On game theory and MI:  
  
How about a Cournot triopoly  - or an n-opoly for that matter? How about   
the free-rider problem.  
  
I do think that game theory raises problems for  a thorough-going   
MI,  because it demonstrates that so many situations involving rational   
agents interacting and aware of one another's rationality exhibit   
indeterminacy. If we need conventions or norms to solve the equilibrium   
selection problem incident to individualist explanation, then the program   
of explaining conventions and norms themselves on completely individualist   
grounds seems  hopeless.  
  
Kevin Quinn