Three Prisoners Problem ("probability paradox")

Discussion in 'Archives' started by Cruel__Machine, Nov 2, 2007.

Three Prisoners Problem ("probability paradox")
  1. Unread #1 - Nov 2, 2007 at 3:37 AM
  2. Cruel__Machine
    Referrals:
    100

    Cruel__Machine Guest

    Three Prisoners Problem ("probability paradox")

    There are three prisoners, A, B, and C. Two of them will be released and one will be executed. A asks the warden to tell him the name of one of the others in his cohort who will be released. As the question is not directly about A's fate, the warden obliges and says, "B will be released." Assuming the warden's truthfulness, what are A's and C's respective probabilities of dying now?


    NOTE: A and C do not have equal probability. (that's what I was initially thinking)

    I was reading some things on probability and found this (and other similar problems). I found it interesting and it gave me a better understanding of probability, so I thought I'd share.

    Thinking about it a bit before looking for the answer.
    Here a nice list of relating problems (includes this one as well):
    http://en.wikipedia.org/wiki/Category:Probability_theory_paradoxes
     
  3. Unread #2 - Nov 2, 2007 at 3:46 AM
  4. Osmosis
    Joined:
    Apr 29, 2005
    Posts:
    2,695
    Referrals:
    2
    Sythe Gold:
    0

    Osmosis Retired Australian Mod
    Banned

    Three Prisoners Problem ("probability paradox")

    It's a binomial I'm guessing?
     
  5. Unread #3 - Nov 2, 2007 at 5:36 AM
  6. Dark Whip
    Joined:
    Oct 27, 2007
    Posts:
    157
    Referrals:
    0
    Sythe Gold:
    0

    Dark Whip Active Member

    Three Prisoners Problem ("probability paradox")

    ^, I seriously don't get this stuff. I'm not really good with maths ;)
     
  7. Unread #4 - Nov 2, 2007 at 12:54 PM
  8. jaamal
    Joined:
    Feb 10, 2007
    Posts:
    1,713
    Referrals:
    1
    Sythe Gold:
    35

    jaamal Guru

    Three Prisoners Problem ("probability paradox")

    If the warden is tell the truth he has 50% chance of being released if it is that simple, if not it is a still 2/3 or a chance for freedom... Also I looked at the Berskon's paradox on the page and that would take forever to figure out with only deductive reasoning... But it couldn't be to large of a number I assume, I don't want to do the problem so if you do it tell me.
     
  9. Unread #5 - Nov 2, 2007 at 1:16 PM
  10. Phil
    Joined:
    Jan 21, 2007
    Posts:
    2,210
    Referrals:
    6
    Sythe Gold:
    5

    Phil Ex-Mod
    ♥ Will Love Govind Forever ♥

    Three Prisoners Problem ("probability paradox")

    What the fuck is this, that wiki doesn't make sense.

    To sum the prison problem up, it says this is the solution.

    A's estimated chances of dying are 1/3.

    The warden tells him that B will be set free (which is 1/3 people)

    Surely, if he knows B is going to be set free then that would leave two people who are up for execution. Which must give the chance 1/2 :S

    Where as wiki says...
    C's chances of dying are now 2/3, and A's are 1/3.

    If i'm an idiot for this sorry, and if someone could explain i'll be grateful.
     
  11. Unread #6 - Nov 2, 2007 at 1:49 PM
  12. Steph
    Joined:
    Apr 25, 2005
    Posts:
    2,680
    Referrals:
    6
    Sythe Gold:
    5

    Steph Azn Goddess / Ex-Admin
    Highly Respected Retired Administrator

    Three Prisoners Problem ("probability paradox")

    The key thing to note is this: The warden's answer has no effect on prisoner A's chance, meaning it will always remain at 1/3.

    So, if prisoner B is released, prisoner C now has a 2/3 chance.

    And don't forget this is a paradox: the answer seems to be something obvious (1/2), but it is not (1/3 or 2/3, depending on which prisoner you are referring to).

    It might help you understand the solution a bit better if you write out all the possible outcomes, which Wiki handily does farther down the page. However, the explanation also adds in another possible case (the guard can say B or C is being freed), so that might confuse you more =S
     
  13. Unread #7 - Nov 2, 2007 at 2:04 PM
  14. Phil
    Joined:
    Jan 21, 2007
    Posts:
    2,210
    Referrals:
    6
    Sythe Gold:
    5

    Phil Ex-Mod
    ♥ Will Love Govind Forever ♥

    Three Prisoners Problem ("probability paradox")

    I kind of understand it now, i was thinking more complecated than i should have.

    The original predictive chances were already measured taking into account what the guard said to person A.

    Thanks steph, and if you read this again read your PM's :p
     
  15. Unread #8 - Nov 2, 2007 at 2:12 PM
  16. jaamal
    Joined:
    Feb 10, 2007
    Posts:
    1,713
    Referrals:
    1
    Sythe Gold:
    35

    jaamal Guru

    Three Prisoners Problem ("probability paradox")

    The way your thinking never changes the probability of death, it was 2/3 of a chance before so if you know part of equation the outcome still comes out the same, if the evidence is true then it is 50% chance of freedom or death now, but that is if you factor in new evidence. Althoughhe chance are still the same as the start. Tell me if I am wrong.
     
  17. Unread #9 - Nov 2, 2007 at 2:19 PM
  18. Eric
    Joined:
    Feb 23, 2007
    Posts:
    3,344
    Referrals:
    13
    Sythe Gold:
    25

    Eric Grand Master
    Banned

    Three Prisoners Problem ("probability paradox")

    I've read it all, and I can't understand how A's chance are any different from C's

    I think math is just setting itself for epic fail :D
     
  19. Unread #10 - Nov 2, 2007 at 2:27 PM
  20. eXact
    Joined:
    Oct 22, 2007
    Posts:
    162
    Referrals:
    0
    Sythe Gold:
    5

    eXact Active Member

    Three Prisoners Problem ("probability paradox")

    EDIT: Ok read the wiki article and come up with this:

    It could very easily have been C that asked the question, since their circumstances do not differ except for the fact that A is the one that asked the question, hence p(A) = P(C) ?

    Im confused....
     
  21. Unread #11 - Nov 2, 2007 at 2:49 PM
  22. Shredderbeam
    Joined:
    Jan 26, 2006
    Posts:
    8,579
    Referrals:
    15
    Sythe Gold:
    664

    Shredderbeam Hero

    Three Prisoners Problem ("probability paradox")

    I can understand why the probability is 1/3 for A, but, if the problem is viewed as a whole, not from any individuals perspective, then wouldn't the odds still be 50:50 for A and C?
     
  23. Unread #12 - Nov 2, 2007 at 2:58 PM
  24. eXact
    Joined:
    Oct 22, 2007
    Posts:
    162
    Referrals:
    0
    Sythe Gold:
    5

    eXact Active Member

    Three Prisoners Problem ("probability paradox")

    Shredder, thats what i thought as well. Wiki the Monty Hall paradox, its kinda on the same lines and the diagrams make sense.

    Basically, its got to do with the fact that we're assuming the warden had to give an answer. If A and B were to be set free, the warden could only say B (he cnat tip of A), if B and C were to be free he would say either B or C, while C and A cant be set free. Thus the fact he chose B , favours the fact A and not C will be released. Correct me if im wrong any ;)
     
  25. Unread #13 - Nov 2, 2007 at 11:59 PM
  26. Cruel__Machine
    Referrals:
    100

    Cruel__Machine Guest

    Three Prisoners Problem ("probability paradox")

    The problem arises when you exclude the third person and don't refer to the original probability.
    I suppose I should've posted the Monty Hall problem, it's easier to understand:
    http://en.wikipedia.org/wiki/Monty_Hall_problem

    And if you still don't understand in that form, I could probably explain it.
     
  27. Unread #14 - Nov 3, 2007 at 12:49 AM
  28. jebckr
    Joined:
    Sep 15, 2007
    Posts:
    1,250
    Referrals:
    0
    Sythe Gold:
    0

    jebckr Guru
    Banned

    Three Prisoners Problem ("probability paradox")

    But if he had said C is to be released wouldn't the same apply to B? If so then A is getting favored chances just by asking the question. (assuming his fate won't be revealed.

    I think that after you look at all the crazy things you people are saying you'll still come back to 50:50.
     
  29. Unread #15 - Nov 3, 2007 at 1:40 AM
  30. Cruel__Machine
    Referrals:
    100

    Cruel__Machine Guest

    Three Prisoners Problem ("probability paradox")

    Yeah...
    Yes, asking the warden increases A's probability of being released.
     
  31. Unread #16 - Nov 3, 2007 at 2:02 AM
  32. Wallace & Gromit
    Joined:
    Jun 11, 2007
    Posts:
    84
    Referrals:
    0
    Sythe Gold:
    0

    Wallace & Gromit Member
    Banned

    Three Prisoners Problem ("probability paradox")

    hhhhmmmmmm. yes agreed
     
  33. Unread #17 - Nov 3, 2007 at 8:22 AM
  34. eXact
    Joined:
    Oct 22, 2007
    Posts:
    162
    Referrals:
    0
    Sythe Gold:
    5

    eXact Active Member

    Three Prisoners Problem ("probability paradox")

    A's chances stay the same; its C's that decrease ;)
     
  35. Unread #18 - Nov 3, 2007 at 8:38 AM
  36. Eric
    Joined:
    Feb 23, 2007
    Posts:
    3,344
    Referrals:
    13
    Sythe Gold:
    25

    Eric Grand Master
    Banned

    Three Prisoners Problem ("probability paradox")

    I think math just looses. It's obviously 50:50
     
  37. Unread #19 - Nov 3, 2007 at 9:33 AM
  38. X.Jake.X
    Joined:
    Nov 2, 2007
    Posts:
    122
    Referrals:
    0
    Sythe Gold:
    0

    X.Jake.X Active Member

    Three Prisoners Problem ("probability paradox")

    my brain hurts now lol.
     
  39. Unread #20 - Nov 3, 2007 at 10:39 PM
  40. Cruel__Machine
    Referrals:
    100

    Cruel__Machine Guest

    Three Prisoners Problem ("probability paradox")

    noes :'(
    math is winn0r.
     
< Trading Level 100 Main For Ddser!!! | selling amazing str/mage pure DT and MM done! ~look!~ >

Users viewing this thread
1 guest


 
 
Adblock breaks this site