The Confidence Illusion

You have probably met the person who sounds certain enough to end the discussion.

The voice is steady. The explanation is clean. There is no visible hesitation.

Someone else, who may actually know more, says, “I’m not completely sure, but here is how I see it.”

Who sounds more persuasive?

Often, the first person.

Humans use confidence as a cue because it can be informative. Someone who has done a task a thousand times should usually be more confident than a beginner. The trouble begins when we confuse the signal of confidence with the evidence behind it.

Confidence and correctness are related imperfectly

Confidence is a metacognitive judgement: an estimate of how likely we think we are to be right. Good judgement requires calibration — being confident when evidence is strong and cautious when evidence is weak.

But confidence is influenced by many things other than accuracy: familiarity, personality, incentives, group dynamics, recent success and the simplicity of the story we can tell ourselves.

This explains why a weak argument delivered fluently can sometimes defeat a stronger argument delivered carefully.

Simple stories feel more certain

Complex reality contains caveats. Strong expertise often reveals those caveats rather than removing them.

A doctor may say, “These symptoms are consistent with X, but there are two alternatives we should rule out.”

A financial adviser may say, “This allocation should reduce volatility, but it does not eliminate drawdown risk.”

A founder may say, “The customer interviews are encouraging, but we still have no evidence on repeat purchase.”

To an impatient listener, caveats can sound like weakness. In fact, they may be evidence that the person understands the problem.

Overconfidence has consequences

Behavioural finance research repeatedly finds overconfidence among the most studied biases in investment decisions. A 2025 systematic review of 63 empirical studies identified overconfidence, herding and loss aversion as dominant themes in research on investor behaviour, particularly in emerging markets.

The same basic pattern appears far beyond investing. We underestimate project timelines, overrate our ability to spot deception, believe our memory is more precise than it is and predict our emotional reactions with more certainty than the evidence justifies.

The cost of overconfidence is not only being wrong. It is failing to look for disconfirming evidence because the question already feels closed.

Confidence can be contagious

In meetings, confident language can reshape group judgement.

“Maybe” gets less airtime than “This will definitely work.”

A senior person's certainty can become especially powerful because status and confidence arrive together. Junior team members may begin to edit their own uncertainty around the leader's conviction.

The danger is greatest when the decision is difficult and feedback is slow. In those settings, confidence can persist for a long time before reality corrects it.

A better question than “How sure are you?”

Ask: What would change your mind?

A person with calibrated confidence can usually describe the evidence that would weaken their view.

You can also ask:

• What part of this conclusion is fact and what part is inference?
• What base rate are we ignoring?
• What evidence points in the opposite direction?
• If we are wrong, what is the cheapest way to discover it early?

The goal is not to become timid. Decisions require commitment. The goal is to separate conviction from certainty.

Questions worth sitting with

• Where in your life does confidence routinely influence you more than evidence?
• Which opinion do you hold strongly but rarely test against disconfirming information?
• Are the most confident people around you also the most calibrated?

Leave points

• Confidence is useful information, but it is not proof.
• Expertise often increases awareness of uncertainty rather than eliminating it.
• Strong narratives can create false certainty because they feel coherent.
• Overconfidence becomes dangerous when it reduces curiosity.
• “What would change your mind?” is often more revealing than “How sure are you?”