Something that I’ve been chewing on (for months, really, but it became more stark recently, thanks to a discussion with ): the nature of skill in probabilistic domains is really different from the skill in non probabilistic ones.
Conversation
Replying to
For instance, when training a new manager, I can tell if there’s improvement when they deal with a certain class of personnel problems and they come out ahead. I don’t need to see more than 1 or 2 instances of this to know that they’ve ‘solved it’; it’s not going to be a problem.
1
But there are whole categories of things where even 1 or 2 instances of success gives you very little information on whether the person really has the skill to do so. Investment is like this, as is 1 or 2 hands of poker.
1
When I wrote my tacit knowledge series, I said that what I was interested in was what we called expert intuition.
But investment also contains a component of skill that is tacit. It’s not expert intuition — but it *is* tacit, so what is it?
1
1