I think a better way to describe it is ven diagrams:

conditional probability is just like, what proportion does A represent given B has already happened.

A might be small in the ven diagram box, but take up a larger area when constrained to only the part that B is in