K" odds were correct, couldn't we just make our idents on four points? And couldn't we just run all our prints through AFIS with only four points marked, and be confident that the odds of a #1 candidate being wrong would be 1 in 10 to the 27th?

But wait! Haven't you ever seen a latent that had four points that could be said to "match" four points in another person's finger? Heck, we all have. It's not that rare at all. Statistics are beautiful, aren't they? Metaphors, too.

Maybe the critics are right when they talk about "smudged, fragmented" prints with distorted, minimal information. Or maybe not. Is it really like a card game?

[R. Gambler]

In every fingerprint, the detail and ridge arrangement is different. We know this by the research of Dr. William Babler, as well as Cummins and Midlo among others. The friction ridge formation and arrangement is the culmination of genetic factors and differential growth that is affected by environmental factors. In over 100 years of trying to locate such a match and even with the advent of computer searching of databases, no two fingerprints have been found to be the same. The possibility of two fingerprints being the same asymptotes to zero. Practically speaking, the possibility of such a chance occurrence does not exist.
When conducting a fingerprint comparison, the friction ridge expert must differentiate two fingerprints. When conducting a comparison of inked fingerprints that have excellent clarity, there is no problem. When comparing a latent print that lacks clarity (“smudged and distorted”) and that has minimal ridge detail, the comparison, of course, becomes very difficult. Experienced experts have all seen two prints (with minimal detail) that appear to “match”. The real question is not whether a fingerprint is or might be duplicated; but rather, at what point can that expert differentiate the fingerprints being compared. Hence the need for training and experience in fingerprint comparison.