What the data actually shows
The curse of knowledge was named and studied by Camerer, Loewenstein and Weber, who showed in economic experiments that better-informed people could not fully set aside what they knew when predicting the judgments of the less-informed. Knowing the answer contaminated their estimate of what someone without that knowledge would think. The bias was systematic, not occasional, and it persisted even when people were motivated to correct for it.
A vivid demonstration is Elizabeth Newton's 'tappers and listeners' study. People asked to tap out the rhythm of a well-known song estimated that listeners would recognize it around half the time; in reality, listeners identified it only a small fraction of the time. The tappers could hear the full tune in their heads, so they could not imagine how sparse and ambiguous the bare taps sounded to someone who could not. Their own knowledge made the message feel far clearer than it was.
Chip and Dan Heath popularized this work in 'Made to Stick,' arguing the curse of knowledge is a central reason expert communication fails — experts forget what it is like to be a novice and pitch their explanations to people who already share their context. The pattern shows up across teaching, technical writing, medicine, and management: the more expert the communicator, the easier it is to overshoot the audience without realizing it.
Why this feels different from how it actually is
It feels like the listener is being slow rather than that the explanation is mispitched, because to the expert the missing steps are not visibly missing — they have been automated into intuition. What was once effortful reasoning has become a single obvious leap, so the expert genuinely does not see the gap they are skipping over.
It also feels like jargon is plain language. Specialist terms that took years to absorb stop registering as specialist; they feel like the natural words for the thing. The expert is not showing off — they have simply lost easy access to the everyday phrasing a beginner would need, the same way the tappers could no longer hear how bare their taps sounded.
And both sides tend to misread the failure as being about ability rather than perspective. The novice may feel stupid; the expert may feel they are surrounded by people who do not pay attention. In fact neither is the issue — the difficulty is structural, a predictable consequence of how thoroughly knowledge rewires the way a topic appears to the person who holds it.
The very expertise that makes someone good at the work is what makes it hard to remember what a newcomer does not yet see.
What the research says to do about it
The most reliable countermeasure is to get concrete and to externalize the beginner's perspective rather than imagine it. The 'Made to Stick' work emphasizes concrete examples, plain language, and stories over abstraction, precisely because abstraction is where the curse of knowledge does its damage. Replacing 'the principle' with a specific worked example forces the hidden steps back into view.
Testing the explanation on an actual novice beats trying to simulate one in your head, since the bias is exactly the inability to simulate the uninformed mind accurately. Watching where a real beginner gets lost, or asking them to explain it back, surfaces the skipped steps that you can no longer see on your own.
It also helps to deliberately name and unpack jargon, and to ask 'what would someone need to already know for this to make sense?' before explaining. Slowing down to make the implicit explicit runs against the grain of expertise — which is built on making things implicit — which is exactly why it has to be done on purpose rather than left to intuition.
What the research says does not help
Simply telling experts to 'explain it more simply' has limited effect, because the bias operates below awareness — they already believe they are being clear, just as the tappers were sure their song was recognizable. Without concrete feedback from a real novice, good intentions do not reveal the steps being skipped.
Repeating or restating the same explanation more slowly or more loudly rarely closes the gap, since the problem is not pace or volume but a missing rung in the ladder. If the explanation starts above where the listener is standing, saying it again from the same height does not help them reach it.
Reading the difficulty as a sign of low intelligence — in either direction — does not help and is not accurate. The novice is not failing to grasp something easy, and the expert is not arrogant for skipping steps. Treating it as a character or ability problem misses that it is a predictable, well-studied bias that affects everyone who knows a subject well.
Restating the same explanation more slowly doesn't add the missing rung in the ladder.
What this looks like in real life
The tapper who's sure the song is obvious
Tapping out a familiar tune, you can hear the whole thing in your head — melody, words, everything — so it feels unmistakable. To the listener, all that arrives is a bare sequence of knocks. That gap between what you experience and what you transmit is the curse of knowledge in miniature: your own knowledge fills in everything the other person never receives.
The brilliant colleague who's a frustrating teacher
The deeper someone knows a subject, the more the beginner's starting point becomes invisible to them — so the explanation lands at the wrong altitude, pitched to people who already share their context. Both sides tend to misread this as being about ability: the novice feels slow, the expert feels unheard. In fact the difficulty is structural, a predictable side effect of how thoroughly knowledge rewires the topic for the person who holds it.
Real numbers in context
The tappers-and-listeners study is the memorable number here, and it deserves to be quoted with care: tappers estimated listeners would recognize the tune around 50% of the time, while actual recognition was only a small fraction of that — a gap that captures how vividly your own knowledge fills in what the listener never receives. Treat the exact figures as illustrative of the pattern rather than as precise constants.
More broadly, the curse of knowledge is robust as a direction of bias rather than a single statistic: across studies, better-informed people systematically overestimate what the less-informed know or can infer. The transferable lesson is that the difficulty experts have explaining simple things is normal and predictable — a feature of expertise itself, not a flaw in the expert or the learner.