Shall We Play a Game? The Promise (and Pitfalls) of Wargames for Policy

Wargames are having a moment. Newspapers and nightly news warn that wargames “map a huge toll” in a war over Taiwan and that in order to win, the Air Force will have to buy new bombers and fighters. Congressmembers, like Congressmen Mike Gallagher, have even argued that wargames should be a standard part of Congressional budget and legislative processes. 

So what are wargames?  And should we believe what they say?

While often called “simulations” or “exercises,” wargames are distinct from computer simulations of combat, field exercises featuring actual military forces, or organized brainstorming sessions. Instead, wargames are interactive events that display four characteristics: human players, immersed in scenarios, bounded by rules, and motivated by consequence-based outcomes.

Wargames go back millennia, with evidence of games in ancient Rome, early Iraq, and China. They took on a central role in the modern conduct of war with the Prussian development of Kriegspiel, a boardgame that simulated combat to train officers. A century later, the United States’ embrace of wargames for military planning between World Wars I and II became a pivotal part of the Navy’s success in the Pacific. During the Cold War, the U.S. military turned again to wargames to understand the impact of the nuclear revolution. U.S. defense wargaming continued after the Berlin Wall fell, with games designed to test new ideas about warfare and aid acquisition decisions.

But while wargames have primarily looked at military or crisis situations, their use goes far beyond “war.” Governments use games to simulate natural disasters and to assess economic cooperation; consultants use wargaming to test new business strategies; and scholars use gaming to study how human behavior affects various social and political phenomenon like governance, revolutions, and insurgencies.

Games are closely tied to policy decisions. Because they have such a strong experiential quality, they can be compelling evidence for policymakers, faced with difficult decisions in rare or unexpected scenarios. Senior decisionmakers often rely on lessons learned from games in which they previously participated, and the evocative results from games—especially ones played with high realism—can be far more persuasive than lab-produced models or simulations. Indeed, games can be such powerful devices of influence that organizations have been known to change rules to influence outcomes and leak results when it benefits their cause.

Wargames are increasingly prevalent and have a long history of influencing policy, but what makes a good game?  And how should we evaluate wargaming data when making policy?  

In general, for games that are analytic, or designed to answer questions, a good game should have five qualities.

First, they need to be enjoyable and believable so that players act in the game as they would in real life (this is what we call external validity in social science).

Secondly, they need the right players with the correct expertise and demographics to mirror real world decision-makers (in social science, we call this having the right sample to represent our real-life population).

Third, there need to be enough players and game iterations to be able to make conclusions that go beyond the unique characteristics of any one instance. The social science term for this is generalizability.

Fourth, good wargames control for bias within their scenarios and rules—things like baking in certain outcomes or creating scenarios that unduly influence players’ behaviors. When we are able to control for bias that might invalidate results, we call this internal validity. The best games have both external and internal validity.

Finally, for analytical games, the best designs have good data collection—able to explain not only what happened in the game, but why.  

The primary value of using and analyzing wargames is not in generating new or better data about outcomes, but is instead in understanding behaviors and choices leading to these outcomes. Wargames do not predict what will happen in conflict or crisis, but they can tell us why and how one outcome or another occurred.

This can be hard to do and so most games struggle on one of these dimensions (sometimes intentionally). That doesn’t mean that wargames aren’t important for policy, but that we need to do a better job of analyzing across games for patterns of outcomes and behaviors.

Here at the Hoover Institution, we are working to make wargaming an even better tool for policymakers—building the first ever open source archive of wargames and wargaming data, running games on some of the biggest policy questions of our time about cyber, nuclear weapons, AI, and space, and providing wargaming resources for the classroom to build the next generation of policy analysts.