Re: Game theory pathologies

["J. Barkley Rosser, Jr." <rosserjb@xxxxxxx>]

     Well, I don't think this is in either of these books
and it is not made much of in the Mirowski book either,
but can be found in the JEL review article by Mailath
in 1998 on "Do People Play Nash Equilibrium?" or
some title like that.
     The problem raised by Mailath (actually by many, he
is just reviewing and summarizing) is that in general
in many game situations there are multiple equilibria,
and LOTS of them.  So, "rational game theory," if that
is what we mean by Nash equilibrium, does not really tell
us what people will do, even if they are "rational."  This
is where recent developments like evolutionary game
theory come in that supposedly provide mechanisms
for how an equilibrium is to be selected.
     Another major problem, discussed in Mirowski and
evident from almost Day One, is the prisoner's dilemma,
which was first labeled by Nash's major professor at
Princeton, Al Tucker of the Kuhn-Tucker Theorem, although
it was first formulated by Melvin Dresher and Merrill Flood,
two mathematicians at the RAND Corporation (the role of
the RAND Corporation in the development of economics
in the second half of the 20th century is a major theme of
his _Machine Dreams_).  Anyway, the Nash equilibrium
for the one-shot prisoner's dilemma is to "confess" (not
cooperate) as is widely known.  However, right up front
Dresher and Flood conducted experiments on the prisoner's
dilemma and found that participants cooperated more often
than they "confessed."  Nash was quite upset about this result.
      Of course there has been a lot of discussion of the
prisoner's dilemma among political scientists as well as
economists, with Robert Axelrod's _The Evolution of
Cooperation_, 1984, providing a classic discussion.
Barkley Rosser
Editor, Journal of Economic Behavior and Organization
James Madison University
----- Original Message -----
From: "Tracy Lightcap" <tlightca@xxxxxxx>
To: <pkt@xxxxxxxxxxxxxxxx>
Sent: Tuesday, April 30, 2002 11:22 AM
Subject: Game theory pathologies


> Steven Block asked for a readable critical analysis of game theory based
on
> its applications. He should see:
>
> Donald Green and Ian Shapiro. 1994. Pathologies of rational choice theory.
> New Haven: Yale
>
> Jeffrey Friedman, ed. 1996. The rational choice contorversy. New Haven:
> Yale.
>
> Both have to do with with rational choice applications in political
> science, thus covering slightly broader ground than game theory, but the
> critique is sufficiently general to cover the bases. The present
> disciplinary unrest in political science - yes, we have it too - is
> directly connected to these books.
>
>
> Tracy Lightcap
> Associate Professor
> Political Science
> LaGrange College
> LaGrange  GA  30240-2999
> 760.880.8226  FAX 706.880.7040
> tlightca@xxxxxxx
>