Actively seek disconfirming evidence

Deliberately look for evidence that your current belief is wrong.

Why it works

Confirmation bias causes people to weight confirming evidence more than disconfirming evidence, producing overconfidence. Superforecasters counteract this by making disconfirmation search an explicit habit: they ask "what would prove me wrong?" before seeking supporting evidence. This does not eliminate the bias but changes the evidence set the estimate is drawn from.

How to do it

1. After forming a belief or forecast, write: "What evidence would most strongly suggest I’m wrong?"
2. Actively search for that evidence before finalizing the estimate.
3. If you find compelling disconfirmation, update the probability downward; if you genuinely can’t find it, increase it modestly.
4. Bring disconfirming cases to your team or accountability partner before committing.

Evidence

Confirmation bias is one of the most replicated findings in cognitive psychology. Actively seeking disconfirming evidence as an intervention has been studied; the results are modest — it reduces overconfidence but does not eliminate it. (observational)

Actively seeking disconfirmation is effortful and produces discomfort — compliance under time pressure or high-commitment conditions drops significantly.

Sources

• Wason (1960), "On the failure to eliminate hypotheses in a conceptual task," Quarterly Journal of Experimental Psychology
• Tetlock & Gardner (2015), Superforecasting, on superforecaster devil’s-advocate habits

Common mistake

Seeking superficial disconfirmation ("I considered the other side") rather than genuinely hunting for the argument that most threatens the current belief.