Check out this post at Choice and Inference if you liked my earlier posts on probability puzles (here, here, and here).  The puzzle is given as a possible counterexample to reflection for credences, but is worked out nicely in the comments over there.  I was too late to the party to add what I learned from jd2718, but the solution to the second ace puzzle is pretty much the same as the solution to the dice puzzle that had initially fooled me when jd2718 presented it to me in comments.