Game Theory
Game theory! This really is an ideal way to introduce some
mathematical, logical thinking in a way which IMO would be a definite asset
to us. How do 'non-zero sum games' like the prisoners' dilemma relate to
the online give and take?
But we have to distinguish constructive, initiative-taking
participation in a list (perhaps some risk-taking), from a more boorish,
self-interested, combative kind of domination at
the expense of the group's purpose and attention,
or resources. Are there objective criteria for this or is it only "in the
eye of the beholder" which side a list's activity falls on?
The neat thing about the prisoner's dilemma as a model of
cooperative behavior, simple as it is, is that it's utterly amoral -- solely
based on the question "how do I get the maximum personal benefit out of this
ambivalent situation -- whether to cooperate with the other player, in which
case both get a moderate reward (taking turns on a list?) or whether to 'defect'
which in the model means gambling on grabbing the maximum reward
(attention/bandwidth?) which only works if the other player has gambled on
cooperation and the moderate reward(let him/her get away with
it?)
.
So at least in this model a moderate, cooperative approach
doesn't depend on satisfying an arbitrary standard of etiquette or anyone's
personal taste -- the outcome is based only
on the participants struggling to find out what works out best for them
individually. No imposed rules or regulations other
than "what's ultimately going to work out best for you?"
I believe that Hobbes gave us a presocietal anthropological fantasy in which
all sorts of mayhem occurred because each individual was trying to advance
his or her own narrow interests. It was a dog eat dog (not dogfood) world,
and you could be killed on the streets for your Nikes. Dissatisfied with
this state of affairs the people agreed that they would exchange what advantages
they got from pursuing their narrow interests for the stability of the
cooperative alternative in the prisoners dillemma. Government, a monarch
in Hobbes case, a listowner or moderator in
ours, was given the task of enforcing the cooperative
alternative.
Note that government also enforces the noncooperative choices
in the prisoner's dillemma. Businesses long ago noticed the advantages of
the cooperative choice. Rather than pursue individual interests, they joined
together to set prices and divide markets so that each got a reasonable piece
of the pie. Everyone was happy except the consumer who had to live with high
prices, bad service, and lack of choice. Government stepped in with antitrust
laws.
A Medline search on 'game theory', citations since 1994, yielded over 400
hits, the first 25 of which I downloaded. It's about 30k in length so I won't
inundate the list with it but will send it to James and anyone else who's
interested in it.
In the prisoners dilemma, public safety is factored out, and when these two
naive prisoners both take the altruistic (or silly) alternative, they both
escape doing the time for the crime and will be out on the streets burglarizing
people again shortly. The prisoners benefit greatly from the cooperative
option, but society,if it has any stake in truth and
justice, suffers.
When one declares the interests of victims and neighbors
to be irrelevant to the problem in the prisoner's dilemma, one comes close
to making deception a virtue. When a list, through its listowner or by other
methods, declares a point of view uninteresting or "not our thing," it is
withdrawing rather than confronting and risks a result in which everyone
suffers. Although the cooperative alternative often exudes moral attractiveness
it can be used to achieve some very unintelligent and unattractive results.
In the prisoner's dilemma it is illogical, untruthful, and ultimately immoral.
When veiwed in the full context of society, the prisoner's dilemma may be
no dilemma at all, but simply the demonstration of a small minded strategy
for avoiding social responsibility.
Although the listserv system permits people to meet and interact with others
from vastly different cultures and viewpoints, it also permits people to
seclude themselves into extremely insulated and isolated groups where the
cooperative alternative is enforced with a vengence.
In the social and political arena many of these narrowly
focused and very cooperative groups, groups which maintain the level of
cooperation by factoring any disturbing member out of the game, are made
up of well educated socially liberal people. They are tolerant and respectful
of opposing views, but simply choose to spend time with others who share
their outlook. They are relativists at their
worst, claiming that all points of view are more or less equally valid, and
thereby avoiding the stress of looking carefully at those who act or believe
differently. This insidious and divisive twist on the virtue of "tolerance"
becomes a form of elitism that inhabits the
dark side of American liberalism. (BTW I am a liberal) The cooperative choice
comes to resemble the price fixer more than the altruist.
Game theory with only 2 participants doesn't make sense to me
in trying to relate it to group
dynamics
Which is to say that after the fact, rationality may be projected
upon the facts of the choice-making using tools such as the prisoner's dilemma,
when in fact it was the choice-maker's anxiety,
denial and avoidance (say, of moving into the 'truer' :-) dilemma of a depressive
position,) that clouded over any sense of rational judgement predicated on
anticipated benefit
The dilemma is ours. How can this illogical state of affairs
exist? Although, as I suggested earlier, there may be no dilemma at all when
the totality of cirmcustances is seen. Compare it to the chicken and eqq
question. The answer is easy--Darwin--but you must step outside the question
to see it.
A program for playing an indefinite sequence of Prisoner's Dilemma
games was develped by Anatol Rapoport. Given on page 37 of Five Golden Rules:
Great Theories of 20th-Century Mathematics -- and Why they Matter, copyright
1996 by John L. Casti, "It consists of two rules: (1) cooperate on the first
encounter, and (2) in subsequent rounds, do whatever your opponent did on
the previous round." Up against other strategies Tit for Tat will win, even
against itself.
Return to Index