The problem is the denominator. If say 200 tests were done then 20 would be contaminated. So we are saying that for just one of those contaminated samples (5%) to be contaminated by Sollecito is beyond any possibility? The more tests you do the more likely you are to find contamination.
Well, no. In fact, something like 400 tests were done in the apartment. Seven samples were taken from the corridoor floor on Dec. 17. alone, another dozen were taken from the corridoor floor before that. Several samples were taken in the kitchen and from the floor in other rooms.
That there is no DNA from Sollecito on the floor is manifest.
From all those 400 tests we can infer that no large-size source of Sollecito's DNA was present in the apartment.
And we already
know that no other sample from the apartment was contaminated from Sollecito's DNA.
We have a small object (the bra clasp) and no evidence of the existence of a source and of a contamination path, the factual circumstances that would make this alleged contamination from Sollecito's DNA become probable are missing. We have to consider the probability that such source existed and found its path as small odds, and even if there was a double touch with glove fingertip of an alleged source external to the room, first, and the bra clasp after, we would have small odd to obtain this transfer. We should consider the need of a combination of these small probabilities, two or three unlikely circumstances each itself not probable, and we would obtain a (very) improbable event.
If we considered the odds of that improbable event as 1:10 (but we know they are far less) still the evidence would be significant, there would be no denominator problem. Because - pay attention here - as I said we already
know that the other samples are not contaminated from Sollecito's DNA. In terms explaining specificity, you need to explain why right
that sample was contaminated and not others, why not the other twenty picked from the floor for example; you cannot use the sheer number as a shield in order to say "one can happen, just because they were many", you need to actually explain why that only one happened, and that one specifically. Because you can see how many samples taken that same day would have not as incriminating as that one. The floor in the kitchen may have little meaning. But Sollecito's DNA in Meredith's room? And on her bra? Why did it happen on
that one?
What about the other unknown DNA profiles found? Why are these qualitatively different? Why are these other profiles innocent contamination, but Sollecito's exclusively the result of transfer during an assault on MK?
First there are no other profiles found, they are alleles, not profiles. The same difference between finding an old tire rather than a functioning car. You can infer there was a car attached once, of course, but you don't have any. But this is not the problem why they are qualitatively different. They are different because they hae a different logical, as well as statistical value: the known datum is different, those are random, completely generic; while Sollecito's matches a suspect. It's a remarkable coincidence.
If you know the DNA comes from a paramedic or a forensic, which you may even know was there, then you can reasonably infer it was just contamination. But if it matches a suspect, and it is not supposed to be there, and it's there and only there on that incriminating object, it's a specific indication.