How can simulations help me take better actions?

This is the root of a tree of questions about using simulations to inform real-world decisions, each one answered through an interactive dashboard built for a specific decision problem. The dashboards are all powered by illustrative research models, not production-ready decision-support tools.

So, how can a simulation help with a decision at all?

The answer in short: a simulation lets us try an action before we take it. We change something in the simulated world, run it forward, and watch what follows, which is exactly the experiment that reality rarely lets us repeat.

To see why this is so useful, it helps to think about what an action looks like inside a simulation.

An action is a change to the simulation at some moment in time: a substitution made, a prescribing policy switched, a flood defence built, a battery charged or discharged, a support grant offered. Everything downstream of that change then plays out according to the rules of the simulated system.

References and further reading

Epstein, Joshua M. 2008. “Why Model?” Journal of Artificial Societies and Social Simulation 11 (4): 12. https://www.jasss.org/11/4/12.html.
Hardwick, Robert J. 2022. “Stochadex: A Simulation Engine Written in Go.” GitHub. https://stochadex.github.io/.
Sutton, Richard S., and Andrew G. Barto. 2018. Reinforcement Learning: An Introduction. 2nd ed. MIT Press.