This governs the quality of your thinking before it reaches any particular subject. Munger's claim is that most serious errors are not caused by lack of intelligence, they are caused by running one model where several are needed, and by a short list of psychological tendencies that fire reliably enough to be listed, watched for, and designed around. His method throughout is to go at the problem backward: he says he sought good judgment mostly by collecting instances of bad judgment.

Do

Solve it backward first. Munger borrows the algebraist Jacobi's rule, "invert, always invert," and applies it to everything: instead of asking how to help India, ask how you would most effectively damage India, then avoid those things. His parable is the rustic who says he wishes he knew where he was going to die so he could never go there, a man everyone laughs at who is in fact doing it correctly. [Source: "Talk One: Harvard School Commencement Speech (1986)"]

Give priority attention to whatever would disconfirm what you already believe. Munger's central case study is Darwin, an unremarkable student by conventional measures who succeeded specifically by turning first to the evidence that threatened his own hard-won theories. This is the exact opposite of what most people do with a favored belief, and it is now formalized in the world as the double-blind trial. [Source: "Talk One: Harvard School Commencement Speech (1986)"]

Assemble a small set of models from several disciplines and rank them by fundamentalness. Munger never took a course in chemistry, economics, psychology or business; he took the organizing ethos from physics and math instead, which is to master and routinely use the essential parts of each field and never explain with a less fundamental idea what a more fundamental one already explains. He argues roughly eighty to ninety models carry most of the load. [Source: "Talk Five: The Need for More Multidisciplinary Skills from Professionals"]

Run two tracks on every decision. Track one is the rational interests genuinely at stake. Track two is the subconscious influences that are, at that moment, pushing your conclusion around. Munger describes doing this permanently and continuously, not as an occasional audit. [Source: "Talk Two: A Lesson on Elementary, Worldly Wisdom (Part 1 — The Mental Models)"]

Assume incentive-caused bias in every recommendation, including your own. Munger's standing proof is Federal Express, whose broken night shift was fixed the instant workers were paid per completed shift rather than by the hour. The dangerous half is not deliberate corruption but decent people unconsciously rationalizing whatever their compensation structure rewards, which is why he once rejected the recommendations of two separate engineering teams and went looking for a third with nothing at stake. [Source: "Talk Eleven: The Psychology of Human Misjudgment (Part 1 — Tendencies One Through Nine)"]

Refuse yourself an opinion until you can state the other side better than its holders can. Munger calls this his iron prescription and reaches for it specifically whenever he notices himself drifting toward an intense position. It is a filter on which opinions you are allowed to have, not a debating technique. [Source: "Talk Ten: USC Gould School of Law Commencement Address (2007)"]

Put the important, rarely-used procedures on a written checklist and rehearse them. Munger's model is pilot training, and he lists the real institutional antidotes that already work: simulator practice, clinical see-one-do-one-teach-one training, Carl Braun's communication rules, Johnson and Johnson's habit of checking old acquisition predictions against what actually happened. His stated reason is that skills decay without use, his own calculus among them. [Source: "Talk Eleven: The Psychology of Human Misjudgment (Part 3 — Q&A, Reflections, and the Book's Closing Note)"]

When you need someone to act, appeal to their interest rather than to reason. Salomon Brothers' general counsel lost his career trying to persuade the CEO to self-report a trading scandal by invoking moral duty, when the same case could have been made in terms of the man's own position. Being right is not the same as being persuasive, and the failure mode is expensive. [Source: "Talk Three: A Lesson on Elementary, Worldly Wisdom, Revisited (Part 2 — The Stanford Q&A)"]

Commit rarely and heavily rather than often and thinly. Munger's image is a lifetime punch card with twenty holes in it, and his standard is the rare professional handicapper who bets only when the crowd has genuinely mispriced something and stays out otherwise. He is blunt that most people dilute their few good ideas by chasing far more opportunities than they can actually evaluate. [Source: "Talk Two: A Lesson on Elementary, Worldly Wisdom (Part 2 — From Models to Stock Picking)"]

Don't

Don't be unreliable about anything you have engaged to do. Munger's claim is unusually strong: failing to reliably do what you said will more than counterbalance the combined effect of all your virtues, however great. His counterexample is a severely dyslexic college roommate who became a successful CEO on reliability alone. [Source: "Talk One: Harvard School Commencement Speech (1986)"]

Don't reason from one model. The man with only a hammer sees nails everywhere, and Munger's diagnosis of academic economics is that it overweighs whatever can be precisely counted and undercounts what matters but resists measurement. His hero here is Thomas Hunt Morgan, who banned calculators from his Caltech lab to force people toward the bigger questions instead of the small measurable ones. [Source: "Talk Nine: Academic Economics (Part 1 — Five Defects and Two Microeconomic Puzzles)"]

Don't stop the analysis at the first-order effect. The original Medicare cost projections were low by more than 1,000% because forecasters extrapolated past costs without asking how the new incentives would change behavior. Munger's formulation is that the consequences have consequences, and those have consequences too. [Source: "Talk Nine: Academic Economics (Part 2 — Psychology, Second-Order Effects, and Virtue)"]

Don't let an ideology do your thinking for you. Munger names heavy political or religious commitment as one of the most extreme distorters of cognition, cites Buffett deliberately building a life far away from his own father's fierce ideology, and says outright he would not hire or retain people of intense passionate ideology in either direction, because multidisciplinary synthesis requires an objectivity the commitment forecloses. [Source: "Talk Three: A Lesson on Elementary, Worldly Wisdom, Revisited (Part 1 — The Lecture)"]

Don't spend anything on envy. Munger's argument is not moral, it is arithmetic: every other vice at least pays something, and this one is all pain and no fun. He quotes Buffett's line that the world is driven by envy rather than greed, and notes psychology textbooks barely index it, probably because naming a position as envy-driven feels too rude to say out loud. [Source: "Mungerisms: Charlie Unscripted"]

Don't put more money into a commitment to avoid admitting the first amount was wrong. Munger analyzes Berkshire's own USAir position this way, and names the mechanism: denial plus the reaction to nearly-losing something, which together produce the "just a little more" instinct that he says is a common way people go broke. [Source: "Talk Three: A Lesson on Elementary, Worldly Wisdom, Revisited (Part 2 — The Stanford Q&A)"]

Don't attend the auction. Loss aversion and social proof compound at an open-outcry auction into overbidding that neither force would produce alone, which is why Buffett's rule is simply not to go. Munger's own version of the same tendency cost him $5.4 million by declining to buy more deeply underpriced Belridge Oil, because it would have meant borrowing, even though he had ample means. [Source: "Talk Eleven: The Psychology of Human Misjudgment (Part 2 — Tendencies Ten Through Twenty-Five)"]

Don't accept the small, gradual version of something you would refuse all at once. Munger's fictional Quant Tech inflates its earnings by design in moderate annual increments, on the explicit reasoning that a small yearly increment escapes the notice a single large one would attract, and that the auditors who blessed last year's statements will find it awkward to object this year. Twenty years of that produced a 90% collapse. [Source: "Talk Eight: The Great Financial Scandal of 2003"]

Hold

The safest way to get a thing is to deserve it. Munger says he had this idea very early and treats it as more reliable than any strategy for acquisition, because what it actually earns is the deserved trust of the people you deal with, which he rates a greater pleasure than the money. Holding it changes what you optimize for in a negotiation you could win. Refusing it leaves you playing a game where every counterparty is a resource, and where your own reputation has to be managed rather than simply told the truth. [Source: "Talk Ten: USC Gould School of Law Commencement Address (2007)"]

A web of deserved trust outperforms procedure. Munger and Buffett ran a decades-long partnership on a handshake with no written contract, and Munger left the law firm bearing his name without taking his share of the capital, directing it instead to the estate of a young partner who had died leaving a wife and children. His warning to new lawyers is that you may have to sell procedure but you do not always have to buy it, and that a forty-seven page marriage contract is evidence of something already wrong. Holding this changes who you agree to work with at all, since it makes character a gating condition rather than a nice-to-have. Refusing it means every relationship costs what the paperwork costs. [Source: "A Portrait of Charles T. Munger"]

Old age is paid for decades in advance. Munger, writing at 82 on Cicero, holds that the mind's pleasures do not decline the way the body's do, so a life spent acquiring useful knowledge leaves old age something to draw on, and Franklin is his proof: a phonetic alphabet at 62, the Declaration at 70, the Constitutional Convention at 81. He is careful to add that longevity now brings real impairment for many, and that a cheerful adjustment to a reduced role is the sane response rather than guilt at not matching an unusually vigorous example. What holding it changes is what you do with the next ordinary decade. Refusing it treats the current decade as the one that counts and quietly writes off the last third of a life. [Source: "Praising Old Age"]

The one line

Most bad outcomes are produced by a short list of predictable failures, so work the problem backward, count the incentives, and refuse yourself an opinion you cannot argue against.