School sign in Sarajevo

the prisoner's dilemma

looking at

Which way to peace?
Nature of peace education
Peace education in the post cold war era
Alternative futures
Educating for a sustainable future
Towards a peace education curriculum.
Democratic education
Books, references and resources
on peace education
Understanding conflict

email page to a friend

The prisoner's dilemma has attracted not only games theorists and evolutionists, but military strategists as well. Our version is as follows:

Adam and Eve are in isolation cells, suspected of a significant crime. ‘The Governor’ visits them in turn, inviting each to grass on the other. Neither knows what the other has done. There are four possible outcomes.
Adam: Eve did it, and I’ve got proof BETRAYAL (of Eve)
Eve: I’m not saying anything CO-OPERATION (with Adam)
Adam is released. Eve gets the maximum sentence

Adam: I’m saying nothing CO-OPERATION (with Eve)
Eve: It was Adam who did it BETRAYAL (of Adam)
Eve is released. Adam gets the maximum sentence

Adam: Eve did it, and I’ve got proof BETRAYAL (of Eve)
Eve: It was Adam who did it; I’ve got proof BETRAYAL (of Adam)
Both receive medium sentences, rewarded for giving evidence

Adam: I’m saying nothing CO-OPERATION (with Eve)
Eve: I’m not saying anything CO-OPERATION (with Adam)
Both receive negligible punishment, through lack of evidence

The prisoners’ respective self-interest dictates that both see betrayal as a way out; but both are therefore likely to betray and suffer. If both co-operated they would suffer less. ‘The Governor’ hears them muttering ‘You bastard!’, or perhaps, ‘If only I’d known you’d say nothing!’ – but trust in these circumstances is hard to come by.

Translated into a two-card ten-round gambling game, the dilemma is more marked.

A co-operates

B co-operates

Banker gives both a modest sum

A sells out

B co-operates

Banker gives A a large sum and B a fine

A sells out

B sells out

Banker takes a small fine from both

A co-operates

B sells out

Banker gives B a large sum and A a fine

The Selfish Gene.
Richard Dawkins (Oxford)
Winers and other Losers in Peace and War.
Arnold Arnold (Paladin)
Christmas Truce.
Malcolm Brown and Shirley Seaton (Papermac)
The Evolution of Co Coperation.
R Axelrod(Basic Books)

A bright player will quickly see that it will pay to play the ‘Sell Out’ card every time. If the opponent is bright too, they’ll both play ‘Sell Out’ until such time as they can swallow hard and play ‘Co-operate’ – hard to do when you know that if your opponent hasn’t played that card he will be in the money.

Games theorists set to work with their computers to see what strategies might be devised to crack the dilemma. They enjoyed themselves co-operating in this enterprise. Since the idea was to win, the result was unexpected: the most successful strategy IF ENOUGH ROUNDS WERE PLAYED was one in which the player (a computer)

a. was never the first to play ‘Sell Out’
b. quickly forgot that the opponent had played ‘Sell Out’ (so no continuing reprisals)
c. did not care if it won less points/money than its opponent
d. co-operated with the other player to fleece the ‘bank’ rather than each other

The informal truces in the First World War are a clear example of this kind of reciprocity. The soldiers co-operated in the interests of everyone’s survival, and against the interests of their commanders (or ‘war’ itself).
Biologists point to many examples of species that work together on the same principle that co-operation is best for all: the price for it is lower than the cost of competition.

From: WORKING TOGETHER a handbook for co opeartion

Peace Pledge Union 1, Peace Passage, London N7 0BT. CONTACT