OpAnalytics.ca logo

Seminar War Game Course

Defence Technology Agency, Devonport, April 2013

The Zefra Scenario

The Zefra scenario provides the background for a short role-playing game. It will be used as an example during discussions about planning and conducting analysis of a seminar war game.


Professional games can be conducted for a variety of purposes:

Within the operations research community, gaming is generally viewed associated with judgement-based methods or, colloquially, 'soft' operational analysis. NATO recently published a report on such methods: NATO Guide for Judgement-Based Operational Analysis in Defence Decision Making - Analyst-Oriented Volume: Code of Best Practice for 'Soft' Operational Analysis.

Many of the other 'soft' operational analysis methods covered in this code of best practice are appropriate for analysis that may be conducted for a professional game that is a component of an operations research project. These include such methods as influence diagrams, cognitive mapping, decision trees, drama theory, and the Delphi method.

Another comprehensive handbook for the operations research community is the ABCA Analysis Handbook. While it includes considerable stress on and guidelines for issues like developing an analysis plan and a data collection and management plan, this handbook also provides a useful checklist of mundane (but easily overlooked) items too, e.g., specifications for a computer network for analysts, manning for positions on the analysis team (and prerequisite skills), templates for the table of contents for various documents, schedules and 'battle rhythms' for analysts (including shift changes when working 24/7). This handbook was intended for large analysis activities (division level) with multinational participants; it may be over the top for activities that are substantially smaller in scale, but helpful nonetheless to reduce oversights when planning analysis for a professional game.

The commitment to develop analysis for some games will be dependent upon its associated purposes. If those purposes can be achieved by merely by leaving players with certain memories (say in training them), then relatively little analysis may be required. But if results are to be used outside the circle of player-participants, e.g., to provide evidence that a course of action or a future concept is good or bad, then substantial data collection and analysis should be included.

Typically each game will have a combination of objectives: for example, it may be to educate an external audience on certain issues while the participants are also gaining a deeper understanding of those same issues. The analysis team must be aware that delivering a good analysis product may not be the sole objective.

For some games, it may seem sufficient after the conclusion of a game that all the participants can depart with a shared memory of what transpired (perhaps to meet some training objective). Note: even when games are used purely in a training role, there should still be mechanisms to record activity and collect feedback. For instantance an After Action Review, is usually a critical part of any training program.

Thus, to achieve the objectives in some games, they may require little or no analysis. For example, games may be intended solely for team building; once the nascent team has developed mutual trust and understanding, nothing further may be required. Another objective that may be adopted for gaming is education; once an audience has absorbed the significance of the issues covered in the game, an analysis report may have no incremental value. However, for games that are intended to illuminate some significant decision or to assist with problem solving, a record of the proceedings, associated analysis, and a report will be its legacy and the game's value will be measured again the influence of the game's report.

As a precursor to the analysis that follows a game, scenarios must be carefully crafted for the game so that they permit data collection on relevant points. If some issue needs to be explored and yet there are no events in the scenario on that issue, there will be no suitable data to collect, and the analysis will fail. So designing the secenario must be done in conjunction with the plan for analysis (see the page on under scenarios).