SECTION 01 · 530 GUIDES
Fundamental ideas, mental models, and frameworks for clear thinking. From first principles to systems theory, the concepts that shape how we understand the world.
WHAT YOU'LL FIND
Concepts are the building blocks of clear thinking. This collection draws on psychology, economics, philosophy, cognitive science and decision theory, the goal is not memorisation but internalisation.
CONCEPT OF THE WEEK · 22 SEP 2026
When a measure becomes a target, it ceases to be a good measure. Named for economist Charles Goodhart's 1975 observation about monetary policy, generalised by Marilyn Strathern in 1997, it explains why call-centre handle times fall while customers get angrier, and why test scores rise while learning doesn't.
THREE GUIDES THAT APPLY IT
6 PAIRS · ONE-SENTENCE DISTINCTIONS
PSYCHOLOGY & BEHAVIOUR · 266 GUIDES
Every cognitive bias and fallacy we've covered, grouped by the kind of thinking it distorts. Each entry links to the guide that defines it, with sources.
GUIDES · NEWEST FIRST
principles-laws9 Nov 2026The Streisand Effect: How Suppression BackfiresThe Streisand effect explained: why trying to hide or censor information amplifies it, through reactance, information cascades, and the famous 2003 case.14 MIN READ
principles-laws2 Nov 2026Parkinson's Law: Work Expands to Fill the Time AvailableC. Northcote Parkinson published the line as Civil Service satire in The Economist in November 1955, backed by arithmetic: between 1914 and 1928 Royal Navy capital ships fell by roughly two thirds while Admiralty officials rose by nearly 80 percent. His mechanism was that officials multiply subordinates and then make work for each other. The article is candid that controlled experiments isolating the law are scarce, the support being convergent.14 MIN READ
principles-laws26 Oct 2026Metcalfe's Law: Why Networks Get More Valuable as They GrowBriscoe, Odlyzko and Tilly argued in IEEE Spectrum in 2006 that treating every possible connection as equally valuable vastly overcounts, and proposed that value grows in proportion to n log n instead, a curve still faster than linear but far gentler than the square. Metcalfe answered in 2013 with a decade of Facebook data he said fit the quadratic form. The article traces the law from an Ethernet sales slide to the exponent that remains genuinely unsettled.14 MIN READ
principles-laws19 Oct 2026Little's Law ExplainedFix any two of work in progress, throughput and time in system, and the third is settled. A team finishing five tasks a week with ten tasks open averages two weeks per task, and nothing about how the team works internally is needed to know that. The practical lesson is that cutting work in progress shortens delivery immediately, while raising throughput means buying capacity. Also covers the two conditions the law needs: a stable system and long run averages.6 MIN READ
principles-laws12 Oct 2026The Law of Diminishing Returns ExplainedBiologists derived the same curve independently: Charnov's 1976 marginal value theorem predicts a forager leaves a patch when its intake rate falls to the average available across the environment, and Cowie's 1977 study of great tits found they go close to that optimum, lingering when travel between patches is costly. The article separates increasing, diminishing and negative returns, and argues the way past the curve is to unfix the binding constraint.13 MIN READ
metrics-measurement-evaluation28 Sep 2026Leading vs Lagging Indicators ExplainedA leading indicator is a bet: it encodes your theory about what drives the outcome, which is why it has to be both predictive and something you can actually influence. This guide contrasts the two kinds of measurement across timing, actionability and certainty, then covers the three ways leading indicators fail: they get gamed when the number becomes the target, the assumed link to the outcome quietly breaks, and too many of them dilute focus.7 MIN READ
mental-models21 Sep 2026The Map Is Not the Territory: Why Models MisleadBorges imagined cartographers so ambitious they drew a map the size of the empire, coinciding with it point for point, and later generations left it to rot in the desert. The fable's point is that a map's omissions are its function, not its defect, which is why adding detail rarely repairs a model that misled you. The article sorts the real failure modes into omission, bias and staleness, each with a guarding question.16 MIN READ
mental-models14 Sep 2026Second-Order Thinking: Considering the Consequences of ConsequencesFrench colonial Hanoi paid for rat tails as proof, so people caught rats, cut off the tails and released them to breed, and officials began noticing tailless rats in the streets. The historian Michael Vann documented it from the colonial archives, which makes it the well-attested case where the Delhi cobra story Horst Siebert named in 2001 is not. Covers Bastiat and Merton on unanticipated consequences, the premortem, and where to stop the chain.14 MIN READ
mental-models7 Sep 2026Occam's Razor: When the Simplest Explanation WinsWilliam of Ockham never wrote the Latin sentence attached to his name; later writers coined it, and Aristotle had argued in the Posterior Analytics for the demonstration resting on fewer postulates. The razor earns its keep probabilistically, since each added assumption is another chance to be wrong, and statistically, since fewer parameters resist overfitting. Medicine keeps a counterweight: Hickam's dictum, that a patient can have as many diseases as they please.15 MIN READ
mental-models31 Aug 2026Inversion: Solving Problems by Thinking BackwardGary Klein's premortem has a team assume a project has already failed and write down why. He cites 1989 research by Mitchell, Russo and Pennington finding that prospective hindsight, imagining an event has already happened, raises the ability to correctly identify reasons for future outcomes by 30 percent. The grammatical shift from might fail to did fail is what unlocks it, and it also makes pessimism the assigned task, so doubts stop feeling disloyal.14 MIN READ
frameworks-models30 Aug 2026Ship of Theseus Explained: The Paradox of Identity Over TimeThat gap between a complete physical description and an unsettled identity question is the actual content of the puzzle, because it shows sameness over time is not read off the facts but needs a criterion argued for separately. Hobbes sharpened it in De Corpore in 1655 by reassembling the discarded planks into a second ship, splitting continuity from material composition. Covers the mereological, four-dimensionalist, sortal and conventionalist answers.8 MIN READ
learning-science-knowledge13 Aug 2026Interleaving Practice ExplainedBlocked practice drills A A A A then B B B B and feels like steady progress. Interleaving shuffles the types so every problem forces you to work out which approach applies, the part blocked practice quietly skips. Learners consistently rate blocking as more effective even when interleaving produces better results on later tests, so the discomfort is the signal rather than the problem. The guide covers when to block first and how spacing compounds the effect.6 MIN READ
frameworks-models27 Jul 2026Theory of Constraints ExplainedEliyahu Goldratt's theory of constraints starts from one observation: any system pursuing a goal has at least one limit that caps what the whole can achieve, so improving anything else changes nothing. The five focusing steps follow: identify the constraint, exploit what you already have, subordinate everything else to its pace, elevate its capacity only if that is still not enough, then repeat. The hard step is the third: letting other resources sit idle.6 MIN READ
frameworks-models24 Jul 2026The Overton Window Framework: How Ideas Become Politically PossibleJoseph Lehman of the Mackinac Center put it bluntly: lawmakers are not in the business of shifting the Overton window, they are in the business of detecting where it is and moving into line with it. That is why activists, journalists and think tanks hold leverage elected officials lack. The window moves through anchoring with extremes, repetition, norm change and crisis, and none of those require persuading anyone that an idea is correct.14 MIN READ
decision-making18 Jul 2026The Sunk Cost Fallacy: Why We Throw Good Money After BadSweis and colleagues reported in Science in 2018 that mice, rats and humans all showed sensitivity to sunk costs in a foraging task, and that the bias appeared only after the decision to pursue a reward had been made, not in the deliberation before it. Arkes and Blumer's 1985 theatre study showed the same pull: season ticket holders who paid full price attended more plays than discounted buyers seeing identical performances.14 MIN READ
decision-making15 Jul 2026The Premortem: Imagining Failure Before It HappensMitchell, Russo and Pennington reported in 1989 that people told an outcome had already happened generated roughly 30 percent more possible reasons for it than people asked what might happen. The study measured the quantity and richness of reasons, not their accuracy. Gary Klein built the meeting around that effect, and its second engine is social: once failure is assumed, the sharpest critic wins the exercise instead of looking disloyal.15 MIN READ
decision-making12 Jul 2026The Availability Heuristic: Why Vivid Beats LikelyAsked whether K appears more often as the first or third letter of English words, most people in Tversky and Kahneman's 1973 study said first; in typical English text it is roughly twice as common in third position. Recall ease, not frequency, drove the answer. Schwarz and colleagues sharpened the point in 1991: people asked for six examples of their own assertiveness rated themselves as more assertive than people asked for twelve.15 MIN READ
decision-making9 Jul 2026The Anchoring Effect: How the First Number Shapes the RestKahneman describes an experiment where people estimated a redwood's height after being given either 1,200 feet or 180 feet, anchors 1,020 feet apart. Their mean estimates came out at 844 and 282 feet, a gap of 562 feet, so roughly 55 percent of the distance between the anchors passed into the answers, which Kahneman calls typical. The defenses with a track record: consider the opposite, and commit to your own number first.15 MIN READ
decision-making6 Jul 2026Satisficing vs Maximizing ExplainedA satisficer sets a threshold before searching and takes the first option that clears it; a maximizer keeps looking until convinced nothing better exists. The difference is the stopping rule, not the height of the standards, and because the full option set is usually unbounded a maximizer has no natural end point. Use five factors to triage any decision: stakes, reversibility, how often it recurs, how many options there are, and the real value gap between good and best.6 MIN READ
decision-making3 Jul 2026Hick's Law: Why More Choices Slow You DownHick in 1952 and Hyman in 1953 found that reaction time tracks the information in a choice, measured in Shannon's bits, so every doubling of options adds the same fixed increment rather than doubling the time. That makes the first few options the costly ones and grouping the real lever. The Iyengar and Lepper jam study, where about 30 percent bought from six jams versus roughly 3 percent from twenty-four, has not always replicated.15 MIN READ
principles-laws22 Jun 2026Goodhart's Law: When a Measure Becomes a TargetCharles Goodhart noticed in 1975 that the Bank of England's stable money-supply relationships broke down the moment it started targeting them, because targeting changed the system that produced the correlation. Marilyn Strathern gave it the phrasing everyone quotes in 1997. Manheim and Garrabrant later split it into four mechanisms, and Choi, Hecht and Tayler found surrogation deepens exactly when pay is tied to the proxy.10 MIN READ
learning-science-knowledge22 Jun 2026The Spacing Effect: Why Cramming Fails and Spaced Practice WinsEbbinghaus found in the 1880s that spreading repetitions across days needed fewer total repetitions than massing them, so spacing is not just better, it is cheaper. The 2006 Cepeda and Pashler meta-analysis, pooling 254 studies and more than 14,000 participants, confirmed the advantage is large and gave a practical rule: set the gap between sessions at roughly 10 to 30 percent of how long you need to remember. Covers Bjork's desirable difficulties and the fluency illusion.12 MIN READ
Decision Making5 Jun 2026Cognitive Bias Evidence Map: Robust vs. Failed-to-Replicate BiasesNot every famous bias is equally well supported. This map rates ten effects by the strength of the record behind them, separating the ones tested at large scale and failed from the ones never tested at large scale at all, like the planning fallacy. The sharpest case is ego depletion: a pre-registered replication across 23 labs with 2,141 participants returned d = 0.04, a confidence interval straddling zero.14 MIN READ
Learning & Knowledge5 Jun 2026Learning Science Evidence Map: What the Research Actually ShowsAn original measurement sits alongside the research ratings: of the usable top search results for learning styles, only 2 of 7 mentioned that the meshing hypothesis has repeatedly failed controlled tests, the lowest disclosure rate measured, while every result checked for highlighting and re-reading noted their low utility. Eight techniques rated on evidence strength and effect type, each tied to a named peer-reviewed source.15 MIN READ
Decision Making1 Jun 2026Better Decisions Under Uncertainty: A Practical Guide and ToolkitA sound decision can still produce a bad outcome, and luck can hide weak reasoning, so the only thing you can improve is the process. The sharpest figure here is a calibration one: when people say they are 90 percent confident, studies consistently find them right roughly 70 percent of the time. From there it works through availability, sunk cost and anchoring, Gary Klein's premortem protocol, Bezos's two-way door test, and a ten-item checklist.12 MIN READ
ethics-values-society-culture1 May 2026Utilitarianism Explained: Ethics of Pleasure and PainWhen an agency sets vehicle safety standards or allocates a health budget it is running that calculation whether it uses the vocabulary or not. Bentham's equal weighting of every person's interests also made him radical for his era, since it led him to argue for women's equal rights and the decriminalization of homosexuality. One correction the page makes: Foot introduced the trolley problem in 1967 to probe the doctrine of double effect.14 MIN READ
data-analytics-insights28 Apr 2026Understanding Measurement Bias: How It Skews ResultsMore data cannot fix a biased sample. The Literary Digest mailed 10 million questionnaires, drew its list from car registrations, phone directories and subscribers, and predicted Landon at 57 percent while Roosevelt took 61. Gallup got it right with about 50,000 using quota sampling. Wald saw the same structure in bomber damage: the undamaged areas on returning planes were where hits proved fatal.34 MIN READ
learning-science-knowledge25 Apr 2026The Testing Effect: Quizzing Yourself for Better RetentionWhy the study method that feels effective is not. Roediger and Karpicke's 2006 benchmark compared repeated study against repeated retrieval with equal time on the material, and the retrieval group performed far better on a delayed test one week on while rating their learning lower. Their 2008 follow-up in Science found the same pattern even though students predicted repeated study would serve them as well or better.22 MIN READ
psychology15 Apr 2026Emotional Labor: Why Easy Days Feel ExhaustingArlie Hochschild named this work in 1983: emotional labor, the suppressing, performing and managing of feeling a job requires. She separated surface acting, faking an emotion you do not have, which research consistently finds more depleting, from deep acting, genuinely shifting your state to match the moment. Knowledge workers do far more of it than they count, in meeting faces, softened feedback and stakeholder maintenance.10 MIN READ
psychology15 Apr 2026Why Procrastination Affects Capable IndividualsWhy procrastination resists productivity systems. Steel defined it in 2007 as voluntary delay despite expecting to be worse off, which makes it an emotion regulation problem rather than a time management failure. Wohl, Pychyl and Bennett found in 2010 that students who forgave themselves for procrastinating on one exam procrastinated less on the next, while self-criticism predicted the opposite.15 MIN READ
psychology15 Apr 2026Rejection Sensitive Dysphoria in the WorkplaceWhat RSD is and why it is hard to interrupt in the moment. The response arrives within seconds of a perceived rejection, often before the triggering content has been processed, which is what makes cognitive intervention difficult. Many people describe it as physical pain rather than sadness. Dodson popularized the term from ADHD clinical work, though Downey's rejection sensitivity research treats it as a dimension rather than a category.14 MIN READ
psychology15 Apr 2026High-Functioning Anxiety: Signs and StrategiesWhat the pattern is and which interventions have real support. High functioning anxiety is not a formal diagnosis but describes a clinically significant pattern where external performance stays high while internal cost compounds. The Yerkes-Dodson law predicts performance rises with arousal to a point then declines, and this pattern typically holds arousal past that point. Sleep and aerobic exercise are the highest-leverage starting points.15 MIN READ
mental-models15 Apr 2026Ockham's, Hanlon's, and Chesterton's Razors ExplainedHanlon's Razor is not folk wisdom, it is a correction for a documented bias: the fundamental attribution error that Lee Ross named in 1977, our habit of blaming character when circumstance explains the behavior. The article pairs each heuristic with its real failure mode: Ockham oversimplifies genuinely complex domains, Hanlon excuses repeated harm that benefits the actor, and Chesterton's Fence can be used in bad faith to freeze any reform indefinitely.11 MIN READ
mental-models15 Apr 2026When to Use First Principles vs Systems ThinkingTwo lenses for different classes of problem. First principles strips a question to verifiable constraints and rebuilds, as when Tesla compared roughly 600 dollars per kilowatt-hour battery packs against about 80 dollars of raw materials. Systems thinking maps stocks, flows, feedback and delays instead, and Meadows ranked 12 leverage points from the weakest, changing numbers, to the strongest, changing the paradigm the system rests on.11 MIN READ
psychology14 Apr 2026Dark Psychology: Recognizing and Countering ManipulationFreyd's group found that people briefed on DARVO before judging a case were significantly less swayed by it, so naming the pattern is itself a defense. DARVO runs deny, attack, then reverse victim and offender, which raises the cost of pursuing a complaint above the cost of dropping it. Against future faking the counter is evidence-based: judge the last thirty days of observable behavior, not the next thirty days of stated intentions.16 MIN READ
mental-models14 Apr 2026First Principles Thinking: Elon Musk's Approach ExplainedWhere the method came from and when it fails. Musk broke a battery pack into cobalt, nickel, aluminum, carbon and steel, priced each at commodity level, and found roughly 80 dollars per kilowatt-hour against an industry consensus of 600. The article is equally clear on the limits: reasoning up from wrong foundations produces confident errors, and in irreducibly complex domains statistical approaches beat it.16 MIN READ
learning-science-knowledge14 Apr 2026Feynman Technique: Accelerate Learning with EvidenceWhat the four steps are and why the method works alone. Nestojko and colleagues found in 2014 that undergraduates told they would teach a passage outperformed those told they would be tested on it, on both recall and conceptual questions, although neither group taught anyone. The preparation frame produces the gain, which is why explaining a concept in plain language to no one still exposes the gaps.19 MIN READ
communication14 Apr 2026How to Say No Without Feeling Guilty: Practical Assertiveness ScriptsEisenberger, Lieberman and Williams put participants in an MRI scanner for the 2003 Cyberball study and found that exclusion from a simple ball-tossing game engaged the dorsal anterior cingulate cortex and anterior insula, regions also involved in physical pain. The study did not examine assertiveness, so it is one possible explanation rather than proof, but it suggests why anticipated rejection makes refusal hard. The article marks what is research and what is script.13 MIN READ
psychology13 Apr 2026Dopamine Detox: Analyzing Its EffectivenessWhat the trend gets wrong and what actually produces the effect people report. Dopamine encodes prediction errors and motivation rather than pleasure, which Berridge separated into wanting and liking, so it cannot be detoxed the way a drug clears from plasma. Receptor changes, where they occur, reverse over months rather than a weekend. What a break really does is interrupt variable-ratio reinforcement.17 MIN READ
mental-models13 Apr 2026Cognitive Biases Everyone Falls ForWillpower does not beat bias; changed procedure does. The entries pair a measured effect with a fix: Englich, Mussweiler and Strack found experienced German judges shifted sentences by about eight months on an explicitly random prosecutorial anchor, and Goldin and Rouse found that screening orchestra auditions behind a curtain explained around 25 percent of the rise in women hired by major US orchestras. Pre-mortems, blind rubrics and base rates do the work.13 MIN READ
learning-science-knowledge13 Apr 2026Nine Methods to Learn Anything Faster and EffectivelyWhich study methods the evidence supports and which only feel productive. Dunlosky and colleagues evaluated ten common techniques in 2013 and rated only two as high utility, practice testing and distributed practice, while highlighting, rereading, summarizing, imagery for text and the keyword mnemonic all landed in the low-utility category. Karpicke and Roediger measured 61 percent against 40 percent one week on.13 MIN READ
communication13 Apr 2026Overthinking and Rumination: Why Thought Loops Persist and What Can HelpWegner's white bear studies showed that suppressing a thought can produce rebound, especially while you keep checking whether it has come back, which is why pushing harder rarely ends a loop. The alternatives rest on different rationales: Edward Watkins on shifting from abstract to concrete thinking, Steven Hayes on defusion, Borkovec's scheduled worry window, behavioral activation. The piece is careful about what each does and does not establish.15 MIN READ
learning-science-knowledge8 Apr 2026Spaced Repetition: The Most Efficient Way to Learn AnythingHow spaced repetition works and what spacing the research supports. Cepeda and colleagues analyzed 254 studies in 2006 and found the optimal gap is approximately 10 to 20 percent of the desired retention interval, so remembering something for a year means first reviewing it weeks later, not days. Ebbinghaus measured the forgetting curve in 1885, and Wozniak's SM-2 algorithm turned the principle into software in 1987.15 MIN READ
communication14 Mar 2026How to Explain Complex Ideas Clearly to Different AudiencesThe test for what belongs in an explanation is removal: take one component out, and if the audience would now draw a materially false conclusion, it belongs in the core. Saying antibiotics kill infections is memorable and falsely broad, while saying they treat susceptible bacterial infections but not viral ones is barely longer and keeps the constraint that makes it actionable. Covers intended action, dependency order, and where analogies break.10 MIN READ
communication13 Mar 2026Feedback Loops in Communication ExplainedSignal feedback confirms a message arrived; meaning feedback confirms it was understood as intended, and conflating the two is the error behind the costliest failures. NASA lost the Mars Climate Orbiter in 1999 because one engineering team used metric units and another imperial, at an estimated $327 million: the signal was received perfectly and the software ran as coded, while two teams used the same numbers to mean different things with no mechanism to catch it.20 MIN READ
communication13 Mar 2026The Curse of Knowledge Explained: Why Experts Misjudge What Beginners KnowKnowing something well distorts your estimate of what others know. The page separates the bias from the illusion of transparency, false consensus and tacit knowledge, and reviews the Newton, Nickerson, Camerer and Hinds evidence along with its limits. In Newton's 1990 tapping study, tappers predicted about half of listeners would name the tune; roughly 2.5 percent did. The fixes are procedural: list assumed prerequisites, test with real readers.12 MIN READ
communication12 Mar 2026Narrative Transportation: How Stories Influence Beliefs and BehaviorGreen and Brock's 2000 experiments found that the more transported readers were into a narrative, the more their beliefs aligned with what it implied, whether the text was labeled fact or fiction. Van Laer and colleagues' 2014 meta-analysis of 76 studies and 132 effects confirmed the association while identifying substantial moderators rather than one universal effect. Argument strength still matters inside a story, and the mechanisms remain proposals.10 MIN READ
communication12 Mar 2026The Ladder of Abstraction: Moving Between Examples and PrinciplesHayakawa built his ladder on a single cow named Bessie to show that words sit at measurable distances from the thing itself, and that no rung is the correct one to speak from. The two failures are opposites: climbing too high leaves claims no observation can check, while staying too low buries the point in particulars nobody can generalize from. Treated here as a linguistic and pedagogical model rather than an experimental finding.12 MIN READ
communication12 Mar 2026Framing Effects: How Presentation Changes MeaningIn 1981 Tversky and Kahneman described the same disease program two ways. Seventy-two percent of subjects chose the option saving 200 for certain over an equivalent gamble, while seventy-eight percent chose the gamble over 400 dying for certain. Identical math, opposite risk appetite. Also here: Levin and Gaeth on 95 percent fat-free beef, and Johnson and Goldstein's opt-out donation rates of 85 to 99 percent against 4 to 28 percent for opt-in.25 MIN READ
communication10 Mar 2026Why Miscommunication Happens Even When Words Are ClearKeysar and Henly had speakers convey ambiguous statements and rate their own success, then measured what listeners took away: speakers consistently overestimated how well they had got through, though the result belongs to that task and should not be stretched across every medium. The page separates six causes, and distinguishes read-back from teach-back and closed-loop confirmation.12 MIN READ
communication10 Mar 2026Understanding Communication Failures: The FrameworkWhy most breakdowns happen at encoding, not transmission. The Mars Climate Orbiter was destroyed in 1999 because trajectory data encoded in pound-force seconds was decoded as newton-seconds: the signal arrived intact and still carried different meaning at each end. Semmelweis showed the same pattern in 1847, since his mortality evidence was never disputed but could not be decoded without germ theory.19 MIN READ
communication9 Mar 2026Signal vs. Noise in Communication: How to Prioritize What Your Audience NeedsSix types of noise are separated out because each needs a different remedy: channel, semantic, structural, relevance, relational, and cognitive overload. The article is careful about its own metaphor, noting that verbose prose is not noise in Shannon's sense, and that content can be signal for a reviewer and noise for an executive. It also defends deliberate repetition: aviation read-backs and SBAR handoffs add it on purpose, because reliability outranks brevity.13 MIN READ
communication9 Mar 2026Five Principles for Clear CommunicationTreating clarity as a property of the message makes it unanswerable, because the same sentence can be exact for one reader and useless for another. Orwell's six rules from 1946 remain influential but have not been validated as a unified empirical framework, and the five principles here are presented as an editorial model rather than an established research taxonomy. Includes Sweller and Cooper's 1985 worked examples with the limit the authors themselves stated.12 MIN READ
decision-making8 Mar 2026Common Decision Traps and How to Avoid ThemPronin, Lin and Ross showed in 2002 that people spot biases readily in others and hardly at all in themselves, and that knowing you are susceptible does not predict actually resisting it. That finding is the whole argument for process over vigilance, since vigilance is exactly what fails under pressure. Covers confirmation bias, sunk cost, anchoring, availability, groupthink and status quo bias, how they compound, and the structural countermeasures that work.24 MIN READ
decision-making7 Mar 2026An Introduction to Second-Order Thinking and Its ExamplesSecond-order thinking traces the consequences of consequences, not just the immediate result. Australia imported cane toads to eat crop beetles and got billions of poisonous toads instead. Howard Marks on first-level versus second-level thinking, the cobra effect, tragedy of the commons, moral hazard, Goodhart's law and induced demand, with techniques for practicing it.31 MIN READ
decision-making6 Mar 2026Decision Frameworks Used by High Performers to Think ClearlyPaul Meehl's 1954 review compared intuitive clinical judgment against rule-based prediction across 20 studies. Actuarial methods matched or beat the experts in 17 of them, and Grove and colleagues' 2000 meta-analysis of 136 studies found clinical judgment won in only 6 percent. This walks through eight frameworks: reversibility, regret minimization, 10/10/10, pre-mortem, decision journals, expected value, inversion and calibrated probability.25 MIN READ
decision-making6 Mar 2026Decision Journaling for Better Judgment Over TimeLevav and colleagues tracked 1,067 car purchases at German dealerships and found buyers deliberated carefully over early options then increasingly accepted defaults by the end of the sequence. Linder and colleagues found inappropriate antibiotic prescribing rose 26 percent in afternoon appointments across roughly 21,900 visits. This page is careful about the contested science: the famous parole board finding is reported as an association, not a measured depletion state.26 MIN READ
decision-making5 Mar 2026Mental Models That Make Every Decision SharperA working set of decision models, each with the conditions under which it fits: second-order thinking, inversion, first principles, opportunity cost, margin of safety and probabilistic thinking. Gary Klein's Recognition-Primed Decision work found experienced experts rarely compare options at all, and Chase and Simon's 1973 chess studies showed masters rebuilt positions from 5-second glances by chunking patterns, not better memory.30 MIN READ
decision-making5 Mar 2026Why Most Decisions Fail Due to Poor FramingFraming decides which decision you are making; analysis only decides how. Paul Nutt spent 20 years studying roughly 400 strategic decisions and reported that about half ended in failure, the most common cause being poor problem formulation rather than poor analysis: decision-makers evaluated options before defining the problem. The page catalogs five framing traps, reframing techniques, and how Kodak framed digital photography as a threat to contain.23 MIN READ
decision-making5 Mar 2026What Is Decision Making Under Uncertainty?Frank Knight separated risk, where probabilities are known, from uncertainty, where they cannot be meaningfully assigned, and most big decisions sit on the second side. This guide maps four levels of knowability, then covers satisficing, reversible bets, scenario planning and Bayesian updating. Tetlock's Expert Political Judgment study, spanning about 20 years and more than 28,000 forecasts from 284 experts, found most specialists predicted little better than chance.32 MIN READSimplified representations of how things work that help you predict outcomes, make decisions, and solve problems. Charlie Munger's “latticework” approach: learn fundamentals from many disciplines and you get a toolkit for better thinking across all of them.
Breaking a problem down to its foundational truths and reasoning up from there, instead of reasoning by analogy. It originates with Aristotle; Musk popularised it in business.
Seeing interconnections, feedback loops, delays, and leverage points rather than isolated events. Most real problems live inside systems where changing one part affects the whole.
Considering the consequences of consequences, asking “and then what?”, to reveal effects that first-order thinkers miss.
Reasoning with likelihoods and distributions rather than certainties. It acknowledges uncertainty and is essential for risk assessment, forecasting, and strategy.
Understand the core principle, recognise the patterns where it applies, practise deliberately across contexts, seek feedback, and build connections between related models.
Approach problems backward: ask how you would guarantee failure, then avoid those failure modes. Especially useful for risk management and strategy.
THE WEEKLY DISPATCH
One new guide in your inbox each week, plus a note on which older guides were revised. Unsubscribe anytime.
We use cookies to enhance your experience, analyze site traffic, and provide personalized content. By clicking 'Accept All', you consent to our use of cookies. View our cookie policy
Choose which cookies you want to allow. Essential cookies are required for the website to function properly.
These cookies are necessary for the website to function and cannot be disabled. They are usually set in response to actions you take, such as setting privacy preferences or logging in.
These cookies help us understand how visitors interact with our website, allowing us to improve our services and user experience.
These cookies are used by Google AdSense to display personalized ads. They help us provide free services by generating revenue from advertisements.