Ken Binmore - Making Decisions in Large Worlds

[gusl]

Ken Binmore - Making Decisions in Large Worlds is a very interesting philosophical paper, which attacks the "school" of "Bayesianites", by which I think he means those who subscribe some type of extreme Bayesianism.

Comments

[candid.livejournal.com]

Ken Binmore is my personal hero.

[gustavolacerda.livejournal.com]

yes, I know.

[gustavolacerda.livejournal.com]

http://en.wikipedia.org/wiki/Kenneth_Binmore

Apparently, most of his work is on rational analysis / ideal-agent models (which I normally associate with Psychology / Neuroscience), which makes sense, since a game-theorist will tend to assume that game-theory works (or look for phenomena in which it works).

Here's the beginning of: A Little Behavioralism Can Go a Long Way

<< There is a school of behavioral economists who have popularized the notion that the neoclassical paradigm of homo economicus is refuted by the experimental evidence. We agree that the idea that human behavior can always be modeled as the rational optimization of money rewards in each and every context is off the wall, but who would want to defend such a wild claim? To make their case, behavioral economists need to address the more moderate claim that people often learn to play like income maximizers—given sufficient time and adequate incentives.
It isn’t enough to look only at the behavior of inexperienced subjects. Nobody denies that they are unlikely candidates for the role of economic man. Nor is it enough to keep pointing at unusual games like the Ultimatum Game, in which subjects do not seem to adjust their behavior much as they gain experience.
Indeed, it seems palpably dishonest to harp continually on such games, while simultaneously turning a blind eye to the very much larger literature in which laboratory subjects are reported as converging on the Nash equilibria of games with money payoffs. >>