Book/Simulation2Book/Simulation3NewsgroupANewsgroupBSim Society Research Info.Paper AssignmentLecture OutlinesBook/Simulation2Book/Simulation3NewsgroupANewsgroupBSim Society Research Info.Paper AssignmentLecture OutlinesRobert Keohane, in AfterHegemony, argues that rational, egoistic actors (nation-states) who find themselves in a self-help and competitive world (the international system), find it in their interest in some circumstances to cooperate with other actors. He argues that cooperation abounds in the international system and sets out to theorize about how to account for its presence. SimSociety is an n-actor, spatial iterated Prisoners Dilemma game. Actors are characterized as rational and egoistic (for the most part). The Prisoners Dilemma game structure creates a situation where actors can find it in their interest to cooperate with other actors. That is, if actors can coordinate their behavior, they can improve on their own payoffs in both the short term and long term, than if they fail to cooperate and defect. Thus, SimSociety provides a good "laboratory" to investigate if and how actors can cooperate and if and how the artificial world of SimSociety becomes cooperative or remains highly conflictual.
SimSociety allows us to create all kinds of worlds (at least in the n-actor spatial Prisoners Dilemma world) and see how they evolve. How are cooperative worlds (worlds where almost all the interactions are cooperative and where the actors that populate the world have strategies that are generally cooperative in nature) created? What kinds of actor strategies are most conducive to generating cooperative worlds? How come? How do cooperative worlds fall apart and why? What kinds of cooperative worlds (types and mixes of actor strategies) appear to be stable and are unlikely to fall apart and become worlds populated with highly defective actors? Why? What kinds of cooperative worlds fall apart. Why? What does all this have to say for Keohane’s ideas about the emergence of cooperation in the international system?
I suggest that to answer these questions you pursue the following general approach. I would start with worlds populated with some fixed number of actors (try 60). After you have developed results that allow you to answer the above questions you may (if you have time and interest) want to see how varying the initial number of actors effects (if at all) your results and conclusions. You can vary the mix of actor strategies that you initially populate the world with. For instance you might start with 10 actors each of All-Cooperate, All-Defect, TFT, Pavlov, Grim, and Mean TFT. Or you may want to start with 30 All-Cooperates and 30 All-Defects. I think you will find that initial strategy mixtures generates some interesting results. You will be able to make strong inferences given summary data generated by 10,000 iterations on levels of cooperation, births and deaths in the system, number of actors in the system, and importantly the average level of CC, CD, DC, and DD probabilities of actor strategies in the system. Here is an opportunity to be thoughtful and creative. Tell a story with data to back it up.
You may choose to work alone or in groups. However, each of you must write your own paper. In you paper, you no doubt will refer to data and perhaps graphs to illustrate your point. The page limit does not include the presentation of raw data or of graphs. I recommend storing the results from your numerous simulation runs on a floppy disk. If not, be sure you have paper copy. Remember to make sure you have a paper copy of the analysis or have it copied to a floppy before you end a computer session (that is whenever you physically logout). It will not be stored on the hard drive.
Book/Simulation2Book/Simulation3NewsgroupANewsgroupBSim Society Research Info.Paper AssignmentLecture Outlines