Circle of Competence

In 1956, when Tom Watson Jr. took over as president of IBM, the company was a thoroughly American business with deep roots in the US market and almost no presence overseas. Watson believed that international expansion was the company's future, and he hired a man named Jacques Maisonrouge to run IBM's foreign operations. Years later, when asked what made Maisonrouge so effective, Watson gave an answer that has been quoted ever since. "He was brilliant," Watson said. "But the most important thing about him, I think, was that he knew the size of his circle of competence and stayed within it." The phrase stuck. By the 1990s, Warren Buffett and Charlie Munger had adopted it as one of their organizing principles for investment. By the 2000s, it had become one of the most-cited mental models in popular thought.

The idea is, on its surface, almost embarrassingly simple. Some things you understand well enough to act on. Most things you don't. The skill of knowing which is which, and confining your high-stakes decisions to the first category, is what Buffett came to call operating inside your circle of competence. Done well, it produces a slow, durable kind of effectiveness that survives bad markets, bad decades, and bad luck. Done badly, it produces the spectacular failures of smart people convinced their intelligence transfers across domains it doesn't.

What makes the principle harder than it sounds is that the skill isn't really "know what you know." That part is easy enough. The hard part is knowing what you don't know, which is structurally different. Inside your circle, you have feedback, experience, and calibration; outside it, you have confident-feeling guesses that look from the inside exactly like real knowledge. The difference between a competent operator and a dangerous one is not their map of what they understand. It's their map of where their understanding ends.

This essay is about that boundary: where it lives, why it's so often invisible, how the world's best decision-makers stay on the right side of it, and where the principle itself can quietly mislead you.

What the circle actually is

The circle of competence is a metaphor for the set of things you understand well enough to make accurate predictions and good decisions about. Inside the circle, your judgment is calibrated. Your intuitions are mostly right. Your sense of "this seems off" is grounded in actual pattern recognition built up over years of feedback. Outside the circle, none of that holds. Your intuitions are guesses, your pattern recognition is matching to the wrong patterns, and your sense of confidence is uncoupled from your accuracy.

Crucially, the circle isn't about credentials, IQ, or general intelligence. A specialist in 19th-century French poetry has a circle of competence too: not in physics or finance or sales, but in 19th-century French poetry. Their judgment in that domain is sharper than anyone outside it; their judgment outside it is no better than yours. The circle is domain-specific knowledge that has been earned through actual exposure, feedback, and correction over time. You can't read your way into one. You can only practice your way into one.

The metaphor of a circle is useful because it has three distinct zones, and each one calls for different behavior. Most popular treatments of the concept collapse all three into a single "in or out" binary, which loses the most important insight: the boundary itself is its own zone, and it's the one that determines whether you're a careful operator or a reckless one.

This three-zone model contains the entire practical insight of the principle. The interior of the circle is comfortable and familiar — that's where most of us already live. The exterior is also relatively safe, because we know we don't know, and we act accordingly. The boundary is where the trouble lives, and most miscalibrated decisions in life come from the same structural mistake: the person making the decision was on the boundary but felt like they were in the interior. The whole discipline of operating inside your circle is, more accurately, the discipline of recognizing the boundary in real time.

Tom Watson, Buffett, and Munger

The phrase itself goes back to Tom Watson Jr.'s description of Maisonrouge in the 1950s, but it was Warren Buffett and Charlie Munger who turned it into a working principle for serious decision-making. Buffett began using the phrase publicly in his shareholder letters in the 1990s, citing it as one of the foundations of Berkshire Hathaway's approach. The Berkshire model, in his telling, is mostly about not investing in things they don't understand: skipping the technology bubble of the late 1990s, avoiding most international companies, declining to bet on currencies or commodities they couldn't model. The discipline isn't picking winners; it's refusing to bet outside the small set of businesses where their judgment was actually useful.

That last sentence is the entire principle compressed into eleven words. Buffett's insight is that the size of your circle doesn't matter much. A small circle of true competence outperforms a vast circle of half-knowledge, every time, over a long enough horizon. What matters is whether you can accurately identify the edge — whether you can tell, when a decision arrives, on which side of the line it sits. Buffett's circle, by his own admission, is small: a few dozen American businesses with predictable cash flows. Munger's is somewhat broader. Neither covers the technology sector or biotech or the venture-capital game, all of which have produced enormous returns for people whose circles include them. The Berkshire view is that this doesn't matter. Their job is to operate inside their own circle, not envy other circles.

This stance has often been criticized as overly conservative, particularly during periods when sectors outside the Berkshire circle were producing outsized returns. The criticism missed the point. The discipline isn't about maximizing the upside of every era; it's about never being on the wrong side of a catastrophic mistake. Over a long enough time horizon, that asymmetry compounds. The investors who operate inside small, well-defined circles tend to survive every bubble and crash, because they don't make positions in the things that crash hardest. The investors who range freely across whatever is currently exciting tend to do brilliantly during one cycle and disastrously during the next. The Berkshire approach is the slow, durable version of investing because it's structurally allergic to the catastrophic-mistake category.

The boundary matters more than the interior

The single most important thing to understand about circles of competence is that the boundary is the operative concept, not the interior. Knowing what you know is the easy part. What separates good decision-makers from confident ones is knowing where what you know stops.

This sounds like a small distinction. It isn't. It's the difference between two kinds of people who, from the outside, look identical. The first type has expertise in their field and acts decisively within it; outside their field, they defer, ask questions, or decline to opine. The second type has the same expertise in the same field but applies their confident style across all topics indiscriminately, treating the confidence they've earned in one domain as if it transferred to all others. The first type is a competent specialist. The second is the dangerous-doctor archetype, the engineer-with-a-political-blog, the founder-turned-policy-expert. The internal experience for both people is similar — the second one feels just as confident as the first. The external outcomes are very different.

This is why "operating inside your circle of competence" is more accurately described as operating with awareness of where your circle ends. The interior takes care of itself; you naturally feel confident there because confidence is calibrated to actual knowledge. The boundary is where vigilance has to be applied deliberately, because the boundary feels exactly like the interior from the inside. There's no internal warning bell that says "you're now near the edge." Either you've cultivated the external habit of checking, or you haven't.

Why the boundary is invisible from inside

If recognizing your boundary is so important, why don't more people do it? The answer is that the boundary is, by its nature, the hardest thing to see. The reason has been documented extensively in psychology, where it's known as the Dunning-Kruger effect: people with limited knowledge in a domain consistently overestimate how much they know, while people with deep expertise tend to underestimate it. The mechanism is simple but devastating: the cognitive skills required to recognize what good performance looks like in a domain are the same skills required to perform well in it. If you don't have them, you can't tell that you don't have them.

The Dunning-Kruger curve has a very specific shape that matters for circle of competence. Confidence rises quickly with the first bit of exposure, reaches a peak somewhere in the early-learner zone, drops sharply once the learner gains enough perspective to see how much they don't know (the so-called "valley of despair"), and then climbs slowly back up as actual expertise develops. The peak overconfidence point is exactly where the boundary lives. You feel maximum competence at the moment you have minimum competence relative to your sense of it, and that's the moment when high-stakes decisions are most dangerous.

This has a practical implication that's worth absorbing carefully. If you've recently started learning about a new field — read a few books, taken a course, watched some videos — your sense of competence in that field is almost certainly inflated relative to your actual ability. The first hundred hours feel like a great deal of progress because you're moving fast against a baseline of nothing. The next thousand hours feel like crawling, because you've started to see the actual contours of the domain and how much remains. People who quit at hour 100 often have stronger opinions about the field than people at hour 5,000, and the asymmetry is almost entirely due to the Dunning-Kruger curve, not actual knowledge.

High stakes, low stakes, and where to risk being wrong

Operating inside your circle of competence isn't a universal rule for all decisions. It's a rule for high-stakes decisions. Most of life is low-stakes, and being wrong about a lot of things doesn't matter much. You can have opinions about restaurants, music, sports teams, and minor political questions all day long without any cost to being wrong. The principle of staying inside your circle of competence is not about ceasing to have opinions. It's about being deliberate about which opinions you act on with significant resources at risk.

The matrix that follows is the practical operationalization of the principle. The diagonal from "low-stakes inside" to "high-stakes outside" is the rough order of how much vigilance each kind of decision deserves. The bottom-right quadrant — high-stakes decisions made outside your circle — is where catastrophic mistakes live, and the discipline of circle of competence is mostly about staying out of that quadrant by either (a) declining the decision, (b) deferring to someone whose circle includes the question, or (c) doing the deep work to bring the question inside your circle before committing.

The "playground" quadrant — top-right, low-stakes decisions outside your circle — is worth a special note. This is the zone where circles legitimately grow. You try a new sport, a new instrument, a new programming language, a new career-adjacent skill, all under conditions where being bad at it doesn't cost you much. The mistakes you make there are how the boundary moves. The danger isn't being outside your circle in this quadrant. It's mistaking time spent in the playground for actual expansion of the circle, which leads us to one of the most important and least-discussed parts of the principle.

How circles legitimately grow

People often ask, after learning about circle of competence, how to expand theirs. This is the right question, but the answer is less satisfying than most people want it to be. Circles don't grow because you read about a topic, take a course on it, listen to a podcast, or chat with experts. Those activities can produce vocabulary, frame, and orientation, but they don't produce competence on their own. Competence comes from action paired with feedback over time, and feedback that's tight and accurate enough to actually correct your mistakes.

This is why the Dunning-Kruger curve is shaped the way it is. The first hundred hours of any new field are mostly accumulating concepts and frameworks, which feel like progress because they're new and structured. Real competence starts to develop somewhere around hour 1,000, where you've made enough mistakes and gotten enough feedback for your intuitions to start being grounded in the actual structure of the domain rather than in the textbook description of it. Most people never reach that point in most fields, and that's fine — you don't need to be competent at most things. But it's worth being honest about the difference between "having read about a domain" and "having a circle of competence in it." They feel similar from the inside. They produce wildly different decision quality.

The legitimate path to expanding a circle is uncomfortable because it requires you to spend a long time being bad at the new thing while feeling like you're not making progress. The fast version — read books, form opinions, act — produces the inflated confidence that the principle is specifically designed to protect you from. There is no shortcut, but there is a structural answer: spend the early years operating in the low-stakes quadrant of the new domain, accumulating mistakes and feedback at a scale where the mistakes don't cost you much, until your judgment in the new area starts to converge with the judgment of people who actually live there. Then, and only then, you can begin treating the new domain as part of your circle for high-stakes purposes.

Where the discipline does its real work

The circle of competence pays off most dramatically in domains with significant downside, slow feedback, or strong incentives toward overconfidence. A few obvious examples are worth walking through.

Investing

The Berkshire approach is the canonical case. Buffett and Munger have spent six decades operating inside a small, well-defined circle and outperforming most investors with much larger nominal "expertise." Their edge isn't superior insight into more domains. It's the discipline of ignoring everything outside their domain, even when those domains are producing 10x returns elsewhere. Most amateur investors, by contrast, range freely across whatever sector is currently exciting and pay the variance tax that ranging-without-edge produces.

Founding companies

Founders frequently destroy companies by extending decisively into adjacent businesses where their original competence doesn't apply. The retailer who decides to launch a media company, the SaaS founder who decides to expand into consumer hardware, the agency that decides to build a product. Each transition involves leaving a circle where judgment was calibrated and entering one where it isn't, usually with the original company's resources at stake. The successful expansions almost always involve either bringing in new leadership whose circle includes the new domain or treating the expansion as a long playground period before committing real resources to it.

Hiring outside your function

Hiring for a role you've never held yourself is one of the most common circle-of-competence violations in management. A technical founder hiring their first sales leader, a marketing executive hiring their first head of engineering, a CEO hiring a CFO without finance background. The standard recommendation — get help from someone whose circle includes the role you're hiring for — is almost always correct, and almost always ignored, because hiring feels like it's a meta-skill that should be transferable across functions. It isn't. The signals that distinguish a great hire from a mediocre one are domain-specific, and your reading of an interview in a function you've never worked in is barely better than chance.

Public commentary

The internet has produced an entire industry of professionals confidently commenting on domains adjacent to but outside their circle of competence. The economist holding forth on epidemiology, the technology executive on geopolitics, the academic on policy, the journalist on science. The content is often well-written and superficially reasonable, but the commentator's sense of certainty almost always exceeds what their actual familiarity with the domain warrants. The audience, lacking a way to distinguish, often takes the content as expert opinion. The cumulative effect is a public discourse where most "expert commentary" is actually edge-of-circle commentary by people who haven't recognized that their circles end before they think they do.

Where the principle fails

The "never grow" trap

Taken too literally, "stay inside your circle" becomes an excuse for never learning anything new, never trying anything outside your existing expertise, and treating the world as if your current circle is the only sensible place to operate. This is a misreading. The principle is about high-stakes decisions, not about life in general. The playground quadrant — low-stakes exploration outside your current circle — is exactly how circles legitimately grow, and avoiding it entirely is its own form of stagnation. The discipline is staying out of high-stakes mistakes, not staying out of new experiences.

Recursive uncertainty

The principle has a structural problem: knowing the boundary of your circle is itself a skill, and that skill has its own circle of competence. People who are bad at self-assessment will be bad at locating their boundary, often placing it far further out than reality warrants. The classic Dunning-Kruger result is exactly this — people with low competence in a domain are also typically poor at recognizing their lack of competence, which means they'll draw the boundary too generously. The principle works best for people who are already somewhat self-aware; it works worst for the people who most need it.

Some decisions can't be deferred

"Don't decide outside your circle" assumes someone else can decide for you. Sometimes no one can. Many of life's most important decisions — major career changes, parenting decisions, end-of-life medical choices, founding a company in a new field — are decisions where no expert has direct knowledge of your specific situation, and where the cost of inaction is comparable to the cost of acting badly. In those cases, the discipline becomes making the decision with full awareness of where your circle ends, not refusing to make it. This is harder than the simple version of the principle suggests.

Markets reward overconfidence in some sectors

Certain professions — sales, leadership, consulting, public commentary — actively reward visible confidence regardless of underlying competence. Operating inside your circle in these contexts can be career-limiting in the short term, even when it's correct in the long term. This isn't a flaw in the principle; it's an honest observation that the social rewards for circle violation are real, and that practitioners of the discipline often pay a near-term professional cost. The compensation comes over decades, in the form of having a track record uncontaminated by confident mistakes, but the early years can be rough.

Connection to other models

The circle of competence sits naturally next to several other essays in this collection. Calibration is the meta-skill underneath it: the audit on whether your confidence inside your circle is actually justified by your hit rate. Hanlon's Razor, in its limits section, hits the same observation about pattern-versus-incident vigilance. Second-Order Thinking overlaps with the high-stakes-outside-circle problem: most second-order failures involve decisions made by people whose first-order knowledge of the system is solid but whose second-order knowledge of how the system reacts is outside their circle. The Sagan Standard is structurally related: extraordinary claims require extraordinary evidence, and you only know which claims are extraordinary if you have a calibrated prior, which you only have inside your circle.

How to actually use it

Operating inside your circle of competence becomes a habit through deliberate self-assessment over time, paired with the practical discipline of slowing down at the boundary. Here's the working version.

Tom Watson Jr. described his colleague Maisonrouge in the 1950s using a phrase he probably hadn't thought much about. Decades later, Buffett and Munger turned that phrase into the operating philosophy of one of the most consistently effective investing partnerships in history. The principle wasn't original to any of them — versions of it have appeared in philosophy, military strategy, and the practical wisdom of every craftsman who ever apprenticed under a master and learned the limits of their tools. But the modern version, with its emphasis on the boundary rather than the interior, is the one most worth internalizing.

The discipline doesn't make you smarter. It makes you slower at the right moments. In a world that rewards confident takes, fast decisions, and the appearance of broad expertise, slowing down at the edge of your circle is the most uncool thing you can do. It also happens to be one of the most reliable predictors of whose track record will look good in twenty years and whose will be a series of impressive-sounding stories about decisions that didn't work out. The track record is everything. The track record is built one boundary check at a time.

Buffett's eleven words remain the cleanest statement of the principle ever offered: "The size of that circle is not very important; knowing its boundaries, however, is vital." Most people get this exactly backwards. They spend their lives trying to grow the circle and almost no time learning where it ends. The few who reverse the priority end up doing very well, very quietly, for very long.