MINDSETOngoing practice86% confidence

The Overconfidence Trap

Expertise builds confidence faster than it builds competence — and experts are the most dangerous victims

Problem it solves

Assuming that confidence signals competence, especially in experts

Best for

High achievers and domain experts making decisions outside their field of expertise

Not ideal for

Contexts where high confidence is warranted and well-calibrated (e.g. narrow technical tasks within one's proven domain)

Overview

Why this framework exists

Overconfidence is one of the most robust findings in behavioural economics and psychology: people systematically overestimate their knowledge, abilities, and the reliability of their judgements. This applies broadly across domains — not just investing but hiring, medical diagnosis, parenting, and business strategy. Crucially, Angner highlights the finding that is most counterintuitive: being highly educated or being an expert in a domain does not protect against overconfidence. It often makes it worse.

The mechanism is this: as you gain expertise in a field, your confidence grows rapidly. But your actual skill — measured by outcomes and calibration — grows more slowly. The gap between confidence and competence widens. An expert who has spent decades in a domain will make gut-level calls that feel calibrated from experience but may actually reflect outdated heuristics, pattern-matching to misleading precedents, or simple familiarity masquerading as knowledge.

The practical danger: people hire based on confidence, defer to the most certain-sounding voice in the room, and trust the boldest forecast. In hiring this produces teams that are systematically overconfident because the selection process filters for projected certainty. In investing it produces the 'I should be able to beat the market' fallacy. In medicine it produces the consultant who diagnoses across the room without examining the patient — and the patient who is impressed by the confidence rather than alarmed by the method.

Core principles

5 total
  1. Confidence and competence are not correlated the way people assume — especially in experts.
  2. Expertise in one domain does not reduce overconfidence, and may increase it through generalised comfort with authority.
  3. People who study decision-making and bias are often worse at decisions than average, because awareness does not eliminate the bias.
  4. Selecting for confidence in hiring produces systematically overconfident teams.
  5. The correct response to an unfamiliar domain is calibrated humility — not extrapolation from adjacent expertise.

Steps

4 steps
  1. Map your actual track record in this specific domain
    Before trusting your confidence in a decision, audit your actual outcomes in similar past decisions in this domain. Not your adjacent domain — this one. Investing confidence drawn from legal expertise is uncalibrated.
    Pro tipIf you have made fewer than 20 similar decisions in this specific context, your confidence is almost certainly not data-grounded.
  2. Separate confidence from method
    When evaluating others' advice (and your own), ask whether the confidence is backed by a sound method or merely by personality and certainty of delivery. The dermatologist was confident and wrong. The hiring candidate was confident and unqualified.
    Pro tipAsk the confident person to walk you through their reasoning, not just their conclusion. Method reveals calibration; conclusion reveals only confidence.
    WarningWe are evolutionarily wired to interpret confidence as competence — the bias is fast and automatic. It requires deliberate override.
  3. Apply a structured process for high-stakes decisions
    In domains where overconfidence is dangerous (investing, hiring, medical decisions, major financial choices), substitute a structured checklist or process for gut instinct. The process constrains the overconfident impulse.
    Pro tipChecklists are not for simple decisions — they are for experts who know too much and trust themselves too far. Atul Gawande's checklist research in surgery applies this principle directly.
  4. Audit your team for inadvertent confidence-selection
    If you make hiring decisions, examine whether your process selects for certainty of delivery. Ask whether the quiet, calibrated candidate was passed over for the bold one. Systematically correct for this by requiring evidence of track record.
    WarningManagers who deny being overconfident will simultaneously confirm that their team is overconfident — Angner's research shows this is the standard pattern.

Checklist

Saved in your browser

Examples

3 cases
The dermatologist across the room

Angner visited a dermatologist for a persistent skin condition. The specialist sat across the room, looking out of the window, and delivered a confident diagnosis while Angner was still giving his history. Angner walked out impressed.

OutcomeOnly later did Angner recognise the fallacy: the dermatologist had zero grounds for that confident assessment. No examination, no data. The confidence felt like competence — but was not. It is a mistake made 'all the time' in hiring, medicine, and investing.
The confident vs. calibrated job candidate

In a hiring scenario, candidate A is boldly confident: 'I know that, I'm great, you can trust me.' Candidate B — perhaps a woman — says 'I've never done this before but I'm a fast learner.' The confident candidate is typically preferred.

OutcomeIf the hire is made on this basis across a whole team, the team is systematically overconfident. The process has filtered for certainty of delivery, not accuracy of self-assessment.
Angner's own self-diagnosis

Angner describes himself as someone who studies decision-making partly because he knows he is bad at it. He drinks too much, exercises too little, eats out too much, saves too little — the standard pattern.

OutcomeAwareness of overconfidence and present bias does not automatically correct them. The behavioural interventions described elsewhere in the episode (temptation bundling, social accountability) are necessary precisely because insight alone is insufficient.

Common mistakes

4 traps
Assuming expertise in one domain transfers to adjacent decisions
Angner's father was an expert engineer and pilot who applied none of that rigour to investing — not because he was careless, but because he was unaware that investing was a different domain requiring different knowledge.
Interpreting confidence as a signal of competence
The dermatologist diagnosed from across the room and Angner was impressed rather than alarmed. This is the standard human response to projected certainty — but it is a systematic error. Confidence and accuracy are weakly correlated.
Believing awareness of a bias eliminates it
Angner studies overconfidence professionally and admits he falls prey to it regularly. Cognitive debiasing requires structural interventions, not just knowledge of the bias.
Building hiring processes that select for overconfidence
Managers who reward confident interview performance end up with uniformly overconfident teams. The manager then attributes this to their team rather than to their own selection criteria.

Origin story

How this framework came to be

Angner's research sits at the intersection of economics, philosophy, and psychology. He has studied overconfidence empirically and written about it extensively. The dermatologist anecdote — a specialist who diagnosed Angner's skin condition from across the room without examining him, and whom Angner initially found impressive — is his personal entry point into recognising how confidence creates the illusion of competence in observers. The hiring research, where selecting for confidence produces entirely overconfident teams, is drawn from the organisational behaviour literature.

Source

Traced to primary
Source · PODCAST
The Economist's Guide To Getting Rich
Erik Angner · 2025
Open source →

Related frameworks

Browse all Mindset →