??? 04/24/04 12:50 Read: times |
#69108 - Coorelation is the issue here Responding to: ???'s previous message |
Joseph Hébert wrote:
Nope. You ask either man which door the other will say is safe. Case 1: The guy you ask always lies. He then tells you that the honest man will tell you that the deadly door is safe. You take the other one. Case 2: The guy you ask always tells the truth. He then tells you that the other man will tell you that the deadly door is safe. You take the other one. In either case, either man will tell you that the other man will point you toward the deadly door. Yes, sure, Joseph, but the man who always lies MUST know, that the other man will always tell the truth. And the honest man MUST know, that the other man will always lie. What I and Raghunathan mean is: There are two unknown quantities: 1.Which door is safe? 2.Which guy will tell the truth? If there's absolutely no correlation between involved quantities, it's impossible to solve the problem with only one question. The problem then is underdetermined. The fact, that the two men know each other, means that each man knows, how the other would answer a question gives the needed correlation, so that the problem can be solved by one and only one question. I was interested in whether there's some kind of question which could introduce this correlation also, so that, even if the men do not know each other, the problem can be solved by one and only one question. Raghunathan stated, that there's no way. And I think I can trust him... Bye, Kai |