I've reviewed Jason Rosenhouse's book on the Monty Hall problem. Overall summary: The book is very good and does an interesting job of discussing not just the original problem but variations of it, as well as what reactions to the problem can teach us about how humans estimate probabilities.