Yep, I’m talking about the frequently referenced forced-choice scenario which exemplifies how and why we make choices.

Remember?  If not, here is an overview:

You and your Partner-in-Crime are captured by the police and put into two different interview rooms and enticed to divulge information that will convict your Partner and set you free.

This table demonstrates the traditional pay-off for the various combination of choice made by you and your Partner:

Your Partner

Cooperate

Compete

You

Cooperate

20 , 20 0 , 30

Compete

30 ,  0 10 , 10

I was first introduced to this game in a statistics course for my BS in Economics where it was presented as a lesson on odds-making.  It is one of the experiences my 20-year-old brain chose to store in it’s “reflect on this frequently” folder, and I have accessed it often.

The current opportunity for this reflection is the course by Howard Rheingold “Introduction to Cooperation Theory” in which the Prisoner’s Dilemma is presented not as a lesson in odds-making but as a lesson in why folks choose cooperation over competition (and choose competition over cooperation).

The course provided my fellow students and I with the game in three formats:

First format I will call “Opponent Personalities” because I played 5 games against opponents with 5 different personalities or “firms”.

Here is the link to this game, see how your experience matches mine:

http://www.gametheory.net/web/pdilemma

One thing I know about myself is that it is in my nature to cooperate even it reduces my personal gain.  So in Prisoner’s Dilemma I would always choose to cooperate even though it means forfeiting the large pay-off.

So, when I face Firm 1 in this game I begin with Cooperate and so do they.  We hold these positions for the entire game of 25 rounds and end with the same number of points.  I find this to be the perfect.  Notice that  this is quite a sizable penalty for being Cooperative because I could have had 10 extra points each round for a total of 250 additional points; moving my point total from 500 to 750.

A huge downfall to this game

This game does not let me know what portion of my income these points represent.  For example, if housing costs 500 points, I really need those 250 additional points for food, clothing, transportation, etc.  Would this affect my choice to Cooperate – YES!  Would this pressure to get the additional points mean that I would Compete, maybe.

In Prisoner’s Dilemma the bargain is for reducing your sentence, so you could argue that my example of choice being effected by the portion of my income these points represent is similar to the years-added-to-prison-term choice being effected by the portion of my life these years represent (am I old or young at the time of imprisonment), or the value of those years (do I have a young family to raise).

Back to the Opponent Personalities version of Prisoner’s Dilemma

Playing against firm 2

I continue to Cooperate and so does Firm 2 so again, after 25 rounds, we end with 20 point each

Playing against firm 3

Here things begin to get interesting because in round 2 Firm 3 decides to Compete and I do get 30 point to their Zero points but it makes me angry that Firm 3 is Competing, so for the next 3 rounds I Compete also and we each get 10 points.  In round 8 Firm 3 Cooperates meaning that they get all 30 points and I get Zero this makes me angry because I thought we could both keep Competing to share the (less but equal) points.  Rounds 9 – 25 I keep Competing and so does Firm 3.

Result in average points: Me = 10, Firm 3 = 13.6

Playing against Firm 4

I carry my disappointment from Firm 3 into my game with Firm 4 even though I know that this is a completely new partner.

I am a bit miffed that I earned less in round 3, so I am going to begin this round with the points on my side (30 points to start).  Then I go back and forth between Competing and Cooperating, trying to guess how Firm 4 will play so that we both win; but I am really happy when they Cooperate and I compete and receive the 30 point prize.

Result in average points: Me = 20.8, Firm 4 = 6.4 (I am secretly really pleased with this result.)

Playing against Firm 5

I begin again by being (generous) Cooperating, but they Compete, so I change into my Firm 4 strategy.  Firm 5 opts for Compete more often than Firm 4 resulting in a more even spread of points, on average.

Result in average points: Me = 9.2, Firm 5 = 11.60

End of game Summary (against all 5 Firms) me 80, them 71.60.  Hee Hee Hee, I’m the winner!

Second format played was called “Serendip” here is the link:

http://serendip.brynmawr.edu/playground/pd.htm

Same Prisoner’s Dilemma scenario, close to the same pay-out system.  There are two rather important differences between this and the Opponent Personalities format detailed above.

First difference:  That the number of rounds played in each game is not fixed (in the previous format there were 25 rounds).  The game ends when the “fiendish cyberspace wizard” decides to bring it to an end.  This changes the strategy if the player is planning their moves based on average winnings.

Second difference:  At the end of the game the “fiendish cyberspace wizard” supplies critique of how I played the game.  I was surprised to see how much these textual comments effected my play.

For example:

Game 1

I play my usual “Cooperate and hope that my Partner does the same” and indeed that was the case.  The “fiendish cyberspace wizard” ended the game after 18 rounds and said “You might want to try a new strategy”.  Of course, I understand the written meaning of “you missed out on bigger gains” and the unwritten message “you are not very smart to choose lesser gains”.

Game 2

I maintain my Cooperate every round, my Partner and I end with the same amount of coins.  The “fiendish cyberspace wizard” stops the game after 10 rounds and asks me if I think this is the BEST strategy; repeating that I might want to try a new strategy.

Game 3

I go back and forth between Cooperate and Compete and it ends with the “fiendish cyberspace wizard” saying”not bad, but I’ll bet you can do better”.

Game 4

I notice that my Partner is choosing to Compete a lot so I do too with this resulting critique from the “fiendish cyberspace wizard” saying I won but that I was “flirting with an Inconceivable foul fate”. Pretty harsh coming from a faceless machine!

I understand that the game wanted me to take more risks to obtain possible rewards, but that is not in my nature.

The third format is called Iterated Prisoner’s Dilemma it’s link is:

http://www.iterated-prisoners-dilemma.net/

By this time I was pretty tired of the virtual interview room and trying to out-maneuver my virtual Partner, so my play was a bit dull.  This format also critiqued my play at the end of each game.  The advice was:  “The best possible Mutual Outcome Rating was not achieved.”  So, I guess I employed my “Cooperate always” strategy in this format also.

Advertisements