My Knowledge Base

Overconfidence Bias

2 min read

Overconfidence Bias

Overconfidence is the systematic overestimation of the accuracy of one’s predictions, the reliability of one’s knowledge, and the quality of one’s plans. daniel-kahneman identifies it as “the most damaging of the biases” in kahneman-2011-thinking-fast-and-slow.

Three Forms

  1. Illusion of understanding — hindsight bias makes past events feel inevitable; this inflates confidence that the future can also be predicted. People who could not predict the 2008 financial crisis are fully confident they “understand” why it happened after the fact.

  2. Illusion of validity — subjective confidence in predictions exceeds their empirical accuracy. Stock analysts who are confident about a prediction are no more accurate than those who are less confident. Expertise in unpredictable domains (finance, politics) does not produce valid intuition; it produces confidently wrong intuition.

  3. Planning fallacy — projects are planned using best-case scenarios (inside view) rather than base rates of similar past projects (outside view). Result: projects are systematically late, over-budget, and underperforming. The corrective is reference-class forecasting — look at the actual distribution of outcomes for similar projects.

The Outside View

The key corrective concept: instead of asking “how will our project go?” ask “how have similar projects historically gone?” — then adjust for specific features. The outside view is systematically more accurate than the inside view for complex projects in uncertain domains.

Graph View

Backlinks