Poor, poor game theory
16 February 2015

Game theory does not receive much love outside academic departments. We are making slow but steady progress in the consolidation of quantitative social science in the mainstream media, but it seems that game theory is not following that same path. On the contrary, the game theory that gets mentioned in opinion pieces usually bears scarce resemblance to one that is the bread and butter of any Economics and some Political Science departments.

On the one hand, there are cases in which game theory is used as a mere descriptor for a given situation, as if game theory was some kind of taxonomy for 2$\times$2 interactions. We hear about how certain settings, typically in international relations, can be labelled as a “game of the chicken” or as a “battle of sexes” without ever talking about what game theory has to say about them, or about the predictions that can be made. As such, game theory is stripped of its analytical power, and it is transformed into a jargon, into a rethorical device to sophisticate —but not to advance— a discussion.

More upsetting are the many occassions in which pieces are baited with predictions coming from a theoretical model but all that is offered to the reader is a trivial game tree with fixed payoffs. No explanation of the structure of the interaction, no indication of the assumptions that generate those payoffs, and no dicussion of the conditions under which the equilibrium will change. All we are offered is a tree with some numbers on the terminal nodes. I tend to think that these abuses of the theory do a worse diservice as they dilute the value of the approach by constructing flimsy models while then claiming that they have the full backing of a mathematical theory.