Kotaku.com.au recently published an article here about rational choice theory, and the huge problems it has faced in cashing out its claim to be a sound method for understanding history, politics and economics.
This is a comment I attempted to post in response to the article.
This is actually a very well researched and well-reasoned article. I did my BA Honours and MA in Political Philosophy on Game Theory / Rational Choice theory, and everything I found in the literature about (a) the ethics of rational choice theory; (b) the plausibility of the most popular/common axioms; and © the evidence from both evolutionary psychology and historical studies suggested everything that the author mentioned here.
You can find a quick summary of my research in two 4 000 word essays I wrote for my BA Honours and MA coursework: https://nd-au.academia.edu/BlairVidakovich
Basically, game theory / rational choice theory has some huge problems. There are two big ones: the first is that game theory is a conceptual framework for MODELLING reality, not actually engaging with it. It abstracts and simplifies the situations researchers deal with to try and help them work out what is significant. The problem with this is that modelling is nothing short of making educated guesses about what you will assume to be significant in the first place. So frequently models contain logical contradictions which lead to bogus results – like the case study about the cliff. With the right assumptions, game theory tells you to almost kill yourself in order to save yourself.
The second problem with game theory is that since game theory is a model of reality, it is largely deductive. That is, its conclusions flow completely from the original assumptions. No new information enters into the decision-making process once you have made your assumptions. This means game theoretical studies frequently ignore enormous amounts of historical evidence about how people and organisations actually think, or how they actually behave.
I think this article is significant because frequently game AI and story-driven events are designed based on models similar to game theory, and this has been widely observed to lead to poor character development and poor story structure.
I really enjoyed this article. More of this please.