[PhatHead]

Well, again, I think it's fair to argue with his reasoning. I haven't read any of his work, so I really couldn't argue it one way or the other.

Has he written about evolution, or just about origin of life/matter/universe? And does he assume a different set of requirements for the first life form than what is generally accepted?

Apologies if you haven't read him, either ... I've read critiques both pro and con and found neither very satisfying. I find the idea of a statistical analysis very interesting, but it strikes me as an extraordinarily complex matter to apply statistics to the origins of life or the universe... too many unknowns, it seems to me, to be able to really determine an "impossibility" threshold in the first place. But that's just me talking.

[Dimensio]

And does he assume a different set of requirements for the first life form than what is generally accepted?

The problem is that he assumes a specific requirement for the first life form, when there's no reason to assume that the first life form must be exactly as he claims. There are diverse hypothesis on the nature and origin of the first life form. Dembski assumes that there's only one possible configuration that would have worked -- when few if any biologists even claim this -- and he starts his calculations from there. He's imposing an artifical restriction with no basis in known reality to keep the result so improbable.