Q-Step : Agent-based modeling

Though models sit at the centre of lines of social inquiry as diverse as game theory, statistical analysis, qualitative analysis, and political philosophy, all involve an attempt to describe core elements of the world in a way that helps us to understand, value, and predict that world. With Agent Based Models, computer simulations of the behaviours of many agents work deductively from simplified assumptions to create dynamic interactions that can be examined over a range of conditions to make inductive arguments about the nature of the world. In this generative reasoning approach, agents with very simple micromotives can lead to complex adaptive systems in which qualitatively different macrobehaviours emerge. How do very simple assumptions about drivers, city dwellers, and voters lead to complex emergent phenomena like traffic jams, housing segregation, and party realignment? In this lecture, I’ll introduce answers to these questions by building models of these problems and highlight tools you can use to develop your own agent based models.