Traditionally, there have only been two possible ways of analyzing the effects of change on a software model:

• Change the model, recalculate, and review the results
• Create a copy of the model, edit that copy, calculate, and review the results

Although either of these methods may be adequate for a relatively small system, the data duplication, editing, and re-editing becomes very time-consuming and error-prone as the size of the system and the number of possible conditions increase. Also, comparing conditions requires manual data manipulation, because all output must be stored in physically separate data files.

Before Haestad Methods: Manual Scenarios