Overconfidence Effect

Type: Cognitive Bias
Local HTML: overconfidence_effect.html


Definition

Systematic overestimation of our knowledge, skills, and predictive accuracy. We believe we’re better than we are.

Svenson’s study: 80% of drivers rate themselves above average. Mathematically impossible.


Three Types

TypeDescriptionExample
OverestimationThinking you’re better than you are”I’m an above-average driver”
OverplacementThinking you’re better than others”I know more than the average investor”
OverprecisionExcessive certainty in estimates”I’m 90% sure” when you’re 50%

Why It Matters

Trading: Overconfident investors trade too much, earn less. Leadership: CEOs overestimate ability to manage acquisitions. Safety: “I can handle it” leads to accidents. Planning: Underestimating risks, overestimating capabilities.


The Dunning-Kruger Connection

Overconfidence is strongest among the least competent. The more you know, the more you know you don’t know.

Experts are often underconfident; novices are often overconfident.


Fighting It

  1. Track predictions — Calibration training
  2. Consider the opposite — “What if I’m wrong?”
  3. Seek feedback — Honest assessment from others
  4. Use base rates — How do others perform?
  5. Assume average — Default assumption: you’re average

  • [[Dunning-Kruger Effect** — Incompetence + overconfidence
  • [[Illusion of Validity** — Confidence exceeds accuracy
  • [[Optimism Bias** — Positive overconfidence

Audio

Podcast episode: Overconfidence Effect


Part of the Cognitive Bias Reference