The Dunning–Kruger effect is not the graph you've seen.

You have seen the chart. A line rockets up to a peak labelled Mount Stupid, plunges into a Valley of Despair, then climbs a gentle Slope of Enlightenment. Confidence on one axis, competence on the other. It is clean, it is funny, it explains your coworkers. It is also not in the paper, was never measured, and gets the actual finding backwards.

The version reproduced everywhere — peak, valley, recovery — shares its silhouette with the Gartner Hype Cycle, a marketing diagram about technology adoption. The “Mount Stupid” naming comes from internet culture, not from Kruger & Dunning.

The popular picture asks you to believe two things: that a little knowledge makes people maximally confident, and that gaining competence makes them crash into self-doubt before recovering. Both are dramatic. Neither is what Justin Kruger and David Dunning reported in 1999.Kruger, J. & Dunning, D. (1999). “Unskilled and Unaware of It.” Journal of Personality and Social Psychology, 77(6), 1121–1134.

01 / What they measuredWhat the original study actually did

Kruger and Dunning gave people real tests — logic, grammar, and judging which jokes are funny — then asked each person to estimate where they ranked against everyone else, as a percentile. Two numbers per person: how good you think you are, and how good you actually are.

Then they grouped participants into four quartiles by actual score and plotted the two numbers side by side. That graph — the real one — looks nothing like a mountain.

Numbers here are read from Figure 1 of the 1999 paper (the logic/“humor” studies) and are approximate but faithful to the shape and gaps reported. The pattern replicated across all four skills tested.

Read the real graph carefully and three things fall out, none of which the mountain captures:

Everyone clusters near “above average.” The self-assessment line barely moves. Whether you're genuinely terrible or genuinely excellent, your guess about yourself lands in roughly the same narrow band. People are bad at locating themselves, period.

The worst performers overestimate the most — but they don't think they're geniuses. They think they're slightly above average. The drama isn't sky-high arrogance; it's a quiet, universal failure to notice you're at the bottom.Dunning's later framing: the skills needed to produce a right answer are often the same skills needed to recognise a right answer. Lacking the first, you also lack the second — a “dual burden.”

The top performers underestimate themselves. Notice the solid line ends above the dashed one. The best scorers assumed others found the task as easy as they did. There is no Valley of Despair — if anything the experts are too modest, the opposite arc.

It was never a story about confident fools. It was a story about everyone, including you, having a blurry self-portrait.

One more way to see it: divide what people think of themselves by what they are — an overconfidence ratio. Above 1.0 you overrate yourself; below 1.0 you sell yourself short. The result isn't a mountain either. It's a slide.

02 / Run it on yourselfSo — how good are you, really?

Reading about a bias is easy. Catching yourself in one is the whole point. Below is a short test of exactly the kind Kruger and Dunning used: questions that feel obvious. Answer them, then — before you see your score — tell us how you think you did. Everyone's guesses are stored anonymously, so the graph at the end is built from real visitors. This page runs its own Dunning–Kruger experiment, and you're a data point.

03 / The twistThe effect explains its own misreading

Here is where it loops. The paper's core claim is that people who lack skill in a domain also lack the skill to see that they lack it. Now consider the millions confidently captioning the Mount Stupid graph as “the Dunning–Kruger effect” — people with little knowledge of the actual research, certain they've understood it.

They are standing on a small hill of knowledge, sure of the view, unable to see what they're missing. The misunderstanding of the effect is an instance of the effect. The map became the territory.

This is why it's such durable folk science. It flatters the sharer (“I'm past Mount Stupid, I see clearly now”) while doing exactly what the research warned about. A self-sealing little trap.

There's a quiet happy ending, though: the record is correcting itself. Wikipedia's article now opens with the real perceived-vs-actual graph — the flat self-assessment line, not the mountain — and the original 1999 paper is a free PDF away. The cartoon still circulates, but anyone willing to read for ten minutes can now find the subtler truth first.

04 / The honest footnoteAnd maybe even the real graph is a mirage

If we're going to be careful, we shouldn't stop at the popular version. The real finding has its own serious challenge — one the confident explainers also never mention.

Plot people's self-estimates against their scores and you'll get that telltale fan shape even from pure random numbers. It's partly regression to the mean and a statistical artifact called autocorrelation: noisy guessers at the bottom can only err upward, those at the top can only err downward. Some of the “effect” may be a property of the graph, not of human psychology.Nuhfer et al. (2017), Numeracy; Gignac & Zajenkowski (2020), Intelligence 80 — “the Dunning–Kruger effect is (mostly) a statistical artefact.” The debate is live, not settled.

The point isn't that the effect is fake. It's that the truth is subtle, contested, and a lot less fun to put on a slide than a mountain.

So the full, lazy-but-honest summary: a real and careful 1999 study found that self-assessment is compressed and the unskilled don't realise it; the internet flattened that into a triumphant cartoon that reverses the finding; and rigorous critics now argue even the careful version is partly an illusion of arithmetic. Three layers — and the folk version skips all three.

Next time someone shows you Mount Stupid, you'll know enough to be a little less sure. Which, finally, is the only honest place the research leaves any of us.

05 / Orders of thinkingHow deep do you want to go?

Every claim in this piece sits one level above the last. That's the real structure of understanding Dunning–Kruger — not a curve but a staircase down into your own uncertainty. Each step is more correct and less satisfying than the one above it. The bar on each rung is how sure you're allowed to feel.

So the question Dunning–Kruger really leaves you with isn't “who's on Mount Stupid?” It's “which rung am I standing on right now, and how would I know?” The only safe answer is: lower than you'd like, and you wouldn't.