Dunning-Kruger Effect Explained: Why the Least Skilled Feel Most Confident

You know the guy. Every family WhatsApp group has one. He watched a fifteen-minute YouTube video about the stock market last Tuesday and now he's explaining to your finance-professional cousin why Nifty is definitely going to crash next month. He's not necessarily lying. He's not even being deliberately arrogant. He genuinely, fully believes he understands this topic better than someone who's spent years studying it professionally.

Meanwhile, that finance-professional cousin sits quietly in the same conversation, hedging every sentence with "well, it depends" and "I could be wrong, but." The person who actually knows the subject sounds less certain than the person who barely does.

We've all seen some version of this. It happens in family conversations, workplaces, social media debates, investing discussions, health conversations, and increasingly, in the endless stream of information that arrives on our phones every day. But this isn't simply a story about arrogance versus humility, and it isn't accurate to reduce it to "stupid people think they're smart." There's a documented psychological finding behind part of this mismatch, and it has a name that's become so overused online that many people who cite it have never actually looked at what the original research found.

The Dunning-Kruger effect is really about something more subtle: the difficulty people can have accurately judging their own level of knowledge or performance in a particular area. And the uncomfortable part is that it isn't a phenomenon that belongs exclusively to other people.

It can happen to all of us.

The Study That Started All This

In 1999, psychologists David Dunning and Justin Kruger published a paper with a title that's almost funny to say out loud: "Unskilled and Unaware of It: How Difficulties in Recognizing One's Own Incompetence Lead to Inflated Self-Assessments." They conducted a series of studies in which students performed tasks involving areas such as logical reasoning, grammar and humour, and then asked the participants to estimate how well they had performed compared with their peers.

The interesting part wasn't simply that some students performed badly. People are obviously going to perform better or worse on any test. What interested the researchers was the gap between actual performance and perceived performance. Participants who performed in the bottom quartile substantially overestimated how well they had done. In the original study, they estimated that they had performed around the 62nd percentile, while their actual performance was around the 12th percentile.

That's a remarkable difference. But the important thing isn't the exact number itself. It's the underlying problem: someone can lack the knowledge needed to perform a task well and, at the same time, lack enough understanding of that task to accurately recognise how poorly they performed.

That's what made the finding so interesting.

The person isn't necessarily pretending.

They may genuinely not have the internal measuring tool required to judge themselves accurately.

The Part Almost Everyone Forgets to Mention

Here's what most online explanations leave out, and it's arguably the more interesting half of the story.

The original research didn't simply show that lower-performing participants overestimated themselves. Participants who performed particularly well also tended to underestimate their relative performance. Those in the top quartile estimated that they had performed around the 74th percentile, even though their actual performance placed them around the 87th percentile.

That doesn't mean every expert is secretly insecure or that talented people always lack confidence. It simply illustrates that self-assessment isn't perfectly calibrated, even among people who perform well.

Dunning and Kruger's explanation for the lower-performing participants was particularly interesting. The skills required to perform well can also be important for recognising mistakes and evaluating performance. If you don't yet understand the rules of a task, you may also struggle to recognise when you've broken those rules.

Think about someone who has watched dozens of videos about photography but has never actually learned how exposure, composition, lighting and perspective work together. They may recognise what a "good photograph" looks like in a general sense. They may even have strong opinions about photographs. But accurately evaluating their own technical ability requires knowledge that they haven't yet developed.

That is a very different claim from saying they're arrogant.

It's a claim about calibration.

Why This Term Gets Misused Constantly

There's an important correction worth making here because the internet version of the Dunning-Kruger effect has drifted quite far from the original research. The popular graph showing confidence shooting dramatically upward with almost no knowledge, crashing at a supposed "peak of stupidity," and then slowly rising as expertise develops is shared everywhere. But that famous graph does not accurately represent the actual data from the original paper.

There has also been considerable statistical criticism of the effect. Some researchers have argued that parts of the apparent pattern can be explained by statistical phenomena such as regression to the mean rather than representing a simple psychological curve of "less knowledge equals more confidence." That doesn't mean the original paper can simply be dismissed. It means the popular interpretation needs more caution than the meme usually provides.

This matters because the internet-friendly version—"the dumbest people think they're the smartest"—is much simpler than what the research actually investigated.

The more defensible lesson is that people can be poor judges of their own performance when they lack the knowledge or skills needed to evaluate that performance accurately. And importantly, this is a domain-specific problem. Someone can be highly knowledgeable in one area and badly miscalibrated in another.

The person who understands software engineering brilliantly doesn't automatically understand economics.

The successful business owner isn't automatically an expert in medicine.

And someone who has spent years studying psychology can still be completely uninformed about a complicated engineering problem.

Knowing a lot about one subject doesn't give you a free pass on every other subject.

My Honest Take: We Only Ever Use This Term to Insult Other People

Here's something I think is genuinely worth admitting. I've rarely heard someone say, "I might be experiencing Dunning-Kruger right now." It's almost always aimed outward, at the uncle in the WhatsApp group, the LinkedIn influencer with strong opinions about a field they've never worked in, the colleague who's supremely confident about something they haven't researched, or the stranger arguing with everyone in a comment section.

It has become a socially acceptable way of saying "you don't know what you're talking about" while sounding academically informed while saying it.

I think that's a shame, because the actual use of this idea is far more uncomfortable and far more useful when turned inward.

Everybody has some domain where they're confidently wrong without any internal alarm going off to warn them. It might be investing. It might be nutrition. It might be technology. It might be parenting. It might be relationships. It might even be something connected to our own profession that we haven't learned as deeply as we imagine.

The genuinely humbling exercise isn't spotting the Dunning-Kruger effect in somebody else's opinion.

It's asking yourself where you might be speaking with more certainty than your actual depth of knowledge justifies.

That's a much less satisfying use of the concept.

Which is probably why almost nobody does it.

Why This Spreads So Easily on Indian WhatsApp

There's a specific reason this pattern feels particularly familiar to anyone who has spent time in Indian family WhatsApp groups.

A forwarded message arrives completely formed.

It may contain a health claim, an investment prediction, a political statement, a technology rumour, or some dramatic piece of "breaking news." The message often doesn't show the process that produced the conclusion. You don't see the original source being examined. You don't see somebody challenging the claim. You don't see the uncertainty that might have existed before the message was turned into a confident statement.

And because the message may have been forwarded by someone you already trust, that confidence can acquire an additional layer of credibility.

The problem isn't necessarily that the person forwarding it is deliberately spreading misinformation. They may genuinely believe it is true. The problem is that confidence has become detached from the amount of evidence supporting the claim.

Health advice, stock tips and dramatic news claims can all travel through the same mechanism. A short message says something with complete certainty, another person forwards it, and within hours dozens of people have encountered the same conclusion without ever seeing the original evidence.

That's why a confident message deserves scrutiny rather than automatic trust.

Not because every confident message is false.

Because confidence itself isn't evidence.

The Familiarity Trap

There's another reason this happens that is easy to miss.

Sometimes we mistake recognising information for understanding information.

You've heard the word before.

You've watched a video about it.

You've read three social-media posts.

Someone explained it to you once.

It now feels familiar.

And familiarity can create a strange feeling of competence.

You know the terminology, so you feel like you understand the subject.

But imagine hearing the word "inflation" fifty times. You may become very comfortable discussing inflation without understanding how monetary policy, interest rates, supply shocks, wages, exchange rates and consumer demand interact.

The same thing happens with artificial intelligence, investing, psychology, nutrition and countless other subjects.

Knowing the vocabulary isn't the same as knowing the subject.

And sometimes the fastest way to expose the difference is to try explaining the idea without looking at your phone.

That's when you discover how much you actually understand.

Why Experts Sometimes Sound Less Certain

One of the strange things about expertise is that learning more can reveal how complicated a subject actually is.

A beginner may learn a simple rule and feel confident.

An experienced person may know the rule, the exceptions to the rule, the situations where the rule doesn't apply, and the evidence that remains uncertain.

So the expert says:

"It depends."

The beginner says:

"Obviously."

The beginner may sound more convincing.

But sounding certain isn't the same thing as being correct.

Of course, experts can also be overconfident, and people with little knowledge can occasionally be correct. Confidence and competence don't move in perfectly opposite directions.

The point is simply that you shouldn't use confidence as a substitute for evidence.

A person who speaks with certainty may be right.

A person who speaks cautiously may be wrong.

You need something beyond their tone of voice to know which is which.

How to Actually Use This Idea Well

The Dunning-Kruger effect becomes useful when it changes the way you evaluate your own knowledge rather than giving you a new label for somebody you dislike.

Notice when your certainty feels effortless. If you learned about a subject yesterday and already feel completely certain that you understand it, that doesn't prove you're wrong. But it may be worth asking whether you've encountered enough complexity to justify that level of confidence. Sometimes certainty comes from knowledge. Sometimes it comes from not yet seeing the reasons to doubt yourself.

Ask what would change your mind. This is one of the simplest ways to test whether you've left yourself room to learn. If you genuinely cannot imagine any evidence that could change your position, consider whether you've investigated the subject deeply enough. Being willing to update your opinion doesn't mean you have no convictions. It means your convictions remain connected to evidence.

Separate familiarity from expertise. Watching ten videos about investing doesn't make you an investment professional. Reading several articles about psychology doesn't make you a psychologist. Learning terminology is useful, but practical competence usually requires deeper study, experience and feedback.

Treat confident forwards with more scrutiny, not less. If a message says something is "100% proven," claims that "doctors don't want you to know," or presents an enormous conclusion without identifying credible evidence, the confidence of the wording shouldn't make you trust it more. It should make you ask where the claim came from.

Remember that uncertainty can be a sign of knowledge. A specialist saying "there are several possibilities" isn't necessarily avoiding the question. Sometimes they are acknowledging genuine uncertainty that a less-informed person doesn't know exists. But don't turn this into another simplistic rule either. Experts can be confident when evidence is strong, and non-experts can sometimes express themselves carefully. Again, the evidence matters more than the tone.

The Opposite Problem: When You Don't Realise You're Better Than You Think

The original findings also raise an interesting question about people who perform well.

If someone finds a task easy because they have spent years practising it, they may not realise how difficult that task is for someone encountering it for the first time.

A senior programmer might solve a technical problem in twenty minutes and think, "That wasn't difficult."

A junior developer watching the process might be thinking, "How did you even know where to start?"

The senior programmer's speed doesn't necessarily mean the task was objectively easy.

It may mean they've built up years of pattern recognition.

This happens in almost every profession.

A surgeon may make a complex procedure look routine.

A photographer may adjust lighting in seconds.

An experienced manager may resolve a workplace conflict with a conversation that looks effortless from the outside.

Experts often compress complicated processes into intuitive decisions because they've performed them so many times.

That can make their own competence difficult for them to notice.

The Confidence Trap at Work

This matters particularly in workplaces because confidence is often rewarded before competence has been properly tested.

Someone speaks confidently during a meeting and everyone assumes they understand the problem.

Someone else carefully explains the limitations, asks questions and points out uncertainties, and can appear less decisive.

Over time, organisations can accidentally confuse presentation confidence with actual capability.

That's dangerous because the two are different.

The person who confidently promises that a project will be finished in two weeks may simply be better at sounding certain.

The person who says, "We need to examine the dependencies before I give you a realistic timeline," may have more experience with what actually goes wrong.

Good workplaces need both confidence and humility.

The goal isn't to make everyone hesitant.

It's to make sure confidence is supported by competence, evidence and accountability.

A More Honest Way to Hold This Idea

Even with the legitimate scientific criticism surrounding how the Dunning-Kruger effect should be interpreted, the underlying question remains valuable because we've all experienced some version of it from both sides.

We've all been the person who confidently thought they understood something and later discovered they didn't.

We've all encountered somebody whose confidence seemed wildly disconnected from their actual knowledge.

And we've probably all underestimated ourselves somewhere too.

The mistake is turning this into a hierarchy of intelligent people and stupid people.

Human knowledge doesn't work that neatly.

You can be highly skilled in one area and completely inexperienced in another. You can be right for the wrong reasons. You can be wrong while sounding incredibly convincing. You can know the answer but explain it badly. You can understand a subject deeply and still be uncertain about a specific question.

That's why the most useful lesson isn't:

"Some people are too stupid to know they're stupid."

It's:

"My confidence is not automatically a reliable measurement of my competence."

That applies to everyone.

Including me.

Including you.

And especially the person who is absolutely certain they don't need to hear it.

Frequently Asked Questions

Q.1 What exactly is the Dunning-Kruger effect?

The Dunning-Kruger effect refers to research by David Dunning and Justin Kruger examining how accurately people can judge their own performance. Their 1999 paper found that lower-performing participants tended to substantially overestimate their performance, and the researchers proposed that limited knowledge can make it difficult to recognise one's own errors and limitations.

Q.2 Does the Dunning-Kruger effect mean unskilled people think they're geniuses?

No. That's an exaggerated internet interpretation. The original research found that low-performing participants overestimated their relative performance, but that is different from believing they are geniuses or deliberately acting superior. The central issue was inaccurate self-assessment.

Q.3 Is the Dunning-Kruger effect scientifically controversial?

There has been substantial debate about how the effect should be interpreted. Some researchers have argued that statistical factors, including regression to the mean, can explain part of the pattern observed in this type of data. The famous graph commonly associated with the effect also does not accurately represent the original study's actual data.

Q.4 Do experts also underestimate themselves?

The original study found that high-performing participants tended to underestimate their relative performance. One explanation proposed by the researchers was that people who find tasks relatively easy may assume those tasks are similarly easy for others.

Q.5 Why do confident people sometimes appear more knowledgeable?

Confidence is immediately visible, while competence usually requires evidence, experience or testing to evaluate. Someone who speaks without hesitation can therefore appear more knowledgeable than someone who carefully explains uncertainty. But confidence alone cannot establish whether a claim is accurate.

Q.6 How can I avoid overestimating my own knowledge?

Seek feedback, test your understanding, look for evidence that could challenge your position, and distinguish familiarity from genuine expertise. One useful question is: "What would I need to know to realise that I'm wrong?" The answer can reveal how much of the subject you actually understand.

Q.7 Is everyone who is confident experiencing the Dunning-Kruger effect?

No. Confidence is not itself evidence of the effect. A highly knowledgeable person can be very confident because they have strong evidence and experience behind their judgment. The concept is about the relationship between actual ability and the accuracy of self-assessment, not about labelling confident people as unskilled.

The Point Isn't to Spot the Dumbest Person in the Room

The next time someone confidently explains something they barely understand, you may be tempted to use the phrase Dunning-Kruger.

Maybe they're genuinely overestimating their ability.

Maybe they're simply wrong.

Maybe they know something you don't.

Or maybe you don't have enough information to judge them at all.

That's precisely why the concept is more useful as a question than as an insult.

The original research wasn't created to give us a clever label for people we disagree with. It was an attempt to understand why our perception of our own ability doesn't always perfectly match reality.

And that's something worth remembering long after you've forgotten the famous graph.

Before saying, "I know exactly how this works," ask yourself how much you've actually studied.

Before dismissing someone else's uncertainty, ask whether they might simply understand the complexity better than you do.

Before forwarding a confident message, ask where the evidence came from.

And before diagnosing somebody else with the Dunning-Kruger effect, perhaps ask the most uncomfortable question of all:

What if I'm the one who doesn't know enough to realise how much I don't know?

That's not a very satisfying question.

But it might be the most useful one.

Search This Blog

Akkiblogpost