Devil's Advocacy

Definition

Devil's advocacy is a structured method of challenging a strongly held consensus by constructing the best possible case for an opposite or alternative explanation. For a stated belief or assertion, the advocate proves its opposite by considering evidence that was disregarded or ignored, and by actively searching for new evidence that contradicts the original position. The method exposes confirmation bias—the tendency to form a conclusion early and then seek only supporting evidence.

In the Book

The handbook grounds devil's advocacy in cognitive challenges identified by intelligence analysis experts: we form conclusions early and then rationalize them with available evidence; we perceive what we expect to perceive; we dismiss information inconsistent with our views; and we become emotionally attached to pre-existing plans. Devil's advocacy counters this by asking: what would make the opposite true? For a stated position like "The U.S. Government should not directly fund private schools," the devil's advocate would enumerate reasons why direct funding should happen, reconsider evidence originally dismissed, and search for new disconfirming evidence. The method then forces reexamination of the original position: what reasons were faulty? What was overlooked? What implicit assumptions does it rest on? The handbook presents this as a tool leaders assign to individuals in decision-making groups, or that red teams employ to challenge plans before war-gaming and execution.

Why It Matters

This concept transforms disagreement from a social problem (somebody being difficult) into an analytical tool. It protects against the confident-but-wrong conclusion that hasn't been adequately tested. By institutionalizing the role of the person who argues the opposite, organizations make it the role's job—not the person's personality—to push back. Devil's advocacy is especially valuable in fields with high stakes for error: strategy, medicine, engineering, policy. It recognizes that the group's first answer is often wrong, and that finding the flaw requires active, disciplined skepticism, not hoping someone will volunteer to challenge the consensus.