Thinking Lab · 4 lessons

Decision Thinking

Free — no account needed to read

Make better decisions under uncertainty

In one line: Judge a decision by the reasoning available at the time, not by how it turned out.

When you need it

The choice is between two things you genuinely want.
It will be hard or expensive to reverse.
It worked out badly and you are about to learn the wrong lesson from it.

The common misreading

What people think: A bad outcome means it was a bad decision.
What is true instead: Uncertainty means good reasoning sometimes loses. Grading yourself on outcomes teaches you to avoid risk rather than to reason well.

After this module you can record a real decision with its assumptions and a date to look again.

The key idea

You cannot control outcomes. You can control the quality of the decision, and only one of those is worth judging yourself on.

Start with a hard question

Write both. Then ask whether the thing you are waiting for is actually knowable in advance.

What decision are you currently avoiding, and what are you waiting to know?

Answer it before reading the lessons — then again after, and notice what changed.

The lessons

Decisions are not predictions — Write what you are optimising for before you look at the options, or the options will write it for you. (6 min)
Trade-offs, opportunity cost and uncertainty — Compare against the best alternative use of the same resources, not against doing nothing. (7 min)
Reversible decisions and second-order effects — "And then what?" asked three times is the cheapest protection against a plan that works exactly as intended. (6 min)
Decision journals and post-decision review — Review assumptions, not outcomes. A right answer for wrong reasons teaches you something false. (6 min)

Practise it

Short exercises. Under each is what a considered answer tends to contain — not a mark scheme.

The trade-off you did not name

You take a job with a 30% higher salary and a two-hour daily commute. Name every trade-off, including the ones that do not appear on a spreadsheet.

Tends to contain: Ten hours a week and what they were for
Tends to contain: Energy, health, relationships
Tends to contain: What the money is actually for
Tends to contain: Whether the comparison was against the current job or against the best alternative

Three futures

You are considering six months of unpaid study. Describe the good, expected and bad cases, and say which one you can survive.

Tends to contain: Concrete numbers, not moods
Tends to contain: Whether the bad case is recoverable
Tends to contain: What would signal early that it is the bad case
Tends to contain: The decision may turn on survivability rather than likelihood

Sort by reversibility

Sort these by cost of reversal: changing your CV, moving city, buying a laptop, leaving a course, telling someone how you feel. Which deserve the slow pass?

Tends to contain: Reversal cost is not the same as size
Tends to contain: Some social decisions are harder to undo than financial ones
Tends to contain: Fast decisions on the reversible ones are a win, not carelessness

Allocate the week

You have ten free hours this week and four things that all matter. Allocate them and state what each hour is bought at the expense of.

Tends to contain: Every hour has a named opportunity cost
Tends to contain: Whether anything got zero and whether that is a real choice
Tends to contain: What would make you reallocate mid-week

Reflect

Looking back at a decision you recorded, which assumption turned out to be carrying the most weight?

Mental models for this

Opportunity cost — The real cost of a choice is the best thing you did not do with the same time or money.
Reversible and irreversible — How much deliberation a decision deserves depends mostly on what undoing it would cost.
Margin of safety — Leaving room between your estimate and the point of failure, because estimates are wrong.
Premortem — Imagining a failure has already happened so the reasons become sayable.
Sunk cost — What has already been spent cannot be recovered by continuing, so it should not enter the decision.
Anchoring — The first number mentioned shapes every judgement after it, even when it is known to be arbitrary.
Thinking in ranges — Describing an uncertain outcome as a set of possibilities with rough weights, rather than one prediction.

Methods for this

The outside view — Estimating from what usually happens in cases like this, before reasoning from your own.
Premortem — Imagining the failure has already happened, so the reasons become sayable.
Separate the decision from the outcome — Judging a choice by what you knew at the time, not by how it turned out.
Widen, test, distance, prepare — Four counters to the four ways decisions usually go wrong.
Sort by cost of reversal — Spending your deliberation where undoing would be expensive.
Know your walk-away — Deciding what you will do if there is no agreement, before you negotiate.