Picture the decision-maker economics textbooks quietly assume you are. Faced with a choice, this creature knows every option that exists, assigns each a precise payoff, and lifts the single best one into their basket — instantly, at no cost, every time. Call them Homo economicus, the perfect optimiser. Now picture what you actually did the last time you chose a restaurant, a laptop, or a job: you looked at a handful of options, found one that seemed fine, and stopped. You did not survey the space. You did not rank everything. You took something good enough and got on with your life.
The gap between those two pictures is the whole subject of this course. The optimiser is a fiction — and not because you’re lazy or stupid. It’s because optimising is often impossible in principle. The options are too many to list, the future too uncertain to score, and the arithmetic too long to finish before the decision is due. A perfect chess player would have to search more positions than there are atoms in the universe. You have bounded time, information, attention, and computing power. Within those bounds, “find the best” is not a plan you can execute. So you do something else — and that something else turns out to be remarkably smart.
That something else is satisficing: a word Herbert Simon coined by fusing satisfy and suffice. You set an aspiration level — a bar for “good enough” — search until an option clears it, then stop and take it. Simon won the 1978 Nobel Prize in Economics for building this into a serious theory of decision-making called bounded rationality, whose core claim is disarmingly simple: real minds don’t maximise, they satisfice, and once you count the cost of thinking, that’s the rational thing to do.
The one-sentence version
Bounded rationality says real agents can’t optimise — they have limited time, information, and computing power — so instead they satisfice: set an aspiration level (a “good enough” bar), search until an option clears it, and stop. Much apparent irrationality is really a smart adaptation to a hard world.
Before you read — take a guess
Before we start — take a guess. According to bounded rationality, WHY don't real people find the single best option when they make a decision?
See it in one picture
Here is the whole tension on one chart. A field of options arrives one at a time, each with a hidden quality from 1 to 100. A satisficer sets an aspiration level — the horizontal axis — and grabs the first option that clears it. A maximiser inspects every option to find the true best. The vertical axis is net value: the quality you end up with, minus the cost of every look you had to take to get there.
Drag the aspiration slider. Too low, and you accept junk on the first look — low quality. Too high, and you search almost the whole field (paying for every look) and often get forced onto a poor last option. The net-value curve humps in the middle: there’s a best aspiration, and it isn’t “demand perfection.” The dashed line is the maximiser, who always gets the top quality — but pays to inspect everything. Now push the cost per look up and watch that dashed line sink below the satisficer’s peak. When looking is expensive, “good enough, fast” beats “the best, eventually.”
Satisficing lab
Satisficer vs maximiser: who actually wins?
Options arrive one at a time with a hidden quality (1–100). A satisficer sets an aspiration level and grabs the first option that clears it; a maximiser inspects everything to find the true best. Drag the aspiration, then raise the cost of each look and watch which strategy wins.
- Accepted quality
- 85
- Looks used
- 3.4
- Regret
- 12
- Net value
- 78
| Quality | Looks | Net value | |
|---|---|---|---|
| Satisficer | 85 | 3.4 | 78 |
| Maximiser | 97 | 25 | 47 |
With 25 options and an aspiration of 70, the satisficer accepts quality 85 after just 3.4 looks for a net value of 78. The best aspiration here is about 81. The maximiser finds the true best but pays for every look — its net value is only 47.
Watch one search
Flip the intuition around: the maximiser wins on raw quality every time — they do find the best jar of sauce in the aisle. But they pay for it in search, time, and (as we’ll see) in regret. The satisficer captures most of the value for a fraction of the effort. That trade — a sliver of quality traded for a mountain of saved cost — is the beating heart of this whole course.
In the lab, raising the 'cost per look' makes the maximiser's net value fall below the satisficer's peak. What real-world lesson does that encode?
What you’ll walk away with
By the end you’ll be able to look at a decision and ask the two questions that matter: should I optimise this, or satisfice it? and what’s my aspiration level — my bar for good enough? Here’s the map:
- Why optimisation fails — combinatorial explosion, unlistable options, and the cost of computation itself, so you feel why the perfect optimiser can’t exist.
- Satisficing & the aspiration level — the model made precise, and how it’s really a reservation-value stopping rule — the same machinery as optimal stopping.
- Maximisers vs satisficers — the psychology, the paradox of choice, and why chasing the best can leave you worse off.
- Fast-and-frugal heuristics — Gigerenzer’s ecological rationality: simple rules that often match or beat complex optimisation because they don’t overfit.
- Transfer — setting aspiration levels — how to pick and adjust the bar, and which decisions to satisfice and which to optimise — plus where the model lies.
Where we're headed
Keep one image in your head the whole way through: that hump-shaped net-value curve. Almost everything ahead is an answer to two questions — is this a decision worth optimising? and if not, where do I set the bar for “good enough” and then stop?
How to use this course
Every lesson opens with a quick guess, teaches the idea through a concrete story and worked numbers, and checks that it stuck. Don’t skip the guesses — committing to an answer before you know is one of the most reliable ways to actually remember. You’ll drive the satisficing lab from several angles and meet sorting and matching exercises along the way.
This is an expert-tier course, so it leans on a few models you’ve ideally met already. You’ll get the most from it if you’re comfortable with optimal stopping (satisficing is literally a reservation-value stopping rule — the aspiration is the bar, and “stop at the first good-enough option” is explore-then-commit with a fixed threshold), the value of information (search has a cost, so gathering more only pays while its expected value beats that cost — satisficing is where the marginal value of another look drops below its price), and circle of competence (bounded knowledge is exactly why you can’t optimise outside what you understand). When you’ve finished the five teaching lessons, a graded final exam pulls it all together — it’s one-way, so once you submit an answer it’s locked.
One habit to build as you go
Whenever you catch yourself paralysed — “there must be a better option, let me check one more” — ask two questions first: does this decision even deserve full optimisation? and have I already cleared my bar for good enough? If it doesn’t and you have, the rational move isn’t more searching. It’s to stop.
Ready? The next lesson goes straight to the foundation — why the perfect optimiser is a fiction, so that satisficing stops looking like a compromise and starts looking like the only game in town.