I think the whole problem of game theory as Psychological Science is now finding is that,
*Man is a
rationalizing creature as opposed to a
rational creature*
You can read on cognitive dissonance theory here.
http://en.wikipedia.org/wiki/Cognitive_dissonanceIf humans were rational why would they smoke? It costs you money, makes you sick, possible makes you pay high health costs, and possibly kills you. What is the game theory explanation of why people smoke? There are other also countless examples such as gambling, where people countinue to play a game where it is mathematically rigged against them, people who continue eating, drinking after warnings from the doctor, people outbidding each other on ebay until they pay more than list...on and on.
A *rationalizing* being would rationalize away smoking.
"I will give up next year."
"My uncle lived until he was 90, and he smoked."
"Meh, the harmful effects are exaggerated."
There are an almost limitless ways in which the human mind
rationalizes things, as opposed to being rational about them.
In my case, I would NOT accept the $10 to ensure that you get Zero. Now you may think this as irational? However, you now know something about my character. How will you treat me in a future deal? What if there is a second round? What if there is a another game at some point in time in which
I get to choose? If somone offered my a mere $10, I would mentally class them as a "tool", somone not to do business with, and not to be trusted; you could possible cut yourself off from any future profits in our relationship by being a "w*nker. So in the bigger picture there's also an element of the "prisoners dilema" for you. On the other hand, if somone was stupid enopugh to accept $10 while I got $4990, I was class them as a "muppet", likely I would try to encourage them to be more of a muppet to see if I could derive more income from them.
So in addition to a lot of people not being rational, even a rational person may look at the biggure picture,
the game of real life is a lot more complicated than can be encapsulted in a simple "game". The whole problem with game theory and where it falls down, is that simple scenarios are taken as axiomatic, and in Tohmas Aquinas fashion, you can build a massive logical edifice on these so called "axioms"; what in computer science would be called "garbage in garbage out".