home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!uknet!comlab.ox.ac.uk!akay
- From: akay@comlab.ox.ac.uk (Andrew Kay)
- Newsgroups: sci.math
- Subject: Re: "Cut & Choose" for several players
- Message-ID: <1993Jan21.101533.19001@acis.comlab.ox.ac.uk>
- Date: 21 Jan 93 10:15:33 GMT
- References: <43725@sdcc12.ucsd.edu> <1993Jan19.213201.24197@zip.eecs.umich.edu> <43773@sdcc12.ucsd.edu>
- Organization: Oxford University Computing Laboratory, UK
- Lines: 37
- Originator: akay@acis.comlab
-
- Anthony Minkoff writes:
- >Now I want a method for distributing the pile among n players, so
- >that no player feels that any other player received a better
- >portion.
-
- and also
- >I.e., the statement "each player feels he got a 'fair' share" is not
- >a sufficient condition in my formulation of the problem. Rather, it
- >is necessary that *no player feels that _any other player_ received
- >a better portion.*
-
- It is not clear to me whether a player is judging (subjective)
- values of shares in themselves, or (subjective) values of
- shares to particular players.
-
- For example, three players A, B, C and shares x, y and z.
-
- A might like x a lot, and be happy for B to get y and C to
- get z. However, A knows that y is of particular advantage to
- C and z to B, so would not be happy for B to get z and C to
- get y.
-
- Actually, what is to stop B and C covertly swapping y and z
- after A has gone home ? I suspect that this question uncovers
- a problem with the whole question of A being satisfied about
- the way B and C share y and z.
-
- I seem to remember that games with more than two players are
- very hard to analyse mathematically, because of the formation
- of coalitions which don't fall within the rules of the game.
- B may try to help C win a game in return for a share of the
- winnings, or to get a date for example !
-
- Interesting puzzle.
-
- Andrew Kay,
- Email: Andrew.Kay@prg.oxford.ac.uk
-