Perry: Its not just an oppinion. To me proving something is true in the natural world means you want to talk about the probability that it is true because thats really all we can do. This view is the same as advocated by eg. Jaynes if you want me to slap a fancy name on it.
As i wrote, probabilities is a subjective thing and depends on your priors. what we DO have is a method of combining probabilities, basically bayes law, and i can point you to various proofs that show why this is a pretty fundamental thing.
I hope you agree with me so far - eg. that when you/me are saying "creation/evolution is true!" we really mean "the probability that creation/evolution is true is very very close to 1" and that it makes sence to talk about probabilities of something being true.
Now, we cant calculate the likelihood of creation or evolution in exact numbers. But we can, at least qualitively, answer questions like: "Does this observation increase or decrease the likelihood of evolution/creation". It turns out that the property we are interested in, in order of making the evidense relevant, is loosely stated: a) Does the theory predict/exclude the observation and b) Is it non-trivial.
So thats just my view. I think this is pretty compatible with Popper, as i think that if a theory make any non-trivial predictions, it is also falsifiable. For example, if my theory said: "Bananas are made of wood", that would automatically make it falsifiable. But i am no expert here, and i dont know if i impose more loose or more strict restrictions on what you would call a scientific theory than Popper does.