>Your reasoning would look fine if it doesn't lead to the same >probability of 0.5 on the original problem. That why the formula of >conditional probability have been used, namely to avoid the "paradox". Oops...:-( :-). Huy