Argument

What to Do When Predicting Pandemics

Simulations have forecast disastrous consequences before. Here’s how to act on the lessons of wargames before they come to pass.

Peter Pellegrino, the U.S. Naval War College's senior military analyst for wargaming, briefs participants in a wargame reenactment of the Battle of Jutland on the college’s tiled floor in Newport, Rhode Island, on May 10, 2016.
Peter Pellegrino, the U.S. Naval War College's senior military analyst for wargaming, briefs participants in a wargame reenactment of the Battle of Jutland on the college’s tiled floor in Newport, Rhode Island, on May 10, 2016. U.S. Navy

In January 2017, incoming Trump administration officials were briefed on a novel strain of the flu overwhelming healthcare systems in Asia. Victims showed respiratory symptoms, governments restricted travel, and hospitals were running out of antivirals, ventilators, and personal protective equipment (PPE) for healthcare workers. The briefing was a part of a simulation, a tabletop exercise conducted during the presidential transition. It was followed in 2019 by a series of exercises called “Crimson Contagion.” Several think tanks and universities held similarly frightening and prophetic wargames, which Bill Gates had earlier dubbed “germ games.”

These simulations could have galvanized investments in pandemic preparedness and primed policymakers for crisis. Yet in March 2020, as the coronavirus spread around the world and throughout the United States, James Lawler, an infectious disease expert and advisor to both Presidents George W. Bush and Barack Obama, lamented: “We are making every misstep leaders initially made in table-tops …We have thrown 15 years of institutional learning out the window.” Meanwhile, the press reported a frustrating incapacity to learn from pandemic wargames and war game-like exercises, which had shown the importance of rapid, not reactionary, responses to a fast-moving disease: early social distancing; timely data capture, testing, and contact tracing; surge production of antivirals, medical equipment, and PPE; clear and authoritative public communication; and international coordination. At almost every turn, the federal government lagged. ?

This is not the first time that prescient simulations have been overlooked. Yet wargames are also more popular than ever. In both government and academia, they have been undergoing a renaissance, spurred in part by the Department of Defense. The decision-makers demanding this growing number of wargames seem to think they are effective—but in what way? Are we doomed like Cassandra to foresee our inevitable fate, yet remain incapable of heeding its lessons? Understanding the potential, and the limits, of wargames and simulations in Washington requires understanding their long and ambivalent history.


A map from Edward S. Miller's book "War Plan Orange."

A map from Edward S. Miller’s book “War Plan Orange.”

Wargames put human players in hypothetical scenarios and task them to make strategic choices. As players advance through multiple rounds or moves, they are forced to live with and learn from the consequences of their decisions. While simulations of combat and crises go back millennia, games first appeared in professional military training in 19th-century Prussia. Shortly thereafter, the U.S. Naval War College (NWC) introduced wargaming to the United States military. NWC officers pioneered innovative tactical and operational naval wargames throughout the 1920s and 1930s. Their efforts informed War Plan Orange, a prescient hypothetical confrontation with Japan that began 35 years before Pearl Harbor. More generally, NWC wargames were extraordinary sources of tactical and doctrinal innovation in the run-up to World War II. Officers anticipated the requirements of air power during amphibious landings and developed new tactics for aircraft carrier operations, mobile basing, and replenishing the fleet.

The games were nevertheless slow to influence strategic thought. At the time that NWC was conducting its wargames, the United States military was wrestling with a novel commitment to the defense of the Philippines—7,600 islands, 5,400 miles from the west coast of the United States and 5,400 miles from Hawaii. In the event of war with Japan, the existing strategy for the Philippines consisted of a daring rescue mission. The U.S. fleet was to rush to Manila, regroup, and seek out the Japanese fleet for one decisive engagement.

NWC wargames had impressed upon the Navy the risks of the strategy, especially after a 1922 treaty ruled out fortifying a naval base west of Hawaii. Simulations consistently revealed that a weary and spread-out American armada would be vulnerable to devastating Japanese air and torpedo attacks.

Climate and Security

Climate & Security: The threat of climate change requires understanding, strategic planning, and transnational cooperation across sectors in order to design and implement solutions on a global scale. FP Analytics’ special report looks at how to manage the cascading security implications of a warming world. Read more.

Yet die-hard defenders of the Philippines, including Governor General Leonard Wood, refused to let joint planners abandon the immediate trans-Pacific defense of Manila, and appealed directly to the Secretary of War. Plans for the dangerous rescue mission endured until mere days after Wood’s death in 1927. And it would still take another decade to fully develop the alternative “island hopping” campaign plans.

Aside from such motivated biases, War Plan Orange continued to suffer from interservice backscratching. While games concluded that the Philippines could not actually be immediately defended, the Navy was readily convinced to defend it because of the resources and status that accompanied an offensive mission in the western Pacific.

Change had to come from generational transition. Throughout the 1920s, historians have found, older naval officers were drawn to the boldness of the planned sprint across the Pacific. Younger officers, who had been engaged in wargames and brought their lessons to fleet exercises, were more cautious. By 1941, almost all of the Navy’s flag officers had played in NWC wargames. As a result, the war plans with which the United States entered the Pacific War had dropped the suicidal thrust in favor of “island hopping” to victory.


U.S. National Security Advisor Walt Rostow shows President Lyndon Johnson options for a military attack on a site in Vietnam during the Vietnam War from the Situation Room at the White House in 1968.

U.S. National Security Advisor Walt Rostow shows President Lyndon Johnson options for a military attack on a site in Vietnam during the Vietnam War from the Situation Room at the White House in 1968. Photo 12/Universal Images Group via Getty Images

Decades later, in the Vietnam era, bureaucratic interests again stymied the healthy translation of simulation lessons into strategic change. The SIGMA series of political-military wargames, conducted by the Joint Chiefs of Staff (JCS) from 1962 to 1967, infamously captured the reality of Vietnam and correctly forecast the frustrations that would unfold for policymakers and legislators over the subsequent decade. In his book diagnosing the failures of the Vietnam War, Dereliction of Duty, H.R. McMaster dubbed the SIGMA wargames “eerily prophetic.” In The Best and the Brightest, David Halberstam called them “a dry run for the actual thing.”

The games, the records of which are now available in archives, were based on a methodology developed by Harvard and Massachusetts Institute of Technology (MIT) researchers and were innovative for their time. Dozens of military and civilian officials assigned to represent color-coded governments (e.g. U.S. Blue, Soviet Red, South Vietnamese Green, North Vietnamese Brown, Viet Cong Black, as was designated in SIGMA I-66, played in September 1966), huddled in separate rooms and presented with the same hypothetical scenarios, debated and selected courses of action to achieve their political aims. A moderator “Control” team would then assess the moves, determine the probable outcome of their combination, and update the scenario to reveal the intended or unintended consequences of each team’s strategy. These immersive games often lasted several hours a day, for three or four days, and consisted of three to six rounds. They concluded with a debrief and player discussion.

SIGMA players confronted several quagmires ahead of their time—a dearth of public support for the Vietnam war effort, an asymmetry of commitment to prolonged conflict, the futility of coercive bombing, and a lack of leverage to negotiate peace. Records from the Kennedy and Johnson Presidential Archives confirm the resulting forecast. A 1962 game found that “if U.S. troops are introduced into such a situation, they will be there for some time.” In a 1964 version of the game, players rapidly ran out of “profitable targets” for bombing and lamented that “the [Vietnamese] people’s rage … turned against the American murderers. Not against any of their own leaders.”

And yet, the national security establishment ignored the lessons of the SIGMA wargames. A scouring of the U.S. Department of State’s Foreign Relations of the United States document series reveals that the game results did not travel far in interagency dialogues. In President Lyndon Johnson’s White House, the SIGMA-sponsoring JCS was out of favor in the policy advising process. And even in the few instances when results did spread, decisionmakers discounted them. In March 1964, academic and National Security Advisor McGeorge Bundy dismissed any lessons because the games were “dealing with higher levels of escalation.” Soon, the real world would be too.

At the end of May 1964, Undersecretary of State George Ball asked: “Why are we contemplating an air action against the North in the face of a recently played war game that demonstrated the ineffectiveness of such a tactic?” Putting a finer point on the unfolding tragedy in a footnote, he wrote, “Are we proposing action against the North because we are reasonably confident it will, in fact, work, or merely because we are becoming reasonably confident that the present course of action will not work and we are not able to think of anything else to do?” The policy that actually followed that question would needlessly take tens of thousands of American and millions of Vietnamese, Cambodian, and Laotian lives.

By contrast, a healthy integration of these wargames into the interagency policy process might have pushed the SIGMA results onto the agenda of an all-important Tuesday Luncheon with the president. Concerns about the applicability of their lessons might then have been fed back to game designers, who could have been challenged to replicate their results under conditions envisioned by decisionmakers. Perhaps the games could then have revealed some motivated biases of policymakers—and prevented unnecessary tragedy in Vietnam.


Naval Postgraduate School students participate in analytic wargames they designed to explore solutions for the Department of Defense’s national security concerns in 2018.

Naval Postgraduate School students participate in analytic wargames they designed to explore solutions for the Department of Defense’s national security concerns in 2018. Javier Chagoya/U.S. Navy

During the lead-up to the Vietnam War and the development of War Plan Orange, simulations exposed the weaknesses of strategic plans. But, as demonstrated, these wargames did not of their own accord change policy. When results came up against entrenched interests, they were ignored. Existing plans had to work.

Yet simulations can have an important impact over time and through the people exposed to them. Games can sensitize players to crisis decisionmaking and influence the choices they go on to make in the real world. In a survey of players who took part in MIT wargames held between 1958 and 1964, 56 percent of those “engaged in policy planning, formulation, or implementation” could recall an instance in which their wargame experience had been of practical value in their job. One player and Chief Scientist of the U.S. Navy’s Special Projects Office, John Craven, even reported that MIT’s 1964 DETEX II wargame was a “significant factor in the decision to spend several hundred million dollars on an improved communications system” for Polaris submarine-launched ballistic missiles. And Robert F. Kennedy so valued his participation in a 1963 wargame centered on a South America crisis that he discussed plans for President Kennedy to attend a simulation (JFK was assassinated before taking part). While contemporary data is scarcer, some players have recalled profound effects. Condoleezza Rice, for instance, reported that as Secretary of State on September 11, 2001, she thought to notify Moscow of U.S. military alerts and explain to friends and foes alike “that the United States has not been decapitated” based on her experience with misperception and escalation during Cold War crisis simulations.

Did Trump administration players learn from their experiences playing so-called germ games during the transition? The answer is likely in the affirmative. But almost two-thirds of those players left government between the transition and the onset of the COVID-19 crisis. They included senior leaders, Secretaries Rex Tillerson, John F. Kelly, and Rick Perry. Moreover, the 2019 Crimson Contagion simulations appear to have suffered the same fate as the SIGMA Vietnam games. Crimson Contagion’s “planning and thinking happened many layers down in the bureaucracy,” the New York Times reported. The powerful tools of simulation and gaming were sidelined. Too few games were played and too few participants remained in leadership roles. It is frustrating, to be sure, but also eye-opening about the impact that wargames can have through consistent iteration and player education. Going forward, we can learn from this history.

First, game results should be compared and distributed widely. Today, for instance, the Department of Defense is reinvesting in gaming to help adapt to an age of strategic uncertainty. Operational games already inform the Joint Operation Planning Process (JOPP), stress test military plans, and reveal unanticipated contingencies. Within the Pentagon, a new repository tracks and stores wargame reports. Nevertheless, contributing to such a repository can become just another checked box. To synthesize their results, an office must be tasked with meta-analysis of searchable reports aimed at distilling lessons from comparable games and actively briefing findings to planners. The best of these analyses will borrow research methods from the social sciences. Games cannot force leaders to listen; they need bureaucratic champions.

Second, iterative wargaming is part of a healthy policy process. Good simulations are not about coming up with a solution to put on a shelf. At best, they inform a continuing process. The more robust the process, the more likely plans are to be useful. No particular games should be expected to predict the future, and policymakers may pay closer attention to replicated findings. Repeated lessons can also be learned incrementally and assimilated over time, especially as players rise in the ranks of their organizations.

A person playing the role of a patient is checked during a two-day flu pandemic simulation at Tan Tock Seng hospital in Singapore on July 21, 2006, part of a large-scale exercise involving Singapore's international airport, a border crossing to Malaysia, and several hospitals.

A person playing the role of a patient is checked during a two-day flu pandemic simulation at Tan Tock Seng hospital in Singapore on July 21, 2006, part of a large-scale exercise involving Singapore’s international airport, a border crossing to Malaysia, and several hospitals. ROSLAN RAHMAN/AFP via Getty Images

Third, wargames empower people. Players develop crisis decision-making sensibilities. They can experiment and make mistakes in exercises. The lessons should linger in their minds. Players, moreover, need to be diverse. Game organizers should not just engage subject-matter experts who can be expected to behave in a realistic fashion. They should also invite those who might bring new perspectives, future analysts and leaders who in any case need to be sensitized to the vicissitudes of crisis response. They might include junior officials, legislative staff, as well as state and local representatives. State and local officials, who have ended up bearing more than their fair burden of the United States’ federalized COVID-19 response, are more commonly included in regional natural disaster preparedness exercises than in DC-driven national security crisis simulations.

Wargames empower people. Players develop crisis decision-making sensibilities.

Finally, games need to be played frequently as new leaders emerge. At the same time, as busy as senior policymakers are, they need to make time for simulations. Rather than seeing a “hypothetical” exercise as taking time away from pressing real-world obligations, they should consider the overlap—for instance, how simulating an acute crisis in the South China Sea can help one to better manage a chronic dispute in the area today.

Simulation designers can leverage technology to help immerse such players quickly or remotely in a scenario and potentially shorten the time commitment. To be sure, wargames with embedded video, augmented-, or virtual-reality experiences would still be simplifications of reality with lower stakes, incomplete information, and compressed decision timelines. Yet scholars have found that “synthetic experiences” and immersive fictions or videos can trigger cognitive processes akin to real-world decisionmaking. Deploying such immersive technology, an uncommon practice today, may also help researchers iterate games more quickly and capture better data on player behavior, instead of relying on frequently incomplete post-hoc reporting.

Inspired by the unfortunate lessons of the pandemic germ games, wargaming should also continue its renaissance. Some offices should clearly continue to focus on potential crises in the western Pacific, as well as in the Baltics and the Middle East. Enterprising wargamers should continue as well to study the effects of nuclear terrorism, cyber-attacks on financial systems and electricity grids, climate migrations, and natural disasters. And the method can be applied further afield to simulate peacetime alliance coordination, trade or arms control negotiations, election legitimacy crises, or governmental transitions. Even if previous experience with games has sometimes come up short, there is nothing preventing us from adapting a powerful instrument to empower future leaders to cope with new challenges.

Reid Pauly is the Assistant Professor of Nuclear Security and Policy at Brown University in the Department of Political Science and the Watson Institute for International and Public Affairs. Twitter: @reidpauly