IX Coach

Circle of Competence

Know what you know, know what you do not know, and act accordingly

What is the circle of competence and how do you use it?

Warren Buffett and Charlie Munger’s circle of competence is the idea that each person has a domain where their knowledge is genuinely deep, and that the highest-quality decisions come from staying inside that domain — or clearly knowing when you have stepped outside it. The hard part is not identifying the circle but accurately estimating its size.

Buffett has described the circle of competence as one of the most important ideas in his investment approach: 'What counts for most people in investing is not how much they know, but rather how realistically they define what they don’t know.' The principle extends well beyond investing — any domain where overconfidence is costly benefits from a clear-eyed assessment of where genuine expertise ends and where guesswork begins. Here are the practices that make this mental model operational, with honest evidence.

Practices

How these practices connect

Every one of these practices treats competence as a boundary that must be actively located rather than a quality you simply possess. The shared mechanism is that confidence and accuracy come apart most sharply near the edge of what you actually know, so the boundary has to be found through honest mapping and feedback, marked explicitly, and then used as the actual switch that decides whether you act alone or bring in outside judgment. Without that boundary-work, the natural default is to treat the feeling of confidence as a proxy for competence, which is exactly the assumption that fails at the edges.

How do I know if something is actually inside my circle of competence?

You map your circle based on where you have been right over time, not on how long you have been exposed to a topic or how interested you feel in it — time spent and enthusiasm are not evidence of accuracy. Then you mark the boundary explicitly as inside, outside, or unsure, because an edge that is never named is invisible and gets crossed without any warning. The most reliable signal is track record under real feedback, not the confidence you feel while thinking about the domain.

Why do smart people make bad decisions outside their expertise?

Being genuinely skilled in one domain does not transfer competence to an adjacent one, but it does transfer the feeling of confidence, since that feeling comes from familiarity and fluency rather than from actual accuracy. People also tend to substitute general intelligence for domain expertise, assuming that being smart in one area makes them reliable in nearby ones. The result is that smart people can be confidently wrong precisely at the edges of their circle, where the mismatch between felt confidence and real accuracy is largest.

Do you have to follow these steps in order?

No, but mapping your circle and marking its edges are the foundation the other practices depend on — you cannot reliably decide when to defer to your own judgment, or notice what you are missing, until you have an honest sense of where your competence actually ends. Expanding the circle and admitting uncertainty out loud can happen at any time and reinforce the mapping, but they are not a substitute for having done it. In practice these operate more as a standing habit you cycle through than a strict sequence.

Is reading a lot about a subject the same as expanding your circle of competence?

No. The circle only expands when you have made real decisions in the new domain, gotten accurate feedback on them, and revised your understanding based on being wrong — reading, taking a course, or simply feeling interested does not do this. Treating exposure time as growth is exactly the mistake this practice is meant to catch, because it lets you believe you have expanded your circle when you have only accumulated unverified impressions. Genuine expansion requires the same kind of corrective feedback loop that honest mapping depends on in the first place.

Why would admitting I don't know something make my decisions better?

Saying 'I don't know' clearly, instead of hedging with a vague 'it could be A or B or C,' signals real epistemic humility to the people around you, and they tend to respond by volunteering information they would have withheld from someone projecting certainty. It also keeps you from getting locked into a position by having stated it publicly, which matters because confident public commitments are hard to walk back even after new information arrives. The social cost of admitting uncertainty is usually smaller than the decision cost of false confidence, though that tradeoff can flip in competitive settings where uncertainty gets used against you.

What's the actual difference between marking the edge of your circle and asking what you're missing?

Marking the edge is a mapping exercise you do in advance, labeling territory as inside, outside, or unsure so the boundary exists before you need it. Asking 'what am I missing?' is a forcing function you apply in the moment, right before a high-stakes decision, to counteract the confident feeling that can mask unseen gaps even inside territory you have already marked as your own. The edge-marking makes the boundary visible; the question checks whether confidence at decision time is still tracking that boundary accurately, since the question only works if you seek perspectives outside your own frame rather than answering it yourself.

How do you choose between these?

The situation that should drive your choice is where you are in a decision, not what kind of person you are. Before you have acted at all, mapping your circle and marking its edges is the right move, since you need the boundary before you can use it. When a specific high-stakes, hard-to-reverse decision is actually in front of you, staying inside the circle and asking what you might be missing are the practices that apply directly, because that is the moment the boundary either protects you or fails to. When you are not under decision pressure and instead have the chance to genuinely learn, deliberate expansion through real feedback is what actually grows the circle, whereas admitting uncertainty out loud is useful continuously, in conversation, any time projecting false confidence would cost you accurate information from the people around you.

What these approaches do not settle

All six practices rest on observational evidence rather than randomized trials, and several are principled extensions of overconfidence and debiasing research rather than independently studied phenomena in their own right — the edge-overconfidence mechanism and the premortem-style 'what am I missing?' question both fall into this category. The Dunning-Kruger finding that anchors the mapping practice has had its original effect size questioned by recent reanalyses, even though the directional pattern of low-competence overconfidence remains supported. And admitting uncertainty out loud is not universally beneficial: in competitive settings, stating uncertainty can be used against the speaker, so the practice carries a real situational tradeoff rather than a guaranteed social benefit.

Practice this with IX Coach

Practice this with IX Coach

Frequently asked questions

What is the circle of competence and how do you use it?

Warren Buffett and Charlie Munger’s circle of competence is the idea that each person has a domain where their knowledge is genuinely deep, and that the highest-quality decisions come from staying inside that domain — or clearly knowing when you have stepped outside it. The hard part is not identifying the circle but accurately estimating its size.

How do you map your actual circle honestly?

Most people overestimate how much of their knowledge is deep versus thin or borrowed. The Dunning-Kruger effect — the well-replicated finding that people with limited knowledge in a domain rate their competence higher than those with genuine expertise — operates most damagingly at the boundary of the circle. An honest mapping process surfaces the inflated portions before they drive bad decisions.

How do you know and mark the edges of your circle?

An unmapped edge is invisible and therefore crossable without warning. Explicitly marking the boundary — "I know retail banking well; commercial real estate I only know the surface of" — creates a cognitive speed bump before decisions in adjacent territory. The edge is where confidence is highest and accuracy is most likely to have diverged, making it the most dangerous point in the circle.

How do you stay inside for high-stakes, irreversible decisions?

The circle of competence is most practically valuable not as an abstract map but as a decision rule: when the stakes are high and the decision is hard to reverse, restrict independent judgment to domains where you are genuinely expert. Outside the circle, the expected value of autonomous decision-making falls — you are not better than average, and you may be worse because you cannot accurately estimate your own error rate.

How do you expand the circle deliberately, not aspirationally?

The circle only expands when genuine feedback corrects your model — when you have been wrong, understood why, and revised. Reading widely, taking courses, or feeling interested in a domain does not expand the circle; making real decisions, getting accurate feedback, and updating from it does. Aspirational expansion ("I’m learning finance") conflates intention with achievement.

How do you ask 'What am I missing?' before major decisions?

Confidence is a feeling generated by familiarity and fluency, not by accuracy. High confidence inside a domain can mask the unseen edges — what you would have needed to know to include the unknown unknowns. The question "what am I missing?" is a forcing function that recruits deliberate reflection to counteract the automatic confidence signal.

How do you admit uncertainty out loud, especially in your own domain?

Admitting uncertainty has two effects: it signals epistemic humility to collaborators, who then contribute information they might have withheld from someone projecting certainty; and it prevents the speaker from being locked into a position by public commitment. The social cost of "I don’t know" is usually lower than the decision cost of false confidence.

How do I know if something is actually inside my circle of competence?

You map your circle based on where you have been right over time, not on how long you have been exposed to a topic or how interested you feel in it — time spent and enthusiasm are not evidence of accuracy. Then you mark the boundary explicitly as inside, outside, or unsure, because an edge that is never named is invisible and gets crossed without any warning. The most reliable signal is track record under real feedback, not the confidence you feel while thinking about the domain.

Why do smart people make bad decisions outside their expertise?

Being genuinely skilled in one domain does not transfer competence to an adjacent one, but it does transfer the feeling of confidence, since that feeling comes from familiarity and fluency rather than from actual accuracy. People also tend to substitute general intelligence for domain expertise, assuming that being smart in one area makes them reliable in nearby ones. The result is that smart people can be confidently wrong precisely at the edges of their circle, where the mismatch between felt confidence and real accuracy is largest.

Do you have to follow these steps in order?

No, but mapping your circle and marking its edges are the foundation the other practices depend on — you cannot reliably decide when to defer to your own judgment, or notice what you are missing, until you have an honest sense of where your competence actually ends. Expanding the circle and admitting uncertainty out loud can happen at any time and reinforce the mapping, but they are not a substitute for having done it. In practice these operate more as a standing habit you cycle through than a strict sequence.

Is reading a lot about a subject the same as expanding your circle of competence?

No. The circle only expands when you have made real decisions in the new domain, gotten accurate feedback on them, and revised your understanding based on being wrong — reading, taking a course, or simply feeling interested does not do this. Treating exposure time as growth is exactly the mistake this practice is meant to catch, because it lets you believe you have expanded your circle when you have only accumulated unverified impressions. Genuine expansion requires the same kind of corrective feedback loop that honest mapping depends on in the first place.

Why would admitting I don't know something make my decisions better?

Saying 'I don't know' clearly, instead of hedging with a vague 'it could be A or B or C,' signals real epistemic humility to the people around you, and they tend to respond by volunteering information they would have withheld from someone projecting certainty. It also keeps you from getting locked into a position by having stated it publicly, which matters because confident public commitments are hard to walk back even after new information arrives. The social cost of admitting uncertainty is usually smaller than the decision cost of false confidence, though that tradeoff can flip in competitive settings where uncertainty gets used against you.

What's the actual difference between marking the edge of your circle and asking what you're missing?

Marking the edge is a mapping exercise you do in advance, labeling territory as inside, outside, or unsure so the boundary exists before you need it. Asking 'what am I missing?' is a forcing function you apply in the moment, right before a high-stakes decision, to counteract the confident feeling that can mask unseen gaps even inside territory you have already marked as your own. The edge-marking makes the boundary visible; the question checks whether confidence at decision time is still tracking that boundary accurately, since the question only works if you seek perspectives outside your own frame rather than answering it yourself.

Research

Related concepts

Articles that cover this