home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky sci.math:17085 rec.puzzles:8000
- Newsgroups: sci.math,rec.puzzles
- Path: sparky!uunet!uunet.ca!wildcan!sq!msb
- From: msb@sq.sq.com (Mark Brader)
- Subject: Re: Aces and Queens (P.S. to: The paradox of the second ace)
- Message-ID: <1992Dec17.095146.13318@sq.sq.com>
- Followup-To: rec.puzzles
- Organization: SoftQuad Inc., Toronto, Canada
- References: <1992Dec15.055832.26324@galois.mit.edu> <COLUMBUS.92Dec15111255@strident.think.com> <COLUMBUS.92Dec16100320@strident.think.com>
- Date: Thu, 17 Dec 92 09:51:46 GMT
- Lines: 72
-
- SPOILER for the aces and queens question, at the end.
- Followups are directed, somewhat arbitrarily, to rec.puzzles.
-
- > This paradox is discussed in one of Martin Gardner's collections.
-
- Can someone identify which one, or provide information about the
- original appearance of this column?
-
- > probability (a), below, is less than probability (b):
- > (a) P(2nd ace | any ace), that is, the probability that a randomly chosen
- > bridge hand has a second ace, given that it has at least one ace
- > (b) P(2nd ace | ace of spades), that is, the probability that a randomly
- > chosen bridge hand has a second ace, given that it has the ace of
- > spades
-
- Here's a way to take some of the surprise out of this result.
-
- From a truly randomly permuted deck, you may as well "deal" the bridge
- hand by taking the first 13 cards off the top of the deck. (Do not
- try this form of dealing at a bridge club, kids!) Then the various
- probabilities involved are probabilities as to HOW HIGH IN THE DECK
- the various aces were located before the "deal". Number the positions
- from 1 (top of the deck) to 52 -- then positions 1 to 13 make the hand.
-
- Okay, consider probability (a). Advance through the deck looking at
- successive cards: position 1, 2, etc. If you haven't hit an ace by
- position 13, you throw out the "deal" and reshuffle. If you do hit
- one, then probability (a) is the chance that you will hit *another*
- ace by the time you get to position 13.
-
- Now consider (b). Start the same way: look through the deck card by
- card. If you haven't hit the ace OF SPADES by position 13, you throw
- out the "deal" and reshuffle. If you do hit it, then probability (b)
- is the chance that EITHER you will hit another ace by the time you
- get to position 13, OR you hit one BEFORE you got to the ace of spades,
- or both.
-
- See? Now it's not at all surprising that (b) is larger. The key thing
- is that in finding out that there is an ace of spades in the hand, it
- doesn't have to be the FIRST ace you come to.
-
- > (I wonder, is this something any decent bridge player would know by
- > instinct?)
-
- Well, I'm a decent (not expert) bridge player, and I found it 100%
- counterintuitive, until I thought of the above line of reasoning.
- Probability questions arise quite often in actual bridge play, but
- that is not one that would realistically do so.
-
- A simple example of a question that does commonly arise (from a hand I
- played a few hours ago), is "given that East and West have between
- them the King, Queen, and two other diamonds, what is the probability
- that West either has exactly two, or has the King or Queen and no
- others?" (About 53% -- fortunately, she did.)
-
- > Here is a variant I haven't seen elsewhere. Consider:
- > (c) P(ace | queen), that is, the probability that a randomly chosen
- > bridge hand has an ace, given that it has a queen
- > (d) P(ace | queen of spades), that is, the probability that a randomly chosen
- > bridge hand has an ace, given that it has the queen of spades
- > Despite what you might guess, (c) and (d) are *not the same*.
-
- Just as with aces, a hand with the queen of spades is more likely to
- have a second queen than just any hand with a queen is. A hand with more
- queens is less likely to have more aces -- there are fewer "slots" left
- to fill. Therefore, (c) is the higher probability. (That was fun.)
- --
- Mark Brader "Many 'business-oriented' packagings of these dialects ...
- utzoo!sq!msb omit the games section. Those responsible will doubtless be
- msb@sq.com reincarnated as worker insects of some sort." -- "J. E. Lapin"
-
- This article is in the public domain.
-