Defining Assumptions from Forecasts

It is sometimes convenient to use results from one simulation as inputs to another simulation. For example, the simulation results from a departmental revenue model can be used as input assumptions to a total company revenue model. It is not necessary for the two models to share the same simulation. You can use the Define Assumption from Forecast feature in Crystal Ball to convert a forecast distribution into an assumption in one of two ways. You can either fit a standard probability distribution to the forecast data, or use the forecast data directly as a custom distribution.

  To define an assumption from a forecast:

  1. Run a Crystal Ball simulation and open a chart for the target forecast.

  2. In the forecast chart menu bar, select Forecast, and then Define Assumption from Forecast.

  3. In the Define Assumption from Forecast dialog:

    • Enter a cell location for the new assumption. You can type it or click the cell selector to point to the cell.

    • Select a distribution type for the assumption. You can select the best fitting distribution or define a custom distribution.

      • If you select Best fitting distribution, the current fitting defaults are used. You can click Fit Options to display the Fit Options dialog described in Fitting a Distribution to a Forecast. If you select Show comparison chart, you can view a chart of each fitted distribution and optionally override the best fit selection (Confirming the Fitted Distribution).

      • If you select Custom distribution (with forecast data), when you click OK, the Define Assumption dialog opens for the custom distribution. It contains the data taken from the filtered range (if any) of the forecast. If you prefer, you can follow the instructions in  Using the Custom Distribution to modify the data.

      • Optional: If you select Custom distribution (with forecast data), you can select Sample sequentially (instead of randomly) to use sequential sampling. For details, see Sequential Sampling with Custom Distributions.

    • Optional: Indicate whether to display the new assumption's type and parameters in adjacent cells, either downward or to the right, and indicate whether to show names (labels) with the values.

  4. When settings are complete, click OK to perform distribution fitting (unless you selected Custom distribution) and open the Define Assumption dialog .

    You can save the assumption with the indicated defaults, or modify the assumption as if you were defining it in the usual way. Most assumption definition features are available. You can enter different parameters and correlate the assumption with another. However, you cannot add the assumption to the Gallery until after it has been created.

  5. To finish defining the assumption, click OK in the Define Assumption dialog.

After you have defined the new assumption, you can select it and select Define Assumption to change the distribution type or perform any other modifications, including adding it to the Gallery.