What experts cannot tell you
Ask someone who is genuinely good at something how they do it, and you will get an answer. It will be fluent, confident, and largely wrong.
This is not dishonesty. It is what happens to a competence as it matures. The steps that once required attention stop requiring it, and once they stop requiring attention, they stop being available to introspection. The expert is not withholding the method. They no longer have access to it.
Why asking harder does not help
The usual response is to press: ask more precisely, ask again, ask a better question. It rarely works, and the reason is structural. You are asking someone to report on a process that has been removed from the part of the mind that reports.
What comes back instead is a reconstruction. A plausible story, assembled after the fact, about what must have happened. It sounds like a description. It is a theory, and it is usually the theory the person was taught rather than the one they use.
What modeling does instead
Modeling starts somewhere else. Not with the account, but with the doing.
You observe the competence where it actually happens, under real conditions. You look for the distinctions the person makes without noticing that they make them, the order in which they do things, the moment they decide something has gone far enough. Then you build a structure out of what you saw, and you test whether that structure produces the result in someone else’s hands.
That last step is what separates modeling from admiration. A model that cannot be handed to another person is not a model. It is a portrait.