What the data actually shows
Research on confidence and status, much of it associated with Cameron Anderson and colleagues, finds that overconfident individuals tend to attain higher social status because confidence is read as competence by others. In other words, looking sure is rewarded with influence and standing, somewhat independently of whether the confidence is justified — which is precisely what makes overconfidence pay.
Within groups and decisions, this shows up as what researchers have described as a confidence heuristic: we weight confident advisors and confident statements more heavily, treating expressed certainty as a proxy for being correct. Work in this area (associated with Price and Stone, among others) suggests people lean on confidence to decide whose advice to follow, even when that confidence is not well calibrated to accuracy.
But confidence and accuracy are only loosely linked. A recurring theme across studies of judgement is that how sure someone feels is an unreliable guide to how right they are — people can be highly confident and wrong, or accurate and tentative. When the cue we rely on (confidence) does not track the thing we care about (accuracy), the heuristic quietly misleads, and the most assertive voice can pull a group toward a worse decision.
Why this feels different from how it actually is
The confidence cue feels trustworthy because it is fast, vivid, and usually available before any evidence is. You can hear certainty in someone's voice instantly; verifying their track record takes effort and information you often do not have in the moment. The brain reaches for the cheap signal, and certainty is the cheapest one going.
It also feels reliable because it is right often enough. Confidence and competence are genuinely correlated in many everyday settings, so the shortcut earns our trust through frequent success — which makes the cases where it fails harder to spot, because we are not expecting them. The failures tend to be the high-stakes, hard-to-verify situations where it matters most.
Finally, confident people are simply more memorable and more emotionally reassuring. Certainty resolves the discomfort of not knowing, so we are drawn to it, and we tend to remember the person who sounded sure more than the person who was carefully hedged — even if the hedged person turned out to be right.
Confidence is worth noticing but worth checking — ideally against track record rather than tone.
What the research says to do about it
The most useful correction is to separate the signal you can see from the thing you care about: treat confidence as evidence of confidence, not of accuracy. When a decision matters, deliberately check the confident person's claim against their track record — how calibrated have they been before? — rather than against how sure they sound now.
Ask for reasoning and base rates, not just conclusions. Because the confidence heuristic latches onto tone, shifting the conversation toward evidence, uncertainty ranges, and past hit rates tends to surface the gap between how sure someone is and how right they are likely to be.
In groups, the research suggests it helps to make space for the less assertive voices and to weight contributions by accuracy where possible, rather than letting the most confident person set the direction by default. Structures that solicit independent judgements before discussion can blunt the pull of the loudest voice.
What the research says does not help
Simply telling yourself to 'ignore confidence' rarely works, because the heuristic operates fast and automatically. The cue will still register; what helps is adding a verification step, not trying to switch off a built-in shortcut by willpower alone.
Assuming confidence means competence is the core error — but so is the cynical overcorrection of assuming confident people are usually wrong. Confidence and accuracy are loosely linked, not inversely linked. Distrusting everyone who sounds sure throws away real signal along with the noise.
Relying on how confident you feel about your own judgements is just as unreliable. The same loose link between confidence and accuracy applies internally: feeling certain is not strong evidence that you are right, which is why checking against outcomes and track record matters for self-assessment too.
When the cue we rely on does not track the thing we care about, the heuristic quietly misleads — and the most assertive voice can pull a group toward a worse decision.
What this looks like in real life
The most certain voice sets the direction
In a meeting, the person who states a view with the most certainty tends to pull the group along, while the carefully hedged voice is discounted — even when the hedged one turns out to be right. That's the confidence heuristic: expressed certainty gets treated as a proxy for accuracy. Making space for less assertive voices, and soliciting independent judgements before discussion, blunts the pull of the loudest voice.
Confidence as evidence of confidence, not accuracy
An advisor sounds completely sure. Before following the advice on something that matters, the useful move isn't to dismiss them for sounding sure — it's to ask how calibrated they've been before, and to ask for reasoning, base rates, and uncertainty ranges rather than just the conclusion. Treat confidence as evidence of confidence, and let track record decide the rest.
Real numbers in context
This is a topic where the honest picture is a pattern, not a precise statistic. Research on overconfidence and status (associated with Cameron Anderson and colleagues) consistently finds that overconfident people tend to attain higher social standing because confidence is read as competence — a robust direction, even if exact effect sizes vary by study and setting.
The confidence heuristic — weighting confident advisors and statements more heavily (associated with Price and Stone and related work) — is similarly best understood as a reliable tendency rather than a fixed number. The key quantitative point is the weak one: the link between how confident someone is and how accurate they are is loose, which is exactly why the shortcut, useful as it is on average, misleads in the cases that matter most. Treat any single figure in this area with caution.