If that were really true, my statistics professors wouldn't have been getting paid the big consulting bucks by folks who didn't share your view of statistics.
I am not sure I understand what you mean by incomplete data set. Could you expand on that some?
Having said that, I will take a guess. The mathematical probabilities I was speaking of were not calculated from a "sample" taken from the fossil record. Rather, they were calculated under the constraints of what we understand the processes to be. For example, if I recall correctly, (without my notes) there are 80 types of amino acids, 20 of which are found in living things. A single protein is built from a combination of 100 of these 20 amino acids in a specific order. Given just that information, it is possible to calculate the probability that a protien would be created by a random combination of 100 amino acids.
As I mentioned in a previous post, in Chapter 7 of "A New Kind of Science", Wolfram spends some time talking about the random generation of certain states. He uses a 10X10 grid of black and white squares to illustrate the probability of a particular solution found at random. He concludes that it would effectively be impossible. And, yes, he does address the fact that nature doesn't have to get it exactly right. Very interesting stuff, actually.
As I said above, I am not sure what you mean by incomplete data set so if this doesn't address your point, providing a further explanation of the incomplete data set you are referring to may help me understand.