There is a practical limit of about 250 fully correlated assumptions (assumptions that are correlated to every other assumption) and up to about 1,000 serially correlated assumptions (assumptions that are correlated to one or two other assumptions) for each workbook. If you have a large number of correlated assumptions:
If Crystal Ball detects inconsistently correlated assumptions when a simulation is running, it first determines whether small adjustments to the correlation coefficients are possible. This process might take a long time, depending on the number of correlated assumptions. Crystal Ball displays the message "Examining the Correlation Coefficients". If you get this message, you should probably stop and redefine the correlations.
If small adjustments to the correlation coefficients are possible, a dialog opens and prompts you to cancel the simulation or continue with adjusted coefficients.
Select one of the following responses:
Click Adjust Coefficients This Time Only to continue the simulation with the adjusted coefficients.
Since adjusted coefficients are not saved permanently, a dialog appears again if you stop and restart the simulation.
To continue the simulation, replacing the original correlation coefficients with the adjusted ones, click Adjust Coefficients Permanently.
Correlation coefficients that were specified using a cell reference in place of an actual value are replaced with a permanent value (the cell references are removed).
If small adjustments to the correlation coefficients are not possible, the simulation stops with an error message prompting you to reexamine the coefficients. To solve the problem, you can usually make large coefficients smaller or change the spreadsheet model to use formulas to calculate assumptions with large correlation coefficients. Creating a report containing just the assumptions might make it easier to spot problems.