Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

A Six Dice puzzle

  1. Aug 13, 2007 #1
    This is reminiscent of similar dice riddles, but the answer to this one appears to defy logic. Unless I'm mistaken, this is not as obvious as it looks.

    There are six unloaded die.
    Each has been rolled out of your sight and hidden behind a cup.
    You are informed that no more than one of them landed as 6 - but it may be that none of them did.

    You are going to remove the cups one at a time.

    Before you do, you label each cup with letters A to F. This designates the order in which you will remove them. You have complete discretion which cup is assigned each letter, but once labeled they must be removed in alphabetical sequence.

    Having labelled them, it is time to remove the cups accordingly from A to F.

    If one of the dice landed as 6, under which cup is it most likely to be found?
    Last edited: Aug 13, 2007
  2. jcsd
  3. Aug 13, 2007 #2
    Under none. You just removed all the cups. Please.
  4. Aug 13, 2007 #3
    Sorry. It's not that kind of riddle! :smile:

    I'll rephrase.

    "Having labelled them, you will remove the cups accordingly, from A to F.
    Under which cup is the dice that landed a 6 most likely to be?"
    Last edited: Aug 13, 2007
  5. Aug 13, 2007 #4


    User Avatar
    Science Advisor

    This sounds like a trick question, but my guess is all are equally likely.
  6. Aug 13, 2007 #5
    I think it's a question with a counter-intuitive answer.

    If I'm correct, they are not all equally likely and one of the cups has a higher probability of revealing the 6 than each of the others.

    But there may be a flaw in my logic.

    Also I amended the wording shortly after I first posted it. The clue may lie in not knowing whether any cup holds a six.
    Last edited: Aug 13, 2007
  7. Aug 14, 2007 #6


    User Avatar
    Science Advisor
    Homework Helper

    There are [itex]5^6[/itex] ways for the dice to land without any 6s. There are [itex]5^5[/itex] ways for each die to land on a 6. Thus the chance that any die lands on a 6 is 1/11 and the chance that no die lands on a 6 is 5/11.
  8. Aug 14, 2007 #7


    User Avatar
    Science Advisor

    But I don't see how that has anything to do with Simon 6's assertion that a "6" is more likely to be under a specific one of the cups.
  9. Aug 14, 2007 #8
    Instead just saying they are equally likely, I prefer you answer me:
    Why should one cup be different from the others?
  10. Aug 14, 2007 #9


    User Avatar
    Science Advisor
    Homework Helper

    If it's right, he's wrong. I tried to make it as transparent as possible.
  11. Aug 14, 2007 #10
    I presume it will have something to do with your outcomes changing as you remove each cup. In the extreme, "if I removed all but one cup and still no 6, then the last cup has 100% blah blah".

    For I can't see the a priori probability of having a six somewhere to be altered by the cup removal order.
  12. Aug 14, 2007 #11


    User Avatar
    Science Advisor
    Homework Helper

    I'll admit that I actually though the probabilities would differ, until I actually tried to model it. It seems to me that it should be symmetric.
  13. Aug 16, 2007 #12
    What you have to do, I think, is find the sample space. Now we can consider the binomial expansion where S=6,andN = not a six: [tex]N^6+6(N^5)S+15(N^4)(S^2)+20.....[/tex] but of all the variations, we are only interested in the first two above: n^6+6(N^5)S, this is in fact only 7 varations, but the probability of (5/6)^6+(5/6)^5 is about 74% of all cases.

    So all we have to do is consider the case of all N's = NNNNNN, about 33%, and then consider 6 more cases as the S moves from the first position to the last, about 40% total, 6.67% for each position. Thus if we had 100 trials under consideration, we would exclude all but 74 of them, and then find 6.67/74 = 9% of those would have a 6 in a given position. or to be exact we could use CRGreathouse's figures:

    [tex]\frac{(5/6)^5*1/6}{(5/6)^6+(5/6)^5} =1/11[/tex]
    Last edited: Aug 16, 2007
  14. Aug 19, 2007 #13
    Having made further calcuations, I draw the following conclusion

    The question as stated - where you're told in advance that no more than one dice landed 6 - does yield the intuitive answer: all have equal probability. I was mistaken to think otherwise.

    However: if you're not limited to rolling one 6, a very counterintuitive answer applies.
    Suppose you're informed: "Two die landed 6".
    Everything else in the puzzle stays the same.
    You label the cups A-F in the order you will remove them.
    The question is then asked: "Under which cup are you most likely to discover a 6?"

    The answer is always "A".

    In other words: before you start removing cups, whichever cup you choose to lift first is most likely to contain one of the 6s. If cup "A" is found not to have a 6, then the highest probability moves to cup "B". This continues alphabetically until the first 6 is discovered. After that, the remaining cups have equal probability.

    That may seem unbeleivable.

    Do others agree?
  15. Aug 20, 2007 #14


    User Avatar
    Science Advisor
    Homework Helper

    The probabilities should be
    A: 1/3
    B: 4/15
    C: 1/5
    D: 2/15
    E: 1/15
    F: 0
  16. Aug 20, 2007 #15


    User Avatar
    Science Advisor
    Homework Helper

    I'm not sure it's counterintuitive. If you're told nothing, A is still the most likely and F the least likely. The only cases when they should be equal is when you know there are less than 2 sixes (3 cases).

    0-6 sixes:
    A: 1/6
    B: 5/36
    C: 25/216
    D: 125/1296
    E: 625/7776
    F: 3125/46656
  17. Aug 20, 2007 #16
    This sounds to me more like conditional probability, something on the order of:
    P(A) = 1/3
    P(B|~A) = 2/5
    P(C|~(A.B)) = 1/2
    P(D|~(A,B,C)) = 2/3
    P(E|~(A,B,C,D)) = 1
    [Edit: corrected this last one:]
    P(F|~(A,B,C,D,E)) = 0/0 (undefined)
    Last edited: Aug 20, 2007
  18. Aug 20, 2007 #17
    I don't think it's counterintuitive at all. If I have two cups and put a marble under each, then I'm most likely (i.e certain) to first find a marble under the first cup I look under. Now the probability that there is a marble underneath is the same for both cups (i.e. 1), just as the probability that there is a six underneath is the same for all cups in the question posed.
  19. Aug 22, 2007 #18
    Naturally! Certainties dispel counter-intuition.

    Translate this to six cups hiding two marbles.

    You're about to lift the cups one at a time from A to F.

    I ask: "Under which cup are you most likely to find the first marble?". Most people won't assume it's the first cup you lift.

    However, suppose the initial question is: "Under which cup are you most likely to find the second marble?".

    Again, I beleive the answer will surprise most people.
    Last edited: Aug 22, 2007
  20. Aug 23, 2007 #19


    User Avatar
    Science Advisor
    Homework Helper

    I'd expect the last cup, naturally.
  21. Mar 27, 2009 #20
    There are six dice, one under each cup. The number of sixes is less than or equal to one.
    The odds of finding a six under cup number A are therefore (less than or equal to)one in six.
    Either it is a 1:6 chance or there is a 0% chance depending on if there is even one six out of all the dice!

    It then follows that as the cups are eliminated, the odds of a six being revealed increase.

    Cup B = 1:5, C = 1:4, &c. so that if you eliminate the first five cups, and there is indeed a six, the last cup has a 100% probability of containing that six.

    This is similar to another riddle/math problem I call "The Game Show" You choose one of three doors, two of which have nothing of value behind them, one containing the "grand prize". The game show host shows you one of the two doors tht you did NOT pick contains nothing. So now it is either the door you picked originally or you could change your guess to the other door. Do you stick to your original guess or switch?

    The best bet is to switch because your first guess you had a 1:3 chance of getting it right. Now you have a 1:2 chance so it would be foolish not to change your bet!
    Think this doesn't have logic to it? Okay, take it on a grander scale. Let's say you pick a number between 1 and 1,000,000,000. Let's say you pick 42. Now the administrator of the "lottery" tells you that they have eliminated every number except 42, 576,341,100, and 923,128,001. The grand prize is $500,000 and you can now either stick to your original guess or switch to either two of the remaining numbers. What do you do?

    Same thing, just more extreme odds. You went from a 1:1,000,000,000 chance to a 1:3 chance.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook