Future-Oriented Modelling & Simulation

Formal representations of systems used to explore how they may behave.

Three linked states of a system, with a feedback loop Three open boxes are stacked and joined by arrows running from each to the next. A fourth arrow leaves the bottom box, curves out to the left and returns into the top box, closing a feedback loop. A marker sits on that loop.

At a glance

Produces
An inspectable representation of a system's dynamics under stated assumptions.
Depends on
A model purpose fixed before model structure, and enough transparency to be argued with.
Fails when
Structure and parameters are treated as given rather than as choices that could be made differently.

Future-oriented modelling and simulation represents a system formally so that its behaviour can be explored under alternative assumptions, conditions and interventions. Where forecasting estimates outcomes, modelling aims to explain the dynamics that produce them - and to show what would have to be true for those dynamics to change.

How we approach it

A model is a structured argument. Its most valuable output is often not a number but a clearer statement of what the argument depends on.

  • Model purpose is fixed before model structure. The same system can be represented many ways, and only the decision at hand can tell us which representation is appropriate.
  • We prefer transparent structure to predictive ambition. A model that can be read, argued with and modified is more useful than one that is only trusted.
  • Structure is tested as well as parameters. Where data are thin, we explore the space of plausible structures and report how conclusions vary across them.
  • Models are documented to the standard of the claim they support, including their limits, and are published or shared where the work allows it.