Mental Models: The Complete Library (100+ Models Explained)

What are mental models?

Mental models are compact explanations of how the world works that you use to think through a situation. They are not facts to memorize. They are lenses: each one makes a certain pattern visible, so you can spot it in a new problem and reason about it.

You already use dozens without naming them. “Don’t put all your eggs in one basket” is a model of risk. “You get what you pay for” is a rough model of incentives and quality. The difference with a deliberate library is that you choose the models, know their limits, and reach for them on purpose when a decision matters.

A good mental model has three traits:

• It is general. It shows up in business, relationships, money and health, not just one field.
• It is simple. You can state it in one sentence and recall it under pressure.
• It changes what you do. If knowing it would not change a single decision, it is trivia, not a tool.

The idea is older than the buzzword. The Scottish psychologist Kenneth Craik argued in The Nature of Explanation (1943) that the mind carries “small-scale models” of reality and runs them to anticipate what will happen. Cognitive scientists such as Philip Johnson-Laird built on that work. The business world picked up the phrase much later, mostly through Charlie Munger.

Why a latticework of models beats one big idea

A single model, used for everything, turns into a blind spot. Munger made this point in a 1994 talk at the University of Southern California business school, later published as “A Lesson on Elementary, Worldly Wisdom.” He argued that you need models from many disciplines (psychology, economics, math, biology, engineering) and that they should hang together in a “latticework” in your head. His warning was the old line: to a man with a hammer, everything looks like a nail.

Here is why the latticework matters in practice:

1. Problems are cross-disciplinary. A pricing decision is partly math (expected value), partly psychology (anchoring), partly strategy (incentives). One lens catches one third of it.
2. Models check each other. Second-order thinking might say “this will backfire in a year,” while opportunity cost says “doing nothing is also expensive.” The tension is the useful part.
3. Models fail at their edges. The map is not the territory: every model leaves things out. Having several lets you notice when one has stopped fitting.

The goal is not to collect models. It is to get to the point where, facing a real choice, two or three relevant ones come to mind without effort.

The five categories of mental models

Mentarium groups models by the job they do in a decision, not by the academic field they came from. That way you can find the right tool when you are stuck, rather than remembering which textbook it lived in.

Thinking clearly

These models help you see the problem before you try to solve it. First-principles thinking breaks a question down to what you know is true and rebuilds from there, instead of copying how things are usually done. Occam’s razor says to prefer the simpler explanation that fits the facts. Hanlon’s razor says not to assume malice where carelessness explains it. They keep you from solving the wrong problem with great skill.

Deciding under uncertainty

Most real decisions are bets. You rarely know how things will turn out, so the question becomes “what is the best choice given what I can know?” Probabilistic thinking replaces “will it work?” with “how likely is it to work, and what happens if it doesn’t?” Base rates ask how often things like this usually go well before you look at the details of your case. Margin of safety builds in room for being wrong.

Seeing consequences

First-order effects are what happens next. Second-order effects are what happens because of that, and they are where most surprises live. Feedback loops explain why some situations spiral (up or down) and others settle. Incentives explain why people do what they do, which is usually more predictive than what they say.

Avoiding mistakes

Charlie Munger often credited the mathematician Carl Jacobi with the advice “invert, always invert.” Inversion asks what would guarantee failure, so you can avoid it. The pre-mortem imagines the project has already failed and asks why. Chesterton’s fence asks you to find out why a rule exists before you remove it. These are the cheapest models to use and often the most valuable, because avoiding a big error is worth more than finding a clever win.

Time and priorities

You can’t do everything, so these models decide what gets your attention. The Pareto principle notes that a small share of causes often drives most of the results. The Eisenhower matrix separates urgent from important. Parkinson’s law warns that work expands to fill the time available, which is why deadlines help.

Mental models vs cognitive biases

Mental models and cognitive biases are two sides of the same coin. A bias is a predictable error your mind makes automatically. A model is a tool you pick up deliberately, and many models are, in effect, a correction for a specific bias.

That pairing runs through the whole library. Every model page names the bias it counters, and every bias page names the model that fixes it. If you want the full picture of the errors, start with the cognitive biases library; if you want to see the models applied to real choices, the decision-making guide puts them in order.

How to build your own Model Stack

Knowing a model and using it under pressure are different skills. The fastest way to close that gap is the Model Stack: three models combined for one situation. One model gives you a single angle. Three, chosen for the situation, give you a reasonably complete view without turning a decision into a research project.

Three examples:

• Quitting a job = Opportunity Cost + Regret Minimization + Reversible vs irreversible decisions. What are you giving up by staying? Which choice will you regret less at 80? How hard is it to undo?
• Launching a product = Pre-mortem + Pareto principle + Second-order thinking. How does this fail? Which 20% of the work drives most of the result? What happens after launch, when competitors and customers react?
• Big purchase = Opportunity cost + Margin of safety + Ten-ten-ten rule. What else could this money do? Can you absorb it if your income drops? How will you feel about it in 10 minutes, 10 months and 10 years?

To build a stack for your own situation:

1. Name the decision in one sentence. “Should I accept the offer from Company B by Friday?” is a decision. “Career stuff” is not.
2. Pick one model that widens the view. Opportunity cost or first-principles thinking are good defaults: they make you see options you were ignoring.
3. Pick one model that looks for failure. Inversion, the pre-mortem or margin of safety.
4. Pick one model about time. Second-order thinking, the ten-ten-ten rule or reversible decisions.
5. Write one line per model. What does each one tell you about this decision? If a model has nothing to say, swap it.
6. Look for the disagreement. Where two models point in different directions, that is the real question you need to answer.

The “widen, find failure, look ahead” pattern is not magic, but it covers the three ways decisions most often go wrong: too few options, unexamined risks, and short-term thinking.

The 5 mental models to learn first

If you are new to this, don’t try to learn a library. Learn five, use them for a month, then expand. These five cover the most ground with the least effort:

1. First-principles thinking. Ask what you actually know to be true, then reason up from there. Useful whenever “that’s how it’s done” is the main argument.
2. Second-order thinking. Ask “and then what?” at least twice. It is the antidote to decisions that look great this week and terrible next year.
3. Inversion. Ask what would guarantee a bad outcome, then avoid those things. Often easier than finding the perfect path.
4. Opportunity cost. Price every choice against the best alternative you are giving up, including your time.
5. Probabilistic thinking. Replace yes/no predictions with rough odds. “70% likely” makes you plan for the 30%.

A simple practice routine: pick one model per week. Each evening, find one decision from the day (even a small one) and write a sentence on how the model applies. After five weeks, run all five against one bigger decision. You will notice that you start asking the questions without the list.

How to use this library

Every model page follows the same structure, so you can scan it quickly:

• A definition you can quote, plus an “At a glance” strip with the origin and what it is best for.
• Trigger questions to ask yourself, the fastest way to apply the model in the moment.
• How to use it, in numbered steps.
• Three examples (business, personal life, and investing or tech). Examples are illustrations, often hypothetical, and say so.
• Common mistakes, because every model can be misused.
• A comparison with the model it is most often confused with.
• The bias it counters, linking the two halves of the library.

The library is growing toward 100+ models. It starts with the ones that come up most often in real decisions and adds more each week. The full list of pages in this section appears below.

If you think in situations rather than models, the guides on mental models for entrepreneurs, investors and programmers pick the models that matter most for each.

Common mistakes with mental models

Mental models are useful, and they are easy to misuse. Watch for these:

• Collecting instead of using. Reading about 100 models feels productive. Using three on a real decision is productive.
• Picking the model after the answer. If you already decided and then go looking for a model that agrees, you are dressing up a bias. Choose the stack before you have a favorite.
• Treating a model as a law. The Pareto principle is a tendency, not a rule that every situation splits 80/20. Models describe patterns; they don’t guarantee them.
• Ignoring your circle of competence. A model applied in a field you don’t understand gives you confident nonsense.
• Using models to avoid deciding. Analysis has a cost. For small, reversible choices, the right model is “decide and move on.”

The Mentarium take

Most mental model lists are written like a museum: a long hall of clever ideas, each behind glass. That is why so many people can name 50 models and use none. Our view is that a model only counts once it has changed a decision you made.

So we organize by the job the model does, and we push the Model Stack because it is a habit, not a reading list. Three models, one decision, one line each. It takes ten minutes, and it is enough to catch most of the mistakes that matter: missing options, hidden risks and short horizons.

If you only take one thing from this page, take the five starter models and use them this week on something real. The latticework Munger described is not built by reading. It is built one decision at a time.