Name: Allen Hathaway
Subject: Prisoner's Dilemma
Date: Sun Feb 22 09:56:09 EST 1998
It has been empirically shown (though I'm not sure mathematically proven) that a strategy of TIT FOR TAT will yield the maximum payoff in an iterated prisoner's dilemma. I have done a study for my anthropology class on a multi-player version of the prisoner's dilemma (free rider dilemma) and noted a difference between genders. I'm now trying to put together a study on a true (one shot) prisoner's dilemma, as well as chicken, deadlock, and stag hunt. I will be looking for biases based on gender, age, and economic status. Any info or references on other work in this area will be appreciated.
Name: Allen Hathaway
Subject: Feedback
Date: Sun Mar 22 20:11:13 EST 1998
Wow, this is a lively site!
Name: Chris Ramshaw
Subject: Prisoner's Dilemma
Date: Sat Oct 24 08:31:18 EDT 1998
Serendip seems to be using a strategy of Tit for Tat in its simulation of a Prisoner's Dilemma game. This makes it very hard to do get an average of more than 3 coins. Given that n turns are required, a good idea might be to co-operate until turn n-1, and then defect on the final turn - this means that Serendip does not have a chance to 'repay' the defection. An extensive discussion of computer simulated Prisoner's Dilemma strategies can be found in Douglas Hofstadter's 'Metamagical Themas', pp.715 - 766.
Name: Ryan Brooke
Subject: Prisoner's Dilemma
Date: Sun Nov 1 23:35:08 EST 1998
I'm not thoroughly familiar with game theory, but prisoner's dilemma is a lot like a game called "win as much as you can." Using the strategy of cooperating every turn yielded 12 coins, and I tied with serendip. The second (and last) time I played I cheated the first turn, then cooperated every turn after that. I beat serendip by three coins, I believe, using that strategy.
Name: meg
Subject: huh???
Date: Fri Mar 5 04:35:53 EST 1999