Professors Without Borders

Although “gaming” may not have a positive connotation among peace researchers, setting it aside could do more harm, both with respect to the software games as well as the theoretical aspects coming from game theory. The latter I have tried to illustrate during the “Gaming for War or Peace: ICT voor Wereldvrede” Symposium, which was organized by the Peace Centre Eindhoven, the Netherlands, and partially funded by the Dutch national council for international cooperation and sustainable development (NCDO). I was an invited speaker, with a presentation entitled “Games against terrorism and for conflict resolution” (slides in Dutch), which I summarise below.

The first step for a game theoretic game in this setting is to identify the ‘players’: what kind of government, what type of aggrieved group (terrorist organization)? Second, which game theoretic games are out there that could be of use? In a conflict resolution setting, a zero-sum game where terrorists are executed and the government has won is, at best, counterproductive. Moreover, such simple games are some 50 years old and lie at the start of game theory, but the field of research has advanced at great pace in the meantime, where more complex scenarios can be modeled more comprehensively. For instance, by now, negotiators know that setting deadlines is not conducive to arriving at a good peace agreement—but Axelrod found out that principle already back in 1981, and not only that, but also that repeated simple games with rationally negative outcomes does lead to cooperation when repeated ‘infinitely’ (the perception of infinity suffices). One can also play with >2 players, build in more variables, calculate optimal coalitions, and much more.

In the presentation, I give one example with 3 players—fundamentalist and moderate aggrieved groups and the government—and three variables: changing payoffs, different power balance among the two potential coalition partners (moderate and fundamentalists, but one equally could switch the assignments to other combinations of players), and different ratio as to how the pie is to be divided among the players. Changes in the latter two leads to the observation that the coalition partners will have motivation to enter the coalition, or defect from it, at different points (i.e., stages in the negotiation phase) depending on the payoff. Conversely, the ‘outsider’ can derive from that point how the power balance between the two coalition partners is.

The last section of the slides is a very small selection of recent advances in computational game theory that has potential for being useful for political science; so, perhaps, in some 5-10 years it will trickle down through to practical applications to, for instance, first run a few simulations so that one is better informed when making the next move in the political arena. If it will result in less political miscalculations, I do not know, but it is worth a try.

More details of the first two sections—in English—can be found here and in the references at the end in the presentation. Further information about the other presentations of the symposium is available here (this link will be updated when the abstracts will be put online).