Re: Prisoner's Dilemma

Michael M. Butler (butler@comp*lib.org)
Sun, 29 Jun 1997 21:52:16 -0700


FWIW, a single Prisoner's Dilemma is a pretty threadbare situation.

The _Evolution of Cooperation_ model assumes
an _iterated_ Prisoner's Dilemma, for starters.

And the consequences have to matter, and you have to be able to identify
your counterpart, and rough parity has to exist, and you have to be able to
not tell if this is the last time you'll ever encounter the other party
(these last two are almost the same thing, looked at in different ways).

That's off the top of my head. Given those things, cooperation or
tit-for-tat are sound evolutionary strategies. _Not_ given them, they are not.

At 08:14 PM 6/29/97 -0700, you wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>
>Long ago on Sun, 01 Jun 1997 Joao Pedro <jpnitya@mithlond.esoterica.pt>
Wrote:
>
> >I'm new to this list so I hope I'm doing this well.
>
>You're doing fine.
>
>
> >The best way for mankind to evolve is with unselfish actions. I
share
> >Richard Dawkins' opinion that the individual is the most
important
> >but for each individual to better prosper and achieve pleasure,

> >cooperation is the best path and selfishness is the worst. It's
like
> >a Prisoner's Dilemma, the best choice is cooperation and
defection
> >will bring ruin to us all. This is not altruism, it is logical

> >reasoning with the objective of determining what is better for my

> >personal existance.
>
>
>That's a very noble idea but it's not quite that simple in a typical
>Prisoner's Dilemma situation. Suppose that if you and your partner you can't
>communicate with both cooperate then you both get 10 days in jail, if you
>both defect then you both get 10 years in jail, if your partner cooperates
>and you defect then you go home today with no jail time at all, and if your
>partner defects and you cooperates then you get 50 years in jail.
>
>Obviously to minimize the total amount of time anybody spends in jail the
>best strategy would be for everybody to cooperate, and if there was some way
>to form a binding agreement with my fellow prisoner I most certainly would,
>BUT there is no way to communicate or influence what the other prisoner
will
>choose to do, and regardless of what he does I'm always better off if I
>defect.
>
>When I say "better off" I assume I don't love your fellow prisoner, if I
did
>my strategy would change dramatically, but you must take that love as an
>axiom, you can't derive it from the prisoners dilemma. At least you can't
>derive it at a level as high as the individual, at the gene level perhaps.

>
>
> John K Clark johnkc@well.com
>
>
>-----BEGIN PGP SIGNATURE-----
>Version: 2.6.i
>
>iQCzAgUBM7ciC303wfSpid95AQFD2ATvXwcxZh63AiABcyfJg0i6pZXFUdpLDqG3
>FhaBY3ijpln6v91ve2ztb7BQaOf5V66nYYYACxxNZJn2PBW8pTmkER59kmQ+Q5hQ
>OOacBeazytiIRfFFSTJUx938M9W55lDRmnmo9gw6NR9AgDINo/DkmBAjxHKlAN6F
>BCLmE3NRm01QnVp7ab54XtI1VqPo6CQnY2/xveHuY9i9T88tkjA=
>=7mPt
>-----END PGP SIGNATURE-----
>
>
BOUNCE WARNING: A simple reply to the above address will fail. If you wish
to send me a _noncommercial_ message, kindly substitute a hyphen for the
asterisk.