Pocket Laws 101: The Mental Models Worth Carrying Around

Also available on Medium.

An observation someone made decades ago can save you from a bad decision. This is a field guide to fourteen pocket laws: simple mental models worth carrying around and reaching for, whenever life throws a familiar situation your way.

They’re less like laws and more like patterns — simple enough to remember, sharp enough that once you’ve noticed one, you start seeing it everywhere: at work, with friends, in your own decisions.

Murphy's Law: anything that can go wrong, will

This is the most famous law of all — and the most misunderstood. Here’s the real idea: if something can go wrong, and you never plan for it, it eventually will. Think about the spare tire. You stopped carrying one — you hadn’t had a flat in ten years, so why bother? Then one day, on a long drive, you get a flat. And there’s nothing in the trunk to help. Or your laptop: it hasn’t crashed in years, so you stopped backing things up — and then one day it crashes, right before a big deadline. That’s not bad luck — it’s just odds. Do something enough times, and the thing you didn’t prepare for shows up. Even small odds are still odds — prepare for the ones that would hurt the most if they happened.

Moore's Law: computing capability roughly doubles every couple of years

Most things grow one step at a time. Other things multiply instead. The most common version of that is doubling — and doubling adds up faster than most people expect.

Think about a pond with a single water flower floating on it. Every day, the flowers double in number. In 30 days, the whole pond is covered. So on what day is the pond half covered? Most people guess day 15. The real answer is day 29 — just one day before the end. Almost all the growth happens right at the very end, because doubling means each new day adds more than every day before it, combined.

This is exactly what happened with computers. In 1965, an engineer named Gordon Moore noticed something: the number of tiny switches on a computer chip was doubling every two years. That doubling kept happening for six decades. Today, the phone in your pocket has more computing power than the room-sized computers NASA used to land astronauts on the moon.

This is Moore’s Law. These days, it’s slowing down — chips are bumping up against the physical limits of how small a switch can get. But the real lesson was never really about chips. It’s that people are bad at guessing how big something gets once it starts doubling — so when something doubles, trust the math, not your gut.

Sunk Cost Fallacy: throwing more in because of what you've already spent

Once you’ve put something into a decision — money, time, effort, whatever it is — it’s gone. You’re not getting it back, no matter what you do next. That means it shouldn’t affect your next decision at all. But it usually does. Think about watching a boring movie. An hour in, you’re not enjoying it. But you keep watching anyway, because you already sat through the first hour. Walking away now would mean “wasting” that hour — even though it’s already gone either way. Or think about a buffet. You’re already full, but you keep eating because you paid for the meal. Eating more food doesn’t get your money back. It just makes you feel worse. This is the sunk cost fallacy: treating money or time you’ve already spent as a reason to keep going, even when it isn’t. The fix is simple, if not easy: ask whether you’d still choose to start this today, knowing what you know now. If the answer is no, the smartest move is to stop — the time or money already spent isn’t coming back either way.

Parkinson's Law: work expands to fill the time you give it

Give a task two weeks and it takes two weeks. Give the same task two days and, somehow, it takes two days. The work itself didn’t change size — the deadline did. Think about packing for a trip. Give yourself the whole day, and packing somehow eats the whole day — extra folding, second-guessing outfits, repacking the bag twice. Give yourself thirty minutes before you have to leave, and the same bag gets packed in thirty minutes. Or think about a school essay: give yourself a month, and you’ll still be finishing it the night before it’s due. Give yourself three days, and somehow, it still gets done just the same. More time doesn’t mean better work. It usually just means slower work. If you want something done well and done fast, give it a tighter deadline, not a longer one.

Pareto Principle (the 80/20 Rule): most of your results come from a small slice of effort

A handful of what you own, use, or do ends up producing most of the good stuff. Look at your closet: a few of your clothes get worn constantly — jeans, that one hoodie, a couple of t-shirts — and the rest just hangs there. Or think about a business: a small number of customers usually bring in most of the revenue, and the rest barely add up to much. This is the Pareto Principle, also called the 80/20 rule: roughly 80% of results come from about 20% of causes. The exact numbers are never exact — sometimes it’s 70/30, sometimes 90/10 — but the pattern shows up everywhere. The trick is figuring out which small part actually matters, instead of spreading your time evenly across everything.

Streisand Effect: trying to hide something draws more attention to it

Trying to hide something is often exactly what makes people notice it. In 2003, Barbra Streisand sued a photographer to get a picture of her house taken off the internet. Before that, almost nobody had looked at it — just a handful of people scrolling through aerial coastline photos. The lawsuit made headlines, and suddenly hundreds of thousands of people rushed to see the exact photo she was trying to bury. Or think about the time you had a crush on someone. You told yourself you’d play it cool. Then they walked into the room — and suddenly you had way too much to say about absolutely nothing, tripping over your own words, and finding your phone deeply, deeply interesting. Everyone noticed in about five seconds. You weren’t hiding the crush. You were broadcasting it, just in code.

Goodhart's Law: a metric stops meaning anything once it becomes the target

Give people a number to chase, and they’ll chase the number, not the reason behind it. Think about weighing yourself every day to get healthier. At first, the number on the scale is just information. But once you start chasing the number itself, you might skip meals, or stop drinking water the night before you weigh in. The number goes down. You’re not actually healthier. Or think about measuring engineers by how many lines of code they write. Once that becomes the target, code gets longer and more complicated. Not because the work is better — because the number is what’s being graded. This is Goodhart’s Law: once a number becomes the goal, people focus on making the number go up, not on actually getting better. The fix isn’t to stop measuring. It’s to keep checking whether the number still means what you think it means.

Jevons Paradox: efficiency gains often increase total consumption

When a process gets more efficient, it needs less of whatever resource it depends on. So you’d expect total demand and consumption of that resource to drop too. Instead, they usually go up. Think about coal-powered steam engines in the 1800s. As engines got better at turning coal into power, people expected coal use to drop. Instead, the opposite happened: the engines were now cheaper to run, so more factories started using them, and steam power spread everywhere. All those new engines burning coal added up to more total coal use, not less.

A worked example: say a factory burns 1 unit of coal per unit of output, producing 100 units — 100 units of coal burned. A new engine cuts fuel use per unit to 0.5. Naively, coal use should halve, to 50. But output is now half as expensive to make, so demand rises — production climbs to 220 units to meet it.

0.5 × 220 = 110

Efficiency doubled, and total coal use still went up. Or think about AI and coding today. AI makes it faster and cheaper to write code, so it’s tempting to think companies will need fewer engineers. But the opposite might happen: cheaper code means teams take on more projects, and all that extra code still needs engineers to review it, test it, and catch what the AI got wrong. The work doesn’t disappear. It just shifts — from writing code to reviewing it.

Hofstadter's Law: it always takes longer than you expect, even accounting for this law

Things almost always take longer than you think, even when you already know that. Think about a home renovation: the contractor says six weeks, so you tell yourself it’s probably more like eight — and somehow, it still takes twelve. Or think about a school report: you give yourself an extra day just in case, and you still end up finishing it at midnight the night it’s due. That’s Hofstadter’s Law, named after Douglas Hofstadter, who made it a joke about itself: any extra time you add just becomes part of the new estimate, which runs over too. The fix isn’t a smarter guess — it’s checking in early and often, instead of waiting until the deadline to find out you’re behind.

Peter Principle: people rise until they hit their level of incompetence

Any time you rise in a skill or a ranking — a game, a sport, a job — you keep climbing based on how well you’ve done so far, until you finally hit a level where you’re not the best anymore. Think about chess ratings: you win games and your rating climbs, and keeps climbing, until you start running into players who are actually better than you — and that’s where it levels off. Or think about the best salesperson at a company: she’s amazing at selling, so she gets promoted to sales manager. But managing people is a different skill, and suddenly she’s just okay, maybe for the first time in her career. This is the Peter Principle: you keep rising based on what you’ve already proven, until you land in a spot where your skills don’t quite fit anymore — and that’s where you stay. The real fix isn’t to stop climbing. It’s to look ahead — figure out what the next level actually needs, and start building those skills before you get there, instead of assuming what worked before will keep working.

Chesterton's Fence: don't remove a rule until you know why it's there

If you don’t understand why something exists, that’s not a good enough reason to get rid of it. Think about a fence in the middle of an empty field, with no gate or animals in sight. It’s tempting to tear it down since it seems pointless. But maybe that side of the field turns to mud every spring, and the fence keeps people from wandering into it without knowing. You’ve just never been there when it floods. Or think about a weird approval step at work that everyone complains about, and nobody remembers the reason for. It might genuinely be useless. Or it might exist because of a huge mistake that happened years before you got there. This is Chesterton’s Fence, named after the writer G.K. Chesterton: before you remove a rule, a process, or a fence, find out why it was put there first. If you can explain the reason and still think it should go, fine — but “I don’t get it” isn’t the same as “the reason for it is gone.”

Parkinson's Law of Triviality (Bikeshedding): groups argue longest about the least important decision

In group decisions, the small stuff often gets way more debate than the big stuff. Parkinson’s own example: a committee has to approve building a nuclear power plant. They spend ten minutes on the reactor design, because almost nobody in the room understands nuclear physics well enough to argue about it. Then they spend forty-five minutes arguing about the paint color for the bike shed out back — because everyone has an opinion about paint color. Or think about buying a car. Almost nobody debates the financing terms or the engine warranty — that stuff is complicated, so most people just trust whatever the dealer tells them. But everyone has a strong opinion about the seat fabric, and that conversation goes on forever. This is Parkinson’s Law of Triviality, also called bikeshedding: the simpler and more familiar a topic is, the more everyone argues about it — because everyone feels qualified to have an opinion. The complicated stuff, which few people actually understand, gets rubber-stamped instead.

Brooks's Law: adding people to a late project makes it later

When a task is running late, throwing more resources at it seems like the obvious fix. Sometimes that’s true — more money or better tools really can speed things up. More people usually doesn’t — it often makes things worse. Think about a software team that’s behind on a deadline. The manager adds three new engineers to speed things up. But those new engineers don’t know the codebase yet — so the engineers who do know it have to stop coding and start teaching. The team that was already behind is now even more behind, just to get the new people caught up. Or think about cooking a big dinner. If you’re falling behind schedule, you might call in extra people to help. But now everyone’s bumping into each other, asking where things are, and waiting for the stove instead of using it. More hands isn’t the same as more speed — sometimes it’s just more traffic. This is Brooks’s Law, named after Fred Brooks, who wrote about it in a famous book on software projects: throwing more people at something that’s already late usually makes it later, not sooner.

New people cost time to train, and more people cost more time to coordinate. Here’s why: each person on a team could need to talk to every other person. With 5 people, each person has 4 others they might coordinate with — 5 times 4 is 20, but that counts every pair twice, since “Alice talking to Bob” is the same as “Bob talking to Alice.” Divide by 2, and you get 10 unique pairs. Now double the team to 10 people. Each person has 9 others to coordinate with — 10 times 9 is 90, divided by 2 is 45 unique pairs. You doubled the people, but more than quadrupled the number of relationships everyone has to manage.

Formula: n(n−1)/2, where n = number of people on the team

The fix isn’t more people — it’s fewer people with more focus, or just accepting the deadline needs to move.

Loss Aversion: losing something hurts more than gaining the same thing feels good

Losing something feels a lot worse than gaining the same thing feels good. Not just a little worse — researchers have found it’s close to twice as bad. Think about finding $20 on the street versus losing $20 out of your pocket. Finding it feels nice for a minute. Losing it can ruin your whole day. Same amount of money, completely different weight. Or think about selling something you own, like an old phone or a game console. You’ll often ask for more money than a stranger would actually pay for it — because giving it up feels like a loss to you, even though you weren’t planning to use it again anyway. This is Loss Aversion, first described by the psychologist Daniel Kahneman: people feel losses roughly twice as strongly as they feel equivalent gains. It’s why people hold on to a falling stock instead of selling, stay in situations they should leave, and fight harder to avoid losing something than they would to gain that same thing. If you catch yourself fighting hard to avoid a loss, ask whether you’d put in that same effort to gain the equivalent amount. If not, the fear of losing is doing more work than the actual stakes deserve.

Putting it together

Here’s the set, in one place:

None of these are predictions — they’re pattern-matchers. You don’t use Brooks’s Law to calculate an exact date; you use it to notice, in the moment someone suggests “let’s just add two more people,” that the instinct is probably wrong. That’s what a pocket law is for: not precision, but a fast, cheap check against a mistake you’ve probably lived through before, even if you couldn’t have named it at the time.