But despite Trump’s reported denial, the plan, which is nicknamed the “bloody nose” option, has prompted a fierce debate about its merits. And while it would change the state of play, it is also consistent with a “game” where convincing your opponent of your intentions is paramount.
The thing that keeps us all from blowing up during a nuclear standoff is something called deterrence: No country carries out an attack against any other because each is convinced that it will face devastating retaliation. It’s this delicate stasis that has forestalled nuclear war ever since there was more than one country with nuclear weapons.
To make deterrence work, a country will send signals to its enemy demonstrating its willingness to fight back — nuclear tests, military exercises, statements from the White House lawn … and, yes, tweets.
The bloody nose, in theory, would be another, more robust, signal of willingness. Talk — and tweets — are cheap, after all. The logic in the tactic’s favor goes something like this: The U.S. would be unlikely to take this drastic step if it were bluffing about its willingness to launch a full attack on North Korea, perhaps even with a nuclear weapon. North Korea would realize that, and perhaps the U.S. could win some concessions. Or so the theory goes.
The argument against the bloody nose is just the same — it could prove very costly. Deterrence can quickly become escalation. North Korea has many options for retaliation, and it could inflict massive casualties on the Korean Peninsula or in Japan even with its conventional weapons. It is also now reported to have the ability to launch a long-distance strike on the U.S. mainland. If North Korea responded to its bloody nose, the U.S. would likely respond, and North Korea would likely respond to that, and so on.
“It’s pretty symmetrical,” Robert Powell, a Berkeley political scientist who studies game theory and war, told me. “The North Koreans would be asking fundamentally the same questions” about strikes and retaliation that the U.S. is asking, so we could expect them to do fundamentally the same things.