A couple of weeks ago at Defense News, Michael Peck asked “Can We Simulate Non-American Wars?” raising the problem of mirror-imaging in policy and planning wargames:
The cardinal mistake in military history has always been mirror-imaging: the assumption, often colored by wishful thinking, that your enemy will conveniently devise the same goals, strategy and tactics as you do. But this doesn’t just apply to enemies. Even allies frequently misperceive each other.
This came to mind last month, when I read an article in Foreign Policy Magazine that described how Pentagon planners predict three possible Israeli options for destroying Iran’s nuclear program: air strikes, a commando raid, or wiping out Iran’s senior leadership. The article troubled me, but I couldn’t put my finger on why – until I realized that the Pentagon expects the Israelis to fight in the same way that the Pentagon itself might fight if it targets Iran’s or North Korea’s nuclear facilities.
It’s not totally illogical; the Israelis fight American-style (or Americans fight Israeli-style), so perhaps there’s a commonality of approach that enables accurate forecasting. But an Israel-Iran war would be a conflict between two foreign nations, with their own domestic agendas, cultural traditions and military methods. If Israel attacks Iran, it may employ technologies and tactics that the U.S. hasn’t thought of. Armies aren’t cloned like Dolly the Sheep.
Michael is absolutely right, of course—that’s why the dangers of mirror-imaging are pretty much in every Intelligence 101 course that trainee government analysts ever take. However, while being aware of the problem is part of the solution, it doesn’t in and of itself resolve it. Methodologically, there are variety of tools that one can use to get a sense of typical behaviours by “the other,” including examination of past military campaigns, published doctrine (fortunately, militaries love to publish doctrine), engagement of subject matter experts, and so forth. Equally important, however, may be the recruitment of personnel with the cognitive abilities and styles that allow them the empathize with the perceptions and motivations of others.
Yet focussing solely on the mirror-image part of the problem leads to another possible deficiency—namely the countervailing risk of viewing an actor’s responses as excessively determined by doctrine or cultural and organizational constraints. A study of pre-war Commonwealth military doctrine, for example, would have given you few grounds on which to have predicted the innovations of the 1942-44 Burma campaign under General William “Bill” Slim. Similarly, the performance of United Nations and mediation and peace operations has varied substantially depending on the personal attributes and styles of senior mission personnel. Trying to understand the behaviour of Hizbullah solely through the cultural lens of Shi’ism or Islamist movements would tend to obscure the almost Google-like innovation the organization has shown at times (but not always).
But do you want to allow game players full flexibility to do whatever they want? Much depends, of course, on the purpose of the wargame. How does one balance exploring the most probable response of an actor with the possible (but less likely) response? Can one do this in a single game (or, for that matter, is a wargame even the best way of exploring all of these contingent possibilities?)
Part of the answer, I think, is to view the behaviour of parties in a conflict (or, for that matter, peace process) as being distributed along a normal curve. Certainly, some types of behaviours or responses may be more common than others, but less doctrinaire or more innovative approaches (of greater or lesser effectiveness) should also have a non-zero probability of appearing. Game rules can variously permit less orthodox behaviours (but at an additional cost in time, energy, or resources), assign probabilities to approval of more innovative approaches (to reflect that less conventional response may be less likely to receive endorsement from the powers-that-be), or have doctrinal responses being the “default” setting which participants must actively decide to alter.
* * *
For more on the subject of simulating the behaviour of adversaries, have a look at the always-excellent Red Team Journal.