Here’s what we know: an Iranian surface-to-air missile system successfully shot down a U.S. Navy operated RQ-4 Global Hawk June 20. Both Iranian and U.S. sources acknowledge the drone was shot down; to the extent that there is a dispute it is about the location of the drone when it was hit and about what happens next. A post-mortem (post-dronem?) analysis of the trajectories, debris, and released footage will be vital to determine the craft’s location. As for escalation, it is hard to predict exactly how any situation plays out.

While four Global Hawks have crashed previously, this event appears to be the first time one of the plane-sized and beluga-shaped aircraft has been destroyed by a hostile actor. The expense of the craft, about $123 million, as well as the deliberate nature of the attack, make it a somewhat unprecedented situation.

But for two rainy days in Washington in October 2015, researchers tried to parse out exactly how a scenario like this may play out.

The Center for New American Security, a D.C.-based think tank, held a series of short, iterative role-playing activities to see how nations might handle transgressions by drones instead of crewed aircraft. Dubbed “Game of Drones” and published in June 2016, the report provides insight into how people, put into the roles of military and national security policy makers, might parse threats. Perhaps more importantly, it is also a window into the kinds of risks politicians could be more willing to take without lives on the line.

Over the course of the exercise, four teams each participated in two games, where they learned the scenario, objectives, and what assets they’d have. Then they took action. The move for opposing pairs were then collected and interpreted by the people running the game, who then told the teams what had transpired. With that information in hand, the teams got to take further action in response to an opponent who they could only communicate with through use of force. The teams, made up of six to eight people, represented two major powers, a small power, and a non-state actor. (Full disclosure: the author was a participant for the nonstate actor team).

So what happened?

“Drones reduced tactical risk to state actors, but heightened strategic and political risks,” Alexandra Sander, then at CNAS, stated in the published wargame report. “Because uninhabited aircraft do not put pilots in harm’s way and are generally less expensive than human-inhabited counterparts, they gave actors the option to use force in many situations where they might have previously been reticent to take action. State teams often viewed drones as a relatively expendable resource to carry out operations that, if conducted using human-inhabited aircraft, might have been prohibitively risky or costly. As a result, drones changed the calculus surrounding the use of force, lowering the threshold for military action in some circumstances because the perceived risk was lower.”

The Northrop Grumman-built Global Hawk is unusual for drones in that it carries a per-unit price tag of around $123 million that puts it at more expensive than not just most drones, but even more expensive than human-inhabited aircraft like the F-35 series. Dollar value isn’t everything, of course. There’s also the role the surveillance craft has, the intelligence it provides, the time it takes to position other assets to fill that gap, and the overall effect of the messaging that comes from a vehicle shoot-down. What likely matters more than the cost of the vehicle is the presence or absence of a human on board.

This finding from the war game seems supported by remarks the president gave June 20, noting that the response would be different if the drone was instead a vehicle with people inside. The lack of a pilot in the vehicle likely influenced the decision to target the Global Hawk with a missile at all.

As reported from the CNAS wargame, “State actors did not hesitate to shoot down and destroy their adversaries’ drones if they were discovered within their territory or posed an obvious threat. This action sent a clear message to adversaries using drones and risked little in the way of backlash.”

Here the exact positioning of the Global Hawk matters for de-escalation, though President Trump seemed to provide a sort of de-escalatory off-ramp by suggesting that the takedown may have been a stupid decision by a general. Almost simultaneous with the president’s remarks came a statement from Central Command that the Global Hawk was outside Iranian airspace at the time. The Iranian foreign minister, meanwhile, tweeted maps claiming the drone was inside Iranian airspace.

The RQ-4 is, like many of the drones iconic to the long wars of the United States in the 21st century, designed to operate in uncontested airspace. To some extent, an operational ceiling of 60,000 feet above sea level provides security, but missiles have been used to hit planes at similar altitudes since the 1960s. With known flight patterns and lack of onboard defensive measures, as well as a dispute over the sovereignty rules regarding transit near and around the strait, it’s easy to see why the Global Hawk was selected as a potential target. To the extent that the point of the event is messaging, this was as clear a way to get that message across and do so using missiles.

“The use of drones complicates messaging and escalation dynamics,” concludes the report. “In expanding the space for action on the spectrum of conflict below full-scale war, drones create increased opportunities for misunderstanding and miscalculation and can even encourage riskier political or strategic choices.”

Messaging with missiles and drones is difficult and new. It does not help that this trial by missile fire comes on the tail of the Trump administration’s withdrawal from the Obama-negotiated Joint Comprehensive Plan of Action, the continued pressure of sanctions, and a flurry of activity in and around the Strait of Hormuz. As clear and abstracted as a message through force could be in the simulated environment of a wargame, the real-life applications are murkier, higher stakes, and don’t come with the tidy analysis written up afterward by omniscient game masters who synthesize the full experience into lessons learned.