Forecasts: Estimating KPIs
Set a model time horizon, chart the outcomes that matter, and use repeated Forecasts to explore the range of possible house-building results.
Defining steps
Begin with the question from this project: how many houses might the model produce over two hours? Define one step as one minute, then set a 120-step horizon. The step is a modeling choice; its meaning must match the wood, stone and building rates you entered.
It is very important to define what a step represents in your model, which depends on the data and your previous analysis of the system.
For this project, one step equals one minute, which does not affect directly in the diagram but it defines how you model it and how you analyze data.
For instance, if you had decided to define that one step is an hour, you would have had to estimate how much wood and stone will be collected per hour, which will give you an estimate of how many houses will be built in an hour, having an effect on the entire diagram.
As a best practice, define the step before you start creating your diagram to avoid any changes afterwards.
Forecasts
One Play shows one possible path through the model. Forecasts repeat the model to show a range of outcomes when its rules include variation. Compare the distribution, then inspect whether the underlying assumptions are credible. A Forecast is a model result, not a guarantee about real players.
Forecasts repeat a model so you can compare outcomes and statistical summaries.
To use Forecasts:
- Add nodes to the chart by switching on Outcome in the Properties panel (in the right-click menu the same choice is called “Track in charts”)
- Choose the number of plays next to the Forecast button (for example 5×)
- Click the Forecast button
- Read the table next to the chart: for each node it lists, at the chosen step, the median, the min–max range and the mean
Always rename your Forecasts so that you can easily communicate what scenario you were testing.
There is no universal minimum number of Forecasts that makes a result reliable. Use enough repeated Plays to see whether the summary is stable for your question, and check the model and input assumptions before drawing conclusions.
For more information about Forecasts and charts, please refer to the documentation page:
Congratulations!
You have finished the Loot and Craft project. Open the completed public diagram to compare its structure with yours, then change a resource rate and see how the possible outcomes shift. The same steps apply to a model of another economy or system.
Open the completed Loot and Craft diagram in Machinations if the embedded view does not load.