Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.3 alpha 4/15/85; site mordred.purdue.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxt!houxm!ihnp4!inuxc!pur-ee!pucc-j!purdue!chk
From: chk@purdue.UUCP (Chuck Koelbel)
Newsgroups: net.singles,net.women
Subject: Re: Win/Win Players vs. Win/Lose Players
Message-ID: <546@mordred.purdue.UUCP>
Date: Thu, 27-Mar-86 20:33:11 EST
Article-I.D.: mordred.546
Posted: Thu Mar 27 20:33:11 1986
Date-Received: Sat, 29-Mar-86 16:03:32 EST
References: <1270@decwrl.DEC.COM> <439@ccivax.UUCP> <12@umcp-cs.UUCP> <12588@ucbvax.BERKELEY.EDU>
Organization: Department of Computer Science, Purdue University
Lines: 23
Xref: watmath net.singles:11259 net.women:9892

In article <12588@ucbvax.BERKELEY.EDU>, weemba@brahms.BERKELEY.EDU (Matthew P. Wiener) writes:
> In article <536@mordred.purdue.UUCP> chk@purdue.UUCP (Chuck Koelbel) writes:
> >What I wanted to point out is that "playing nice" like "tit for tat" does
> >is not always the best strategy.  In a replay of the computer simulation
> >with different (sometimes more complicated) strategies, "tit for tat" did
> >not win - a program that "probed" its opponents to see their reactions
> >did better.  The conclusion is that no one strategy is best. 
> 
> What do you mean by "best strategy"?  "tit for tat" is best against ALL
> the competition.  In other words, it can be beaten, but only barely.  The
> other known strategies, including the "prober", will lose spectactularly
> against certain other strategies, and so on average, do worse than "tit for
> tat".
> 

I meant "best" in the sense that no strategy always has the highest total
at the end of the tournament.  (I believe this was what the original message
meant, as well as the way the tournament was judged.  But it's been a while
since I read the article, so I may be wrong.)  So the "win/lose" players out
there may not want to use tit-for-tat.  For us "win/win" players, it is still
a good option.

							Chuck Koelbel