Argument 2, I figure. Finding out if a randomly chosen ace is a AS or not tells you nothing about p(2A) - since the chance of choosing AS is 50-50 for both 2A and across all 4 of the !2A cases.
That’s assuming a genuine random choice. If there’s a bias that causes the choice to become non-random once hearing the AS question, that might mess things up.
Argument 2, I figure. Finding out if a randomly chosen ace is a AS or not tells you nothing about p(2A) - since the chance of choosing AS is 50-50 for both 2A and across all 4 of the !2A cases.
That’s assuming a genuine random choice. If there’s a bias that causes the choice to become non-random once hearing the AS question, that might mess things up.