Maybe even _that_ reasoning relies upon simple hypotheses being more likely. In this case, the simple, and therefore favoured, hypothesis would be that "models with a simplicity that is similar to that of other successful models are more likely to be true".

Assuming that other successful models are also true and not just a wolf in sheep's clothing.

Let's forget that I said "true", and switch it to "successful" instead.

That sounds like a more complex hypothesis to me.

Maybe it's simpler than "models with a simplicity that is dissimilar to that of other successful models are more likely to be true", though it's not obvious what should be considered simple and not.

Kolgomorov complexity seems like the obviously correct measure, though AIUI it's only defined relative to some fixed Turing machine.

I have an idea of doing a PageRank-style weighting of Turing machines, based on how simple/complex simulating one under another is. That would align with my intuition that most basic universal Turing machines are equally valid, but one that has (say) a short instruction that outputs one particular long sequence of symbols is probably not a valid machine to use.