Because they use people's imagination for filling in omitted details, simulations can be less expensive than constructing actual settings with all of their component parts. As a method, it is often more economical than setting up an experiment. It offers more control than observational research.
Two criticisms of simulations are 1) they are not sufficiently realistic to be convincing, and 2) they can be too realistic and stress the participants.
If a simulation is not sufficiently realistic, it will lack external validity; the results will not be generalizable to the actual situation. Playing a game is not the same as real life. The brief duration and fixed rules of games limit their relevance to real life. Even when role-playing generates a high level of involvement, it is limited in time. At some level, people know that the role will end.
On the other hand, even mild forms of role-playing can be upsetting. People may become so identified with their roles that they say and do things they regret later. The researcher can rarely predict how individuals will respond to a role-playing exercise, making it difficult to protect them from personal revelations or actions that my be harmful to them or upsetting to others. For example, a role-playing exercise with co-workers, friends, or fellow-students might result in behavior that is subsequently embarrassing, such as losing one's temper, weeping, or insulting another person.
When a mayor and a county air pollution officer play an urban planning game, each has the opportunity to observe the other's performance. If some time later, the mayor becomes a county supervisor, it might have serious implications for the county official's job. These potential consequences can limit the use of simulations. The ethical costs may outweigh the research benefits.
On to summary