So everything is Nash isn't it? Nothing settles down until a Nash equilibrium is found.
In a 2x2 game, Nash is simple. Even in context of choices over continua (e.g. picking a point on a function), Nash is still just the point(s) where the optimal functions of two players cross.
And theoretically (need to think more carefully on this), the more players there are, the harder it is to find Nash.
Ad infinitum, it should be impossible to find Nash.
So why is it that we do find Nash in this world. It must be that people choice varies as the number of players increase. That is, payoffs change with the number of players.
Eventually this may lead us to a sort of majority rules world, where if enough people act in a certain way, it would induce even more people to behave in that certain way, and thus we achieve Nash equilibria (or the minority gets squeezed out).
Mark: this is herd effect!
---------
Another (unrelated) idea
What is the link between time series and cross-sectional. Surely there must be a link, i.e. the cross-sectional distribution of strategies today (mark: this links to Nash equilibirum idea) would influence the time-series variation of outcomes. Modelling phenomena with Panel Data isn't enough. The influence of distributions today on future outcomes must be explicitly taken into account.
One can almost think of this as an Einstein-type idea. People have always known space and time as separate dimensions, but no one had thought about how they might interact. Einstein did. And I'm sure his results are more general that most people understand (all results are always more general than people think).
Will the constant for us be the infinite case, where we have an infinite amount of players? In this case, it would be impossible influence anything / any body, since what one can do is of close to no importance. What would Nash equilibrium look in that world??
And would the restrictions imposed by microeconomic foundation axioms necessarily imply restrictions over how many (how little) Nash equilibria there might be??
Subscribe to:
Post Comments (Atom)
0 comments:
Post a Comment