Metagame analysis involves framing a problem situation as a strategic game in which participants try to realise their objectives by means of the options available to them. The subsequent meta-analysis of this game gives insight in possible strategies and their outcome.
Origin
Metagame theory was developed by Nigel Howard in the 1960s as a reconstruction of mathematical game theory on a non-quantitative basis, hoping that it would thereby make more practical and intuitive sense . Metagame analysis reflects on a problem in terms of decision issues, and stakeholders who may exert different options to gain control over these issues. The analysis reveals what likely scenarios exist, and who has the power to control the course of events. The practical application of metagame theory is based on the analysis of options method, first applied to study problems like the strategic arms race and nuclear proliferation.
Method
Metagame analysis proceeds in three phases: analysis of options, scenario development, and scenario analysis.
Analysis of options
The first phase of analysis of options consists of the following four steps:
Structure the problem by identifying the issues to be decided. Identify the stakeholders who control the issues, either directly or indirectly. Make an inventory of policy options by means of which the stakeholders control the issues. Determine the dependencies between the policy options.
The dependencies between options should typically be formulated as "option X can only be implemented if option Y is also implemented", or "options Y and Z are mutually exclusive". The result is a metagame model, which can then be analysed in different ways.
Scenario development
The possible outcomes of the game, based on the combination of options, are called scenarios. In theory, a game with N stakeholders s1, ..., sN who have Oi options (i = 1, ..., N), there are O1×...×ON possible outcomes. As the number of stakeholders and the number of the options they have increase, the number of scenarios will increase steeply due to a combinatorial explosion. Conversely, the dependencies between options will reduce the number of scenarios, because they rule out those containing logically or physically impossible combinations of options.
If the set of feasible scenarios is too large to be analysed in full, some combinations may be eliminated because the analyst judges them to be not worth considering. When doing so, the analyst should take care to preserve these particular types of scenarios :
The Status Quo, representing the future as it was previously expected. The present scenario, which may differ from the Status Quo as it incorporates the intentions that are expressed by the stakeholders to change their plans; the Status Quo necessarily remains the same, but the present scenario may change as stakeholders interact and influence each other's plans. The positions of different stakeholders, being the scenarios they would like others to agree to. Similar to the present scenario, positions may change through interaction. Compromises between two stakeholders, defined as scenarios that, while not the position of either, are preferred by both to the other's position. A compromise does not necessarily have to involve all stakeholders. Conflict points, defined as scenarios that stakeholders might move to in trying to force others to accept their positions.