Even Assad’s enemies are rethinking their strategy. European countries worry about the thousands of Europeans who have traveled to Syria to fight Assad—and their potential return as violent militants. Meanwhile, the United States has dispatched hundreds of advisors to join the battle against ISIS in Iraq. Members of the Obama administration are backing away from the goal of toppling Assad. “Anyone calling for regime change in Syria,” said one official, “is frankly blind to the past decade; and the collapse of eastern Syria, and growth of Jihadistan, leading to 30 to 50 suicide attacks a month in Iraq.”

The devil’s gambit is a chancy maneuver, since the resulting radicals could grow too powerful to control. For a dictator, the sweet spot is an extremist force that’s strong enough to inspire fear abroad, but not capable enough to topple the regime—which is roughly where ISIS is right now. If the militants become too potent, Assad will probably turn on them with a vengeance.

The devil’s gambit is risky, yes—but a calculated risk. A dictator looking down the barrel of regime change may be inclined to gamble. Perhaps Muammar Qaddafi and Hosni Mubarak lost power because they failed to play the game as ruthlessly or effectively as Assad has.