I am currently a Nonresident Senior Fellow at the Atlantic Council, and this past week I was in Washington DC to run a crisis game, one designed in conjunction with my colleagues Bilal Saab and John Watts. The details and findings of the game will be outlined in an eventual Atlantic Council report, and may also be reported by journalists who participated, so I won’t detract from any of that by discussing the scenario or any of the findings here. Instead, I wanted to offer some thoughts on game methodology we used.
In particular, we had been asked by the Council’s Middle East Strategy Task Force to develop a game approach that would explore the impact of two different future US policy postures in the region. From the outset we were committed to assuring that the game design was not pre-cooked to validate a preferred option, but rather represented a fair examination of both approaches. Doing this properly really required two runs of the game, so that each posture would be presented with a series of similar challenges. However, such a desire had to be balanced against real world constraints: the participants would comprise around sixty former (and current) senior policy-makers and subject matter experts, we could really only expect to have them for a day, and we were limited by available space, budget, and other practical considerations. Also, we wanted to maximize the time players had to consider both the scenario and the implications of American policy.
What we decided on was rather different than the usual seminar wargame.
For a start we ran two simultaneous games using the same group of participants. One game (PURPLE) involved one set of US policies, the other game (GOLD) involved an alternative approach. The scenario and initial injects in both games were the same. However, once started the games were free to diverge. You can think of the process as involving two alternative game universes, with the variation built around a different set of US policies.
In terms of role assignments, the PURPLE and GOLD games each had their own US teams. However, the other teams were playing in both games at the same time.
We had concerns about doing it this way. Could players remember the details of the two games, especially when they started to diverge? We addressed this by appointing PURPLE and GOLD team captains in each team. The team captains were responsible for approving immediate tactical responses to the crisis (which could be submitted to the White Cell at any time) and overseeing the development of broader strategic responses (which were submitted in writing at the end of each game turn). The remainder of each team were free to assist both team captains with input and advice. In doing so, they were effectively operating in both alternative universes and hence were in a position to assess what impact differences in US policy had across the two games.
Each team had a White Cell liaison attached (drawn from a group of excellent Atlantic Council interns, plus one visiting Canadian). These acted as note-takers for the group, helped participants stay on track in terms of the game agenda, and communicated with the central White Cell and each other via the Slack messaging system. Slack was also used for formal statements by teams, and by us to insert inject events as needed (which were then being read to team members by their assigned White Cell liaison).
In most of my game designs I am eager to include elements of fog, friction, and various coordination challenges. I am also a very strong believer in building narrative engagement by players and encouraging them to internalize their roles and associated perspectives. In this particular case, each team was assigned to a different meeting room, but participants were encouraged to physically travel to amd meet with the other teams to consult and coordinate actions. There was some concern at asking senior participants to run around two floors of the Atlantic Council offices as if they were participating in model UN, but I’ve generally found people quite willing to do so. Moreover, we were able to allocate the rooms in such a way as to make communications between some rooms easier by placing them in close proximity, while making others more distant and hence increasing their sense of (diplomatic) isolation.
The crisis scenario and briefings were designed with asymmetric information to contribute to intra-group tensions and suspicions, but in such a way that escalation and desclataion were both realistic and possible outcomes. Following plenary welcome speeches and a game briefing, the first turn of the game ran until lunch. As the players ate the White Cell hurriedly collected together and synthesized the actions of the various teams, and a second game turn (with new crisis elements) was then introduced for the afternoon. Finally everyone reassembled in plenary session to share insights and analysis.
How did it go? The participants and observers are really the ones in the best position to judge that, but I was very pleased. The teams were extremly active, meeting with each other, making statements, taking immediate tactical actions, and developing larger strategic responses to the crises we threw at them. The key parties very much internalized “their” view of events, and sometimes became genuinely frustrated and antagonized by the actions of opponents. No one really seemed to be bored, or tuned out—indeed, at the end of the game we had to repeatedly tell some teams to stop playing and report for the coffee break and final plenary session.
I was also pleased with the richness of data we were able to extract from the process. The fact that most of the day had seen participants divided into multiple teams meant that we had many times more discussion than would have been possible in standard plenary sessions. The scenario seemed a fair test of both US policy postures. By playing simultaneously in two parallel games participants were readily able to identify both similarities and divergences. In some cases, US policy drove the two games in different directions. In other cases, differences in US policy were largely drowned out by powerful local and regional dynamics. That, I think, was a useful reminder that many of the levers of American power are far from all-powerful, and that it is frankly very hard to direct the behavior of a complex, adaptive system with so many actors and interests involved.
Finally, I was very pleased with the commitment of the Atlantic Council to run a methodologically-rigorous game. Not all national security gaming manages to avoid sponsor-injected bias, and using games as a mechanism to promote pre-existing policy preferences (something I’ve called “gamewashing”) is far from unusual in the think-tank world either.
When the final report is released—it has to be written first, of course—I’ll certainly link to it here at PAXsims.
Some of the White Cell at work (picture by John Watts).