Originally posted by PBE6
Food for thought:
http://en.wikipedia.org/wiki/Prosecutor%27s_fallacy
Excellent link. It provides an answer in the matter that is clearer than the manner in which I put it, but for those who did not check it out, here is what I took from it...
The usefulness of such testing and the probability of random matches would depend very much on the size of the samples looked through and the odds of the person's guilt without the DNA evidence.
In the event the suspect had other significant evidence against him, the DNA evidence would remove almost all reasonable doubt.
However, if the suspect was randomly found amongst a very large sample, and the prosecutor did not have much else in the way of other evidence, then the DNA evidence would probably be next to useless for the moment, because there would still be reasonable doubt of the person's guilt.
So the question of his chance of guilt does not have a definitive answer, because any definitive answer would be heavily influenced by context we are not given, and thus we cannot give a meaningful numerical value to it.