Interactive scenarios

The method involves cross-impact analysis, a forecasting method based on the interaction between future events. In its “classic” form a rectangular matrix is constructed in which a set of prospective events serve as labels for both the rows and the columns. The probability of each event is considered independently and is also provided by the analyst. The cells of the matrix are filled with judgments about conditional probabilities, that is, the new probability of the event that labels the column, given the occurrence of the event that labels the row. The matrix is played out in Monte Carlo fashion. First an event is chosen at random. Its occurrence is decided on the basis of its probability. If it "happens" then the probabilities of the remaining events are replaced by the conditional probabilities shown in the matrix cells of the decided event's row. Then another event is chosen at random and decided according to its new probability, and so on until all the events have been “played.” The result is a single scenario containing events, which have "happened". This process is repeated many times and the number of occurrences of each event is recorded. This leads to the computation of a new probability for each event, now based on the interactions depicted in the matrix. The influence matrix is, in short, a map of causality. It is also an extra burden for scenario developers to construct, but leads to a deeper understanding of the internal self-consistency of a scenario. One of the operations performed by the software constructed by the Millennium Project is to convert the influence matrix to a matrix of conditional probabilities. The demonstration software allows the user the following options: 1. Display the scenario as it currently stands. This option simply displays the statements in amended form if the users have changed them and all statements that the users have added. 2. Display the scenario statements with cross-impacts. This option brings up the influence matrix in terms of conditional probabilities. 3. Display the scenario statements with probabilities. 4. Add or change a scenario statement. 5. Run a cross-impact solution. Strengths and Weaknesses. As for strengths, there are no other techniques that can systematically permit a user to make changes to a narrative scenario and to have the scenario “automatically” self-adjust to the changes. Furthermore, the construction of the influences forces the analyst to examine the relationships among the statements of a scenario - how one affects the others. This examination can lead to internal self-consistency, a mark of excellence in any scenario. However, the method, as it presently exists, is not developed well enough to be used with any but the simplest and shortest scenarios. It assumes the cross-impacts are linear across all possibilities.