It appears to be intended for using illustrative experiments in an undergraduate-level course in game theory.
This site is based on the perception of game theory as the study of a set ofconsiderations used by individuals in strategic situations. Models are not seen as depictions of how individuals actually play game-like situations and are not meant to be used as the basis for a recommendation on how to play real “games”. My goal as a teacher is to deliver a loud and clear message that game theoretic models are not meant to supply predictions of strategic behavior in real life.
Cheap Talk 
3 comments
Comments feed for this article
January 20, 2011 at 10:11 pm
anonymouse
I was going to say hey that reminds me of this other rubinstein quote someone once pointed me to. a quick search reveals that someone was you too…
https://cheeptalk.wordpress.com/2010/10/06/rubinsteins-afterward/
He seems to be on quite a crusade.
January 25, 2011 at 2:04 am
Lones Smith
“My goal as a teacher is to deliver a loud and clear message that game theoretic models …” predict nothing of the real world.
Damn, maybe some insightful young economists are going to get this totally mind-blowing idea of ignoring the models that have worthless predictions, and INSTEAD just focusing on those that actually explain the real world well. When this radical paradigmatic shift happens, and I am not sure when it will, I predict that it will revolutionize our whole discipline. Now, I might be wrong, and that this “theories that work well” will just be one of those passing fads before we go back to the tried and true “totally bogus theories”. Gosh, now that I think of it, right now, I am going to be the first game theorist to make this shift. I hope it’s not going to damage my career.
January 25, 2011 at 2:31 am
Lones Smith
There are two sorts of theory papers each fitting a different essential role:
Class A: THEORY AS METAPHOR. A classic example is the folk theorem. No one is going to test this.
Class B. THEORY AS MODELED BEHAVIOR. Auction theory has more of this flavor.
The discipline has always proceeded by a motivated blend of both model classes. Ariel Rubinstein has been a prince of class A.
As Box said, “All models are wrong, but some are useful”. Yet the wrong models of flavor A are often ultimately more influential and important than the more useful models of class B, because they act as critical BENCHMARKS for future models of either class.
I am told that Selten often publicly poo-pooed his earlier key benchmark concept of trembling hand perfection — but this was to treat it as a model in class B, and not class A. To get the best models the profession wants in class B, we often need to proceed through a string of class A models. We absolutely needed THPE as a benchmark model.
We all need to be clear about what class our model is, and not proudly boast that it is useless for predictions when it is a class A model, and conversely, be afraid if it is useless if it is intended to be a class B model. When I referee, I often use the phrase that “I want to learn about the world around me, or about the world of models”.