!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> Streamline Training & Documentation: Using Simulated Interactions to Forecast Decisions in Conflict Situations

Wednesday, August 29, 2007

Using Simulated Interactions to Forecast Decisions in Conflict Situations

J. Scott Armstrong, whose work on forecasting came up in yesterday's post, is a proponent of using simulated interaction to forecast what decision will emerge in a conflict involving a small number of parties, such as a hostile takeover of one business by another.

Most of Armstrong's work on simulated interaction has been done in collaboration with Kesten Green (pdf) of Monash University in Australia. They define simulated interaction (SI) as:
the acting out of interactions among people or groups who have roles that are likely to lead to conflict. ... An unaided expert must try to think through several rounds of interaction in order to make a forecast. In contrast, SI can realistically simulate interactions. SI can be used to forecast the effect of different strategies. For interactions involving role conflict, research has shown that SI provides forecasts that are substantially more accurate than those from game theory and from expert opinions (Green, 2002).1 ["game theory" link supplied]
Simulated interaction involves four main steps:
  1. Describe the roles of the key people in the conflict.


  2. Describe the target situation.

    Include information about the parties to the conflict, their goals, relevant history, current positions and expectations, and the nature of their interaction. Make sure that the details in the description cover what is needed to understand and address the issue to be forecast. If the relevant options for decision-makers can be identified, provide a list of the options.

    Expect to spend considerable time crafting the situation description.


  3. Simulate the situation.

    Conduct as many as ten independent simulations, in parallel if feasible. Allow an hour for each simulation.


  4. Derive the forecast.

    Record the final decision made by the role-players in the simulation(s) and, in the case of multiple simulations, combine the decisions to arrive at the forecast. For example, if seven out of ten simulations end in a decision to accede to a hostile takeover, predict that the hostile takeover will succeed.
You can find an admirably streamlined self-certification course in simulated interaction, prepared by Armstrong, here (beta version in MSWord). Armstrong and Kesten provide an extensive complilation of information on conflict forecasting at http://www.conflictforecasting.com/.

__________
1 Kesten C. Green, "Forecasting Decisions in Conflict Situations: A Comparison of Game Theory, Role-Playing, and Unaided Judgement,: International Journal of Forecasting, Vol. 18, pp. 321-344. "Expert opinions" are defined as "predictions of how others will behave in a particular situation, made by persons with knowledge the situation."

###

Labels: , ,