Marginal Revolution book forum, The Logic of Life, Chapter 2
The second chapter is discussed. Fabio Rojas has a great discussion. From this, I'm realizing that I don't have an objection to people being mostly rational, most of the time. I object to equilibrium. Game theory was developed in the framework of equilibrium, and outside of ideal situations like a poker game, I don't see where it applies.
It's like fluid dynamics. We can write down an exact equation for how fluids flow. (Lorentz or somebody, I should know this). But when we try to solve them, they are horrendously complicated, and exhibit chaotic behavior. So instead, for practical results, we using finite simulations, where we don't get the exact answer, but limit ourselves to some time and spatial resolution. That is often good enough, and gets better with every cycle of Moore's law.
But how do we tell when it's "good enough"? With fluid dynamics, we can see some of the results in films - digital water has gotten better, from The Abyss to Titanic to the whirlpool in Pirates of the Caribbean 2 (or 3?). What about economics?
More thoughts: lots of simulation code locked up inside the Wall Street trading firms - they hire physicists and coders all the time.
Companies evolve and learn: we can simulate that, what does it tell us?
Here's the question(s) I really want answered: Are the notions of a mostly rational actor and equilibrium linked? Is equilibrium really that useful? Might it be mostly misleading? What can the mostly rational actor explain by itself?
I suppose equilibrium is just a useful fiction.
Every time I try to draw a conclusion, I feel that I don't know anything.
It's like fluid dynamics. We can write down an exact equation for how fluids flow. (Lorentz or somebody, I should know this). But when we try to solve them, they are horrendously complicated, and exhibit chaotic behavior. So instead, for practical results, we using finite simulations, where we don't get the exact answer, but limit ourselves to some time and spatial resolution. That is often good enough, and gets better with every cycle of Moore's law.
But how do we tell when it's "good enough"? With fluid dynamics, we can see some of the results in films - digital water has gotten better, from The Abyss to Titanic to the whirlpool in Pirates of the Caribbean 2 (or 3?). What about economics?
More thoughts: lots of simulation code locked up inside the Wall Street trading firms - they hire physicists and coders all the time.
Companies evolve and learn: we can simulate that, what does it tell us?
Here's the question(s) I really want answered: Are the notions of a mostly rational actor and equilibrium linked? Is equilibrium really that useful? Might it be mostly misleading? What can the mostly rational actor explain by itself?
I suppose equilibrium is just a useful fiction.
Every time I try to draw a conclusion, I feel that I don't know anything.
Comments