Overconfidence: when trusting yourself is a defence

“Nobody needs to check this for me. I know what I’m doing.”

The sentence landed in a meeting, after a colleague suggested that someone else might look over the report before it went out. It was said firmly, almost kindly, but for a moment the whole room stood still. Nobody said anything. The report went out unchecked, it contained a mistake, and when the mistake surfaced, the blame went to “unclear instructions”.

Most of us know someone like that. Someone who doesn’t doubt, doesn’t ask and doesn’t apologise. Some of us admire them, because we would like even a fraction of that certainty. Others avoid them, because it is hard to breathe around them. Rarely do we ask what is going on beneath that certainty. That is what I want to write about here: when trusting yourself really is trust, when it is a miscalculation, and when it is a shield protecting something very fragile.

Let me say at the outset what this piece is not: it is not a checklist of signs you can use to pin a diagnosis on someone around you. Overconfidence is not a diagnosis. It is a description of behaviour, and very different stories can sit behind it.

What does “too much” actually mean?

Confidence in itself is neither good nor bad. It is an estimate of what you are capable of. The trouble starts when that estimate drifts away from reality — in either direction. I have written several times about low self-confidence, when someone can do more than they dare believe. Overconfidence is the mirror image: an estimate that goes beyond what a person actually knows or can do.

And here is the first uncomfortable piece of news: to some extent, we all do it. In 1981 the Swedish psychologist Ola Svenson asked a group of American students how well they drove compared with others. A full 93 per cent believed they drove better than average. Statistically that is impossible; psychologically it makes a lot of sense. A slightly flattering picture of ourselves helps us get up in the morning, try something new and survive failure without falling apart.

So in this piece, “overconfidence” does not mean “a bit more optimistic than strictly accurate”. It means confidence that has become rigid: that can no longer take in any information that might shake it, and that therefore starts to do damage — to decisions, to relationships and in the end to the person themselves.

Dunning–Kruger: what the study actually showed

Whenever overconfidence comes up today, one name almost always follows: the Dunning–Kruger effect. In 1999 the psychologists David Dunning and Justin Kruger published a paper with the telling title Unskilled and Unaware of It. They gave participants tests of logical reasoning, grammar and humour, then asked them how well they thought they had done compared with everyone else.

The most interesting result came from those who had done worst. On the logical reasoning test they scored, on average, at the 12th percentile — better than only about one in eight participants. Yet they estimated that they were at the 62nd percentile, comfortably above average. The authors explained this as a “dual burden”: the knowledge you need to do a task well is often the very same knowledge you need to recognise that you have done it badly.

Something else tends to get lost in the popular summaries: the best performers slightly underestimated themselves. Because the task was easy for them, they assumed it was easy for others too.

… and why we shouldn’t take it literally

The Dunning–Kruger effect has become an internet meme. A graph circulates online showing “Mount Stupid”, where beginners supposedly stand with soaring confidence before tumbling into the “valley of despair”. That graph is not in the original paper. It is a later illustration — a nice one, but with little connection to the research.

More importantly, researchers later pointed out statistical problems. Part of the pattern Dunning and Kruger described can arise simply from what statisticians call regression to the mean: anyone who scores very low on a test will almost inevitably rate themselves higher, and anyone who scores very high will rate themselves lower, purely because all estimates are somewhat imprecise. In 2020 the psychologists Gilles Gignac and Marcin Zajenkowski tested the effect with different statistical methods and found it to be considerably weaker than usually presented, and partly a statistical artefact.

So what are we left with? A cautious but still useful finding: where we lack knowledge, we also have fewer tools to notice that we lack it. That applies to all of us, not to some special group of “the ignorant”. And one more thing: Dunning–Kruger is about an error of estimation, not about defence. A person who overrates themselves because they don’t know something needs information. A person who overrates themselves so as not to feel something painful needs something quite different. The rest of this piece is about the second kind.

Confidence and arrogance: how to tell them apart

The psychologists Jessica Tracy and Richard Robins studied pride and found that it has two faces. The first they called authentic pride: satisfaction with what you have done — “I worked hard and it paid off”. The second they called hubristic pride: satisfaction with what you are — “it worked because I’m better”. From the outside the two look alike; both show up as an upright posture and a smile. But they are linked with very different traits. In their research, authentic pride went hand in hand with self-esteem, conscientiousness and good relationships. Hubristic pride went with shame, aggression and a fragile sense of self-worth.

It is a very useful distinction in everyday life too. Confidence is about actions and abilities: I know I can do this because I have done it before. Arrogance is about rank: I am more than you. A confident person does not need anyone else to be smaller. An arrogant person only feels big once somebody else is small.

I wrote about how real confidence is often not loud at all in Confidence isn’t volume: what trusting yourself really means. Here I am adding the other side of the same thought: loudness is not proof of confidence either. Sometimes it is quite the opposite.

Fragile high self-esteem

For a long time high self-esteem was simply considered a good thing. Schools and parenting manuals boosted it with praise, expecting children to become more successful and happier as a result. In 2003 a group of psychologists around Roy Baumeister reviewed hundreds of studies and found that the link was much weaker than hoped. High self-esteem was associated with more happiness and persistence, but it did not reliably lead to better grades, better relationships or less risky behaviour. Some people with very high self-esteem were even more aggressive when they felt threatened.

Researchers therefore began to distinguish between secure and fragile high self-esteem. Michael Kernis described fragile self-esteem as contingent (it holds only while you succeed), unstable (it swings sharply from day to day) and defensive (it needs constant validation). In the same year Christian Jordan and colleagues showed an intriguing pattern: some people report very high self-esteem on questionnaires, while indirect measures, which are harder to control consciously, reveal something much lower. It was precisely these people who, in experiments, were most defensive, most inclined to inflate their achievements and most dismissive of people who thought differently.

In other words: the loudest certainty sometimes stands on the thinnest ice. That is not a rule, and it does not apply to everyone who believes in themselves. But it is common enough to be worth a second look. I explain the difference between self-image, self-worth and self-confidence in more detail in Self-image, self-worth, self-confidence: how they differ and how each is built, and you will find a short version in the knowledge base under self-image.

Compensation: when certainty is a shield

As early as the beginning of the 20th century, the Austrian psychiatrist Alfred Adler wrote about what he called compensation. Every child at some point feels small, weak and dependent on others. That is normal, and it drives us to grow. But when that feeling of inferiority is too strong — because the child was often mocked, compared, overlooked — the opposite stance can develop: a sense of superiority that covers what the person cannot bear to feel.

Today we would call it a defence mechanism. A defence is not a lie someone consciously makes up. It is the way the psyche protects itself from a feeling that seems unbearable. And the feeling that excessive certainty most often protects against is shame — the sense that there is something wrong with me, that I am not enough. If the subject interests you, I wrote about it in Shame: the emotion that whispers “There is something wrong with me”.

How does compensation show up in everyday life? Some common forms:

• Being unable to say “I don’t know”. Every question gets an answer, even when there isn’t one.
• Instant counter-attack. A comment about your work isn’t considered but returned — “and what about you?”.
• Making others smaller. Other people’s success is put down to luck, connections or chance.
• Excessive risk-taking. Proving that nothing can stop you, even when it would be wiser to wait.
• Stories about yourself that are always slightly bigger. Not lying, but constantly touching up the picture so that it is enough.

Interestingly, this is often the mirror image of what I described in Impostor syndrome: why successful people doubt themselves. With impostor syndrome, a person fears others will discover they are not good enough, and so they work harder, hide and doubt. With defensive confidence the fear can be the same, but the solution is the reverse: loudness instead of hiding, and instead of doubt, a firmness that leaves no room for questions.

How it sounds in conversation

Defences like these don’t collapse when someone exposes them. They come down — slowly — only where a person no longer needs them. A conversation might go something like this (a composite, not a real person):

“I don’t know why I’m here. My partner says I never admit I’m wrong.”
“What do you think?”
“That I’m right most of the time. I don’t see what’s wrong with that.”
“What would happen if, just once, you said you weren’t?”
“… Nothing. Only then I’d be the one who got it wrong.”
“And who was the one who got it wrong when you were little?”
“Always me.”

Almost everything is in that short exchange. The certainty is not meant for his partner. It is meant for someone who is no longer in the room — perhaps a parent, a teacher or an older brother — in front of whom being “the one who got it wrong” was too painful. A grown man is still competing in a contest that ended thirty years ago.

A view from relational family therapy

In relational family therapy, in which I am training, we see patterns like this as something that arose in relationships and that relationships also keep alive. Christian Gostečnik describes how a child in a family unconsciously takes on a role that helps the family survive tension. One of the children becomes “the strong one” — the one who doesn’t cry, doesn’t ask for help and knows what to do. At a certain moment that role is necessary. Later it becomes a cage: the person no longer knows how to be different, because that would mean giving up exactly what earned them acceptance.

In adult relationships, overconfidence often shows up as an inability to be vulnerable. The partner complains that they never know what is going on inside the other; that they cannot talk with them about anything that might suggest weakness. And because certainty needs an audience, the relationship slowly turns into a stage: one plays the invincible one, the other plays the one who must keep confirming it — or the one who keeps pushing back. Both feel lonely.

That is why, in therapy, we don’t try to “bring anyone down to earth”. Humiliation only strengthens a defence. A space in which you can be wrong without losing your worth, on the other hand, slowly makes it unnecessary. Most often this happens in small moments: when a person says “I don’t know” for the first time and nothing terrible happens.

Exercise: check what your certainty stands on

If you recognised something of yourself while reading, that already takes courage. People with defensive confidence usually don’t read pieces like this to the end. Here is an exercise for one evening; it takes about ten minutes.

1. Recall the last time you were very sure you were right — and someone questioned it. Write it down in two or three sentences.
2. Write down what you felt in the first few seconds. Not what you said, but what happened in your body: heat in your face, tightness in your chest, the urge to answer straight away.
3. Ask yourself: what would it mean if that person were right? Not about the matter, but about you. If the answer is “nothing much, I’d just have been wrong”, your confidence is probably secure. If the answer is “that I’m incompetent” or “that they won’t take me seriously any more”, you have found the spot the shield is guarding.
4. Find where you know this feeling from. Who once let you know that being wrong means being worth less?
5. Next week, once, deliberately say “I don’t know” or “you may be right” — in a safe situation, with someone you trust. Then watch what happens. For most people, nothing happens. And that very “nothing” is the experience the shield itself would never allow.

Where this explanation doesn’t hold

Because the idea that “loudness hides insecurity” is so popular, it easily turns into a weapon. So here are some limits.

Not all certainty is a defence. Some people are sure of themselves because they have genuinely learned something well and tested it a hundred times. A surgeon who doesn’t hesitate isn’t necessarily arrogant. If we read every bit of decisiveness as hidden doubt, we often just reveal our own discomfort around people who dare more than we do.

Overestimating is sometimes just a lack of information. A beginner who thinks they’ll finish a project in a week doesn’t need therapy but experience — and someone who kindly shows them what they can’t yet see. Here Dunning–Kruger, with all its limitations, is a better explanation than Adler.

Culture and gender shape what seems “too much” to us. The same decisiveness that we call leadership in a man is often called rudeness in a woman. Before attributing overconfidence to someone, it is worth asking whether we would notice it in someone else.

Don’t diagnose. In recent years “narcissist” has become a word for anyone who has disappointed us. Narcissistic personality disorder is a clinical diagnosis that only a professional can make after a thorough assessment — and it is far rarer than social media would suggest. This piece is about behaviour and defences we all know, not about a disorder.

And finally: if you live with someone whose certainty has turned into control, humiliation or threats, it is no longer a question of confidence. At that point your safety comes first. In Slovenia you can get help on the SOS helpline for women and children experiencing violence, 080 11 55, and in an emergency call 113 or 112.

If the overconfidence is yours

If you recognised yourself, the first thing I want to say is this: the shield once protected something. You don’t have to tear it off, and you don’t have to be ashamed of it. But perhaps it is time to check whether you still need it everywhere you carry it.

It helps to start noticing the moments when something stings — not to judge them, but to get to know them. It helps to find a few people in front of whom you can allow yourself to be imperfect. And it helps to learn to take criticism as information rather than as a verdict about yourself; I write about that in How to take criticism without falling apart.

If you find there is more pain beneath the certainty than you can carry alone, that is a good reason for therapy. Not because there is something wrong with you, but because what was formed in relationships is most easily unlearned in a relationship.

What to remember

• We all overrate ourselves slightly. Confidence becomes excessive when it is rigid and can no longer take in any information that might shake it.
• The Dunning–Kruger effect is real, but weaker than the internet says, and partly a statistical artefact. It is about an error of estimation, not about defence.
• Confidence is about actions (I can do this); arrogance is about rank (I am more than you).
• High self-esteem can be fragile: contingent, unstable and defensive. When it is, it often protects against shame.
• You don’t break a defence with humiliation. It becomes unnecessary in a relationship where you can be wrong and still be accepted.