Simulation models help scholars conduct thought experiments, forecast and predict policy implications or just think about a complex problem by formalising it. To run a simulation, one needs to define rules that describe what takes place in the simulation. Simulations are based on a large number of variables and simple algorithms, or rules, that describe how these values change over time. A simulation just steps the time - moves the timeframe forward and recomputes new values for each variable.
There are many different ways of doing simulations. System dynamic models focus on modelling the system at large: There is one unit only. It is based on stacks that accumulate values and flows between stacks. System dynamic models provide an overview of the phenomena, but they do not allow more detailed analysis. In agent-based simulations the focus is to define how single units behave and interact with other agents and the environment. Therefore, the focus is often linking individual behaviour to wider macro-level implications. Microsimulations are somewhere between system dynamic models and agent-based simulations. They simulate flows between groups but are based on a more individual approach when focusing on simulations.
Simulation models, particularly agent-based simulations, help to examine complex systems. Complex systems are any kind of system that includes feedback loops, self-organisation, emergence or non-linearity. Like a paradigm, complex systems provide tools and jargon to describe these kinds of systems, even as a divide and conquer kind of toolset to understand and articulate them. These may help social scientists building models from them.