Trap/Cognitive Bias/No. 0233
Curse of Knowledge
The curse of knowledge is a bias in which informed people struggle to judge what less-informed people know or understand. Named by Camerer, Loewenstein and Weber in 1989, it can distort predictions of others’ judgments and lead to explanations that omit needed context.
- Evidence
- Well established
- Read
- 6 min
- Links
- 10 connections
01You've seen this when…
- in life
You teach a friend your usual bread recipe. You tell them to knead until the dough feels ready, then realize they have no idea what ready feels like.
- at work
You send a new hire a setup checklist. They get stuck on the first step because you never mention the access request everyone on your team made years ago.
- out in the world
A city posts a flood warning for the lower catchment. Officials know which neighborhoods that means; residents scan the notice for their street.
02The idea
The missing step is often invisible to the person writing the instructions. They can follow them perfectly, because they silently supply everything the instructions leave out.
The curse of knowledge is the difficulty of setting aside what you know when judging what someone else knows or understands. Your own knowledge slips into your estimate of their starting point. An explanation feels complete because it works inside your head.
Large assumptions can hide in both technical language and plain words. A direction to restart the service might use familiar words while leaving a beginner unsure how to find the right service and its controls or whether restarting it will disconnect customers.
The effect also reaches beyond communication. Knowing the answer can distort your prediction of what someone without the answer will guess. Knowing a task well can distort your estimate of how long a beginner will take.
It overlaps with hindsight bias, but the questions differ. Hindsight bias makes an outcome seem more predictable after you learn it. The curse of knowledge makes another person’s uninformed perspective harder to reconstruct. You can fall into it without any outcome having occurred.
03Why it happens
- Your own knowledge is the easiest starting point. When estimating another person’s understanding, you have immediate access to yours. You must deliberately adjust for what they lack, and that adjustment can stop too soon. This form of egocentric bias can occur even with generous intentions.
- Practice hides the intermediate steps. Experience turns several actions into one familiar unit. Through chunking, checking permissions, selecting a workspace, and finding a command can feel like simply opening the tool. A beginner still has to solve each part.
- The answer changes the clues. Once you know what a diagram means, its labels seem informative. Without that knowledge, the same labels may not tell you where to begin. You can’t reliably test a beginner’s experience by rereading the diagram yourself.
- Feedback can conceal the gap. Listeners nod, copy your words, or avoid interrupting. You interpret that as understanding. Meanwhile, they are juggling unfamiliar terms and steps in limited working memory. Cognitive load can make it harder for them even to identify what they need to ask.
04A worked example
Imagine a finance analyst preparing a monthly-report checklist for a new teammate. She writes a six-step checklist with screenshots, then checks the document by following it herself. Everything works.
The new teammate produces a report for the wrong month. The analyst’s saved workspace always opens with the correct reporting period selected. A new account opens with the current month instead. Selecting the reporting period never became a written step.
What it looks like The teammate ignored a straightforward procedure. The analyst followed the same document and got the right result.
What’s actually going on They started with different knowledge and settings. The analyst supplied an unwritten check automatically and assumed the screenshot’s date filter made that check obvious. Her successful rehearsal tested whether an experienced user could complete the task, not whether the document taught an inexperienced user how.
What would have helped Watching someone with a fresh account use the checklist without coaching. Their first wrong turn would expose the missing instruction. The revised version would explain how to select and confirm the correct reporting period, then describe what a correct finished report should show. It would then be tested again with someone who hadn’t seen the earlier draft.
05How to spot it
06What to do instead
- Find the actual starting point. Ask the person to show how they currently approach the task. Use that demonstration to establish which terms, tools, and background facts they already know. A guess based on their job title can miss gaps.
- Test with someone who lacks the knowledge. Borrow the logic of usability testing: give them a realistic task, watch, and resist rescuing them immediately. Record where they hesitate and what they infer.
- Explain the structure before the sequence. Give a small conceptual model: what the main parts do and how they connect. Then provide the steps. A command is easier to understand when its purpose is clear.
- Check understanding through use. Ask the learner to perform a small task or explain what they would do next. Frame this as a test of your explanation, with the learner helping you assess how well it works.
- Make hidden prerequisites visible. List required access, starting settings, materials, and expected results. Use a fixed reader-check routine that prompts you to consider a beginner’s perspective each time.
- Separate beginner and experienced paths. Offer a short reference alongside fuller instructions. Extra explanation helps beginners but can obstruct experts, a problem studied as the expertise reversal effect.
07When it isn’t the curse of knowledge
Failed explanations have several possible causes. A listener who knows the background may disagree with the conclusion. They may lack access or have too little time or attention to follow the explanation. A speaker may deliberately withhold information. The curse specifically involves misjudging another person’s perspective because your own knowledge intrudes.
Explanations can build on an audience’s existing knowledge. Shared shorthand saves time when everyone has that background. Match the level of detail to a correct estimate of the audience.
Illusion of transparency is another nearby trap: you overestimate how clearly other people can read your feelings or intentions. Here, the problem is what your knowledge lets you understand.
Your knowledge of the answer limits the reliability of imagining yourself as a beginner. Treat perspective-taking as a draft, and observation of actual beginners as the check.
08Roots
In 1989, economists Colin Camerer, George Loewenstein, and Martin Weber approached knowledge as a possible disadvantage. Economic models often assume that a better-informed person can anticipate the judgments of someone with less information. An informed trader, for example, should be able to distinguish what they know from what other traders know.
The researchers tested whether people could actually make that separation. Their experiments asked informed participants to predict judgments made by less-informed participants. These numerical predictions made their understanding of others measurable. Information unavailable to the people being predicted nevertheless pulled the predictions toward what the informed participants knew. Setting that information aside was difficult even when doing so served their interests. The paper named the problem the curse of knowledge.
The idea traveled beyond economics because the same mismatch appears wherever knowledge must cross a gap. Research on experts predicting novice performance, including work by Pamela Hinds, examined how mastery can make a beginner’s difficulty hard to estimate. What began as a challenge to assumptions about information became a practical warning for teachers, managers, and designers: expertise can leave you with mistaken assumptions about what someone else needs to learn the subject.
09How solid is this?
Controlled studies show that privileged knowledge can distort predictions of less-informed people’s judgments and performance. The size varies by task and expertise; it doesn’t mean every expert explains badly, and perspective-taking alone is not a reliable fix.
10Connections
- Often confused with Hindsight Bias, Illusion of Transparency, Illusion of Explanatory Depth
- Countered byConceptual Model, Usability Testing, Cognitive Forcing Strategy
- Part ofEgocentric Bias
- See alsoCognitive Load Theory, Chunking, Expertise Reversal Effect
11Origin and sources
Colin Camerer, George Loewenstein, and Martin Weber named and experimentally examined the curse of knowledge in 1989, studying whether informed people could predict less-informed people’s judgments.
- [1]Camerer, C., Loewenstein, G., & Weber, M. (1989). The Curse of Knowledge in Economic Settings: An Experimental Analysis. Journal of Political Economy, 97(5), 1232–1254.
- [2]Nickerson, R. S. (1999). How we know—and sometimes misjudge—what others know: Imputing one's own knowledge to others. Psychological Bulletin, 125(6), 737–759.
- [3]Hinds, P. J. (1999). The curse of expertise: The effects of expertise and debiasing methods on predictions of novice performance. Journal of Experimental Psychology: Applied, 5(2), 205–221.
Suggest an edit· Updated 2026-10-02