I have kept a list like this for years. It started as somewhere to put things I read and did not want to lose, and it turned into the set of frames I reach for when a problem does not have an obvious shape. Most of them tie back to business in some formal way. A few are on the list because I find them interesting rather than because they are useful. What they have in common is that each one has changed a decision I made at some point, or changed how I look at the world.

I add to the list as I go, so it is never quite finished.

Theories

Bayes' Theorem

Bayes' theorem is about updating a probability as new information arrives. The version I use day to day is simpler than the math. A headline number describes a population, and I am almost never the population. If half a percent of people die of something, my odds depend on my age, my health, and my habits, and the honest answer could be far better or far worse than the headline. Businesses make the same mistake with market data. Industry churn of fifteen percent tells you very little about your churn until you find the subgroup you actually sit in.

Dunbar's Number

Dunbar's number says a person can hold roughly 150 meaningful relationships at once. Mine is closer to fifty, because the relationships outside the office come out of the same budget. Family, friends, the people I actually want to keep up with.

The number matters because a business below the number runs on relationships and a business above the number runs on structure. The crossing is where culture changes, and the change costs more effort than most people plan for. Cross the line without building the structure and the business keeps trying to coordinate through conversations that no longer reach everybody.

The Eisenhower Method

The Eisenhower method sorts tasks by importance and urgency into four quadrants. Important and urgent gets done. Important and not urgent is where the long-term value sits. Not important and urgent is what a manager should be delegating. Not important and not urgent should not exist. Most managers I know live in the urgent column and call it work. The quadrant that compounds is important and not urgent, which covers planning, hiring ahead of need, and thinking about the business rather than in it. That quadrant has no deadline attached, which is why the quadrant gets skipped.

Prospect Theory

Prospect theory, from Kahneman and Tversky, found that people feel a loss about twice as strongly as they feel an equivalent gain. The same pair gave us loss aversion, which is the shorter name for the same finding. Most of the changes I have proposed inside a business have been positive on paper and still met resistance, because the people affected were weighing what they stood to lose at roughly double the weight of what they stood to gain. If you want a change to land, work out what each person is giving up and deal with that first.

Principal Agent Theory

The principal agent problem is that the people you hire have their own priorities, and their priorities are not yours. The standard fix is to align interests through compensation, and the standard fix holds up reasonably well while things are going fine. I watch for the problem hardest when something is going wrong, because that is when it shows. The moment a project starts to fail, most people insulate themselves, protect their position, and push the blame somewhere else. The person who takes the hit on behalf of the group is rare. Charities and the military seem to produce more of them than business does, which suggests compensation is not the only lever.

Two-Factor Theory

Herzberg's two-factor theory says the things that demotivate people and the things that motivate them are two different lists. Take time off, for instance. A company that does not offer it will have demotivated staff. Offering time off does not make anyone more motivated. All the offer does is remove a reason to be unhappy. Pay works the same way past a certain point. I have watched businesses try to buy motivation with money and get nothing back except a higher cost base, because they were topping up the hygiene column and wondering why the motivation column stayed empty.

Euclid's Elements

Euclid built all of geometry on five postulates. A straight line between any two points, a segment extended indefinitely, a circle at any center and radius, all right angles equal, and the fifth one about parallel lines. The fifth looked so obviously true that mathematicians spent two thousand years trying to derive it from the other four. The fifth postulate is optional. Replace the fifth and you get hyperbolic geometry, which is internally consistent, and modify the system further and you get spherical geometry, which is what curved space actually needs. I do not force this one into a business lesson. I keep it because the assumption that looked least worth questioning turned out to be the one doing the work.

Braess's Paradox

Braess's paradox is that adding capacity to a network can make the whole network slower. It turned up in road systems, where opening a new road moved traffic to an equilibrium that was worse for the drivers overall, and it has been reproduced physically with springs and string. I think about the paradox whenever somebody proposes adding a step or a person to a process. On a factory floor an extra station can starve the one behind it and back up the one in front. Adding people to a late project is the version most of us have lived through. A system finds an equilibrium, and an equilibrium is not required to be good.

Human Conformity

Solomon Asch showed people a line and asked which of three other lines matched it. The answer was obvious. He seeded the room with actors who confidently gave the wrong answer, and about three quarters of participants went along with the room at least once, against what their own eyes were telling them. Boards and leadership teams run a version of the same experiment most weeks. If the first two people to speak agree, the third is pushing against the room rather than against the idea, which is a much harder thing to do. I often work to make sure I am the last one to speak.

The Zeigarnik Effect

The Zeigarnik effect is that an unfinished task stays in your head in a way a finished one does not. A waiter remembers an open order and forgets a paid one. In business the effect shows up as clutter. Half a dozen small projects that never quite closed will keep pulling at your attention, and the pulling is real work even on the days you do nothing about them. Running twelve projects at sixty percent costs more attention than running the same twelve to completion. I would rather close four things and start two.

Laws

Sturgeon's Revelation

Theodore Sturgeon was defending science fiction when he said that ninety percent of science fiction is crud, and then ninety percent of everything is crud. Most people know the line as Sturgeon's law. I use the idea in two directions. When somebody tells me they do not like a category, whether a technology, a market, or a management approach, I assume they have mostly met the bottom ninety percent. And when I am evaluating anything, I assume the first few things I see are from the bottom ninety percent and I keep looking.

Parkinson's Law

Parkinson's law is that work expands to fill the time allotted to it. Give a team six weeks for something that needs two and it will take six. The useful consequence is that a deadline is a design decision rather than an estimate. I set deadlines shorter than feels comfortable and treat the discomfort as the point. There is a limit past which short deadlines produce bad work and burnt-out people, and the limit moves with the team and the task.

Entropy

Entropy is the tendency of things to move toward disorder. Business is mostly a fight against entropy. You build something and it erodes underneath you, worn down by competitors, by staff turnover, by changes in the law, and by new technology that makes part of what you built irrelevant. I find the frame useful because it kills the idea that you are building toward completion. There is no finished state. Every business eventually fails or shuts down, and the work is holding the erosion off for as long as the thing is worth holding.

Bushnell's Law

The rule attributed to Nolan Bushnell at Atari is that the best games are easy to learn and difficult to master. I apply the rule to anything I want people to stick with. For customers that means a first step which takes almost no effort and a ceiling high enough that there is still something to shoot for a year in, which is what makes a program sticky rather than a one-off. Internally I use the same shape for targets. A team needs something achievable this week and something worth chasing for a year, and the two have to belong to the same game.

Moynihan's Law

Moynihan's law came out of human rights reporting. The number of complaints heard from a country runs inversely to the number of actual violations, because the places with the worst records produce the least noise. The better things get, the worse they seem. When complaints inside my own business went up I took it personally, and I was reading the signal backwards. People complain when they believe complaining is safe and might change something. The quiet organization is the one to worry about, because silence usually means people have decided that speaking up is not worth the cost.

Amara's Law

Amara's law is that we overestimate the effect of a technology in the short run and underestimate the effect in the long run. The shape is the hype cycle. A trigger, then a peak of inflated expectations, then a trough of disillusionment, then a slope of enlightenment as real adoption happens, then a plateau of productivity once the technology has become ordinary. The internet has been climbing out of its trough for twenty years and still keeps finding new ground, and it got a real push in 2020 when a great deal of work had to go remote. GPS is the cleaner example. I doubt anybody launching it was picturing a receiver in every pocket.

The Woolsey-Swanson Rule

Woolsey and Swanson wrote that people would rather live with a problem they cannot solve than accept a solution they cannot understand. I use the rule less as a management tactic and more as a way to read people. When a report or a board member or a customer keeps rejecting something that would obviously help them, the resistance is often about comprehension rather than merit. And I hold the rule against myself, because I have turned down plenty of things I did not understand and told myself I had a reason.

Principles

The Peter Principle

The Peter principle is that people get promoted until they reach a job they cannot do, and then they stop. Organizations therefore tend to fill up with people operating just past their level. The principle shows up most in management, because the reward for being an excellent engineer or an excellent salesperson is a job that has almost nothing to do with engineering or selling. The fix is a senior track that is not management, paid properly, which more companies talk about than build.

The Pygmalion Effect

The Pygmalion effect is that your expectation of someone tends to be self-fulfilling. Believe a person is capable and you hand them harder work, more information, and more room, and they usually grow into it. Believe they are not and you quietly withhold all three, then point at the result as proof you were right. I have been on both sides of the effect. The uncomfortable part is that you rarely notice you are running the negative version, because from the inside it feels like an accurate assessment.

In-Group Psychology

In-group psychology is the pull people feel toward a group they belong to. A leader is nearly always trying to build one, because people who feel part of something stay longer and work harder. The failure mode is that the in-group gets narrow. At one of my businesses the phrase going around was that Alastair likes this person, and the phrase explained why some people thought they could get away with things and why others thought there was an inner circle they were locked out of. I had not built that perception on purpose, which is why I did not see it. The group has to be the whole team or it works against you.

Tim Ferriss's Big Five

Tim Ferriss put five questions in The 4-Hour Body for spotting bad science, and I use them on far more than science. Is a relative change being used to convince, so that twenty percent better hides how small the real move is? Is it an observational study claiming cause and effect, the way fewer pirates correlate with more global warming? Does it rely on self-reporting, which people are poor at even when they are trying to be honest? Is the control group real or nominal? And do the funders have a stake in the answer? I run those five over vendor decks and consultant pitches more often than over papers.

Steelmanning

Steelmanning is building the strongest version of the argument against your own position, stronger than your opponent would build it. The straw man is the opposite, and the straw man is what most arguments run on. Doing the work properly is uncomfortable, because a real steel man sometimes wins. I ask for one in board discussions where the room has already agreed, since agreement arrived at that quickly usually means the idea has not been tested yet.

The Hedonic Treadmill

The hedonic treadmill is the tendency to return to a baseline level of happiness after almost anything, good or bad. A raise feels like a raise for about a month and then becomes the new normal. So do bonuses, and so does company spending on travel and perks, which is how a business ends up with a cost base that cannot be trimmed without the trim reading as a punishment. A meal eaten is soon forgotten. I apply the same test in my own life before buying anything, asking whether I want the nicer watch and the bigger house or whether I want the fortnight of feeling different that comes with them.

Span of Control

Span of control is the number of people one person can actually manage, and the number depends heavily on the work. Eight people doing near-identical repeatable jobs is manageable. Eight people running separate functions with their own problems is not, and the manager either goes shallow on all eight or deep on three and neglects the rest. When somebody tells me a team is underperforming, the span of the manager above them is one of the first things I look at, because it is a structural problem that usually gets treated as a performance problem.

Negative Confirmations

A negative confirmation is treating silence as agreement. If I do not hear back, I will assume we are proceeding. Silence is not agreement, it is an absence of information, which is the same problem an IT team has when a system cannot know whether a packet arrived unless something comes back. Negative confirmation gets used in law in places where I think it should not be. The practical effect on me is that I keep a running list of everything I have sent and not heard back on, and I chase all of it, because my own silence is not confirmation either.

Biases

The Dunning-Kruger Effect

Dunning-Kruger is usually drawn as a curve where confidence spikes early, crashes as you learn how much you do not know, and recovers as you approach competence. That curve is a later invention and does not appear in the research. What Kruger and Dunning actually found is less dramatic and more useful. Self-assessment barely moves with skill. People in the bottom quarter rated themselves around the sixtieth percentile when they were closer to the twelfth, and the strongest performers slightly underrated themselves. For hiring, the practical version is that confidence carries almost no information about competence in either direction.

The Endowment Effect

The endowment effect is that we value something more once we own it. The classic experiment gives people a mug and finds they want roughly two and a half times as much to sell it as they would have paid to buy it. That gap explains a lot of failed deals. In M&A a seller prices the business on what it took to build and a buyer prices it on what it will produce, and both sides think the other is being unreasonable. The same gap shows up in customer behavior, in what people will pay for something they already feel some claim on. It makes me wonder how anything changes hands at all.

The Hawthorne Effect

The Hawthorne effect is that people change their behavior when they know they are being watched, usually by working harder. The name comes from lighting studies at the Hawthorne Works plant in the 1920s, where productivity rose whatever they did to the lights, though a later reanalysis of the original data found the effect much weaker than the legend. The catch is that attention fades, and anything built on attention alone decays alongside it. Before I put oversight on a process I ask whether the task actually generates something, whether it can be iterated on, and whether there are real people available to do the watching.

Razors

Occam's Razor

Occam's razor says the simplest explanation is usually the right one. In a business the simplest explanation for a missed number is usually that somebody did not do the thing, rather than that a complex set of market forces conspired. I have sat through long and elegant explanations for results that came down to a person not making the calls.

Hanlon's Razor

Hanlon's razor is never to attribute to malice what is adequately explained by stupidity. I would soften stupidity to something closer to overload, bad information, or competing priorities, which covers more of the real cases. Very few people in a business are working against you on purpose. Most of the behavior that looks like sabotage is somebody overloaded and dealing with their own priorities first.

Rules of Thumb and Readings

The Beveridge Curve

The Beveridge curve plots unemployment against unfilled jobs, both as a share of the labor force. The relationship is not linear. As available jobs dry up the unemployment rate climbs much faster than you would expect. I use the curve mostly to work out roughly where we sit in the cycle, which matters most when I am looking at starting or buying something exposed to the broader economy. A restaurant, say. COVID pushed the whole curve outward, with plenty of open jobs and plenty of people out of work at once, because the jobs and the people did not match up. The curve has since mostly shifted back. Some businesses do better in a downturn than in a boom, because you can always sell potatoes.

Okun's Law

Okun's law says that for every one percentage point fall in unemployment, GDP rises around two to three percent. It is called a law and behaves more like a rule of thumb, because plenty of other things move at the same time. Workers can put in more hours, which lifts GDP without touching unemployment at all. Okun himself flagged that the relationship holds while unemployment sits in a normal band of roughly three to seven percent and gets unreliable outside it.

The Sahm Rule

The Sahm rule uses unemployment to call the start of a recession. When the three-month moving average of the unemployment rate sits half a percentage point or more above its low of the previous twelve months, a recession has usually already begun. It is a real-time indicator rather than a forecast. Knowing roughly where we are in the cycle changes how I price and where I deploy capital. It does not make me pessimistic. The worst year of the Great Depression was a fall in GDP of about thirteen percent, which means most of the economy carried on, and I have written about why that number is the one worth looking at.

P-Value

A p-value is the probability of seeing a result at least as extreme as the one you got if nothing real were going on. Below 0.05 is the usual bar for calling a result significant, and the larger the effect you are looking for, the smaller the sample you need to find it. The reason p-values matter in business is that most of the numbers people quote confidently come from samples far too small to mean anything. A close rate calculated off eleven deals. A first pass yield off one morning's production. Both get quoted in meetings as though they settle something.

Age Pyramids

An age pyramid shows the distribution of a population by age, and the shape tells you where a city, a state, or a country is heading. Consumer spending peaks in people's thirties, and retirement spending looks nothing like young spending, so the shape tells you which businesses have a tailwind and which are fighting the demographics. A bulge of young children makes a daycare a reasonable bet. An inverted pyramid points toward a shrinking economy. A large surplus of young men has historically gone alongside more violence. I look at the pyramid before I look at a market.

Paradoxes

Zeno's Paradoxes

Zeno's paradoxes are a set of arguments that motion is impossible. Give the tortoise a head start, and by the time Achilles reaches where the tortoise was, the tortoise has moved a little further, and so on forever, so Achilles never passes. Of course he does. The argument is airtight and the conclusion is wrong, which is why the paradoxes stay on my list. Business produces the same kind of debate, most often with lawyers who want cover for every eventuality. The logic runs perfectly and lands somewhere no commercial person would go. At some point you have to notice that Achilles ran past the tortoise.

The Stockdale Paradox

The Stockdale paradox is holding two things at once. Unwavering faith that you will prevail in the end, and the discipline to confront the most brutal facts of your current position. Jim Collins named it after Admiral Stockdale, who told him the optimists were the ones who did not make it out of the prison camp, because they kept setting dates that came and went. Turnaround work runs on the same balance. Lose the faith and the team stops trying. Lose the brutal facts and you build a plan for a business that does not exist.

Jevons Paradox

Jevons's paradox is that making something more efficient often increases total consumption of it rather than reducing it. Jevons noticed the pattern in 1865, when more efficient steam engines led to far more coal being burned, because efficiency made coal worth using in more places. Fuel economy in cars did something similar, and so did cheaper computing. I expect the same with AI. The cost of a unit of intelligence is falling fast, and the response so far has been to use a great deal more of it rather than to spend less.

One quote

Modern self-help contains an in-built flaw. To continually improve yourself, you must continually locate the ways you are broken.

That is Tim Ferriss, and the same flaw sits inside continuous improvement in a business.