This is a very silly and quite blatant misuse of statistics on your part.
Suppose I was looking out a window and I said to you (you not being able to see out the window) "Oh look, I see an albino pigeon".
If you were skeptical of this claim, the intelligent thing to do would be to get up and look out the window.
What would be very foolish would be to sit on your bottom, refusing to look out the window, and say "Your claim is completely useless unless you provide me with statistics as to the prevalence of albinism in pigeons. If it turns out albinism is very uncommon I plan to fix the belief that you have not seen an albino pigeon, because the odds of picking an albino pigeon by pure chance out of the pigeon population would then be very low".
This would be very foolish for two reasons. Firstly, you can tell an albino pigeon by the fact that it's white. Secondly, the reason this particular pigeon was noteworthy in the first place was because it was an albino pigeon.
What you are trying to pull here is exactly analogous: Amanda Knox's confession has the specific characteristics of an internalised false statement - it's an albino pigeon. We are examining it precisely because there is strong evidence in this case that the conviction was a miscarriage of justice - we are looking at it because it's an albino pigeon.
This is the wrong forum in which to try bluffing with fallacious statistical arguments. The Knox case was not randomly selected for examination from the complete pool of all court cases, and Knox's false statement has specific characteristics indicative of an internalised false confession.