But you're still hung up on the probability of a specified endpoint--maybe one of a number of possibilities, but you're still trying to estimate the probability of ending up with a human, or a clam, or some other specific organism. I'm surprised that someone so adamant about the need for math skills can't see the difference between the probability of a specific outcome--no matter what specific outcome you choose--and the probability of some outcome at all.
I'll try another analogy. You're standing at the mouth of a river and saying, "The chances of this river forming are astronomical." And you're right: it required that tree to fall over at that particular point, and that bank to wash away in the extra rainfall that summer, and and and. And now you're saying, "I don't care, pick a different river. The chances for that one are astronomical, too." And you're right again. The chances of any particular river forming are infinitesimal, requiring countless random acts, and there's no way you could predict the river's course by standing at its source.
And yet, given the snowpack in the mountains, it's practically a certainty that rivers will form.
For even a small life form and even starting with some fairly advanced molecules, there are a very large number of combinations that will never be more than rotting goo for every combination that is viable.
You are missing the point that any form of life is not simply one astronomical trial that went right but an astronomical sized array of trials, for each of which the odds of any result being compatible with life is astronomically small.
Numbers mean something -- you can't just wave your hand over them and dismiss what they are telling you. Go on to the link at the end as well.
I hear what you are saying but the distinction between “some outcome” and “any outcome” is wrong. That would never fly in fault trees, failure mode analysis or quantum mechanics. The requirements for a change of state are described almost exclusively by probability functions in those fields. The change from one species to another is a change through several states. Since the change could have gone in several directions or no where, those possibilities each has a probability that can be calculated.
Both Huxley and Hoyle performed their calculations using a state (endpoint). I’m sure others have too and I haven’t seen any criticism by anyone on any evolutionist web site claiming faulty logic for doing that. They were criticized for their assumptions concerning randomness and the role of natural selection.