Decisions You Can Return To

Nadia wants to draw again. She has a borrowed sketchbook, a pencil and two possible times: Tuesday evening or Saturday morning. By Thursday she has made a third list of reasons and has not drawn anything. The question has quietly grown from When will I practise? to Am I the sort of person who ever follows through?

Those are different questions. This book works with the first. It offers a way to name an ordinary choice, record what you know, act within agreed limits and review what happened. The aim is a usable record, not a verdict about your character.

Nadia is fictional. So are the other worked examples. Their outcomes illustrate how a decision record can be read; they are not research participants or evidence of a success rate. The decision cards in this book are editorial tools, not validated psychological tests.

Begin with a low-stakes choice you control. Choosing a folder structure, a practice slot or a small volunteer task can be suitable. Do not turn another person's privacy, your health, an essential payment or a binding commitment into an experiment because a worksheet has space for it. Some choices cannot be undone, and returning to a decision does not restore the original situation.

Write one sentence: For the next defined period, I will choose between these specific options in order to achieve this observable purpose.

Nadia writes: For the next two weeks, I will try a drawing slot that allows me to finish two small practice sketches. She has not promised to become an artist or to love every minute. She has made a question small enough to answer with experience.

Keep the first version. Later you will need to know what you actually chose, not what the outcome makes you wish you had chosen.

A choice can be yours without all its consequences belonging to you. Nadia shares a kitchen table. Saturday morning is not empty merely because her own calendar is blank. She asks the people who use the space whether a thirty-minute slot would work. They agree a specific time. This is an agreement in the fictional example, not a rule that households should organise themselves around one person's worksheet.

On your card, distinguish the person making the choice, people whose agreement is needed and people whose advice might help. These roles are not interchangeable. A colleague may offer useful advice without being entitled to make the choice. Someone whose time you need may be entitled to say no without producing a better alternative for you.

Avoid writing other people's names into an empty resource column before asking them. If your plan needs a friend to cover a task, mark it as a request pending confirmation. The option is not ready to compare with a fully available alternative until that dependency is resolved or honestly labelled uncertain.

For a private filing experiment, the affected person may be only you. For a volunteer newsletter, subscribers' information, another editor's time and the group's publishing process may be involved. A personal experiment does not grant permission to copy a mailing list or change a shared system.

Your output for this chapter is a boundary sentence: I may decide this part; I must ask about that part; this other part is outside the trial.

Nadia may choose to draw. She must agree access to the shared table. She will not take photographs of family papers on it for a progress post. A clearer boundary makes the trial easier to interpret because it does not conceal a cost someone else is paying.

Give each important statement a label. A fact has a basis you can name. An assumption is something you are relying on without confirming. A question is something you intend to check. These labels describe the status of the statement, not how confidently you say it.

Nadia's fact: she has a pencil and a sketchbook. Her assumption: she will feel more willing to draw on Saturday than after work on Tuesday. Her question: is the table available at the proposed time? When the household agrees, she updates that last item with the date. She does not change her earlier note to pretend it was always known.

Fischhoff showed that knowing results can distort retrospective judgments. [S2]

That research statement is a reason to be modest about recollection, not proof that this particular card will remove a bias. The practical suggestion here is simple: retain the original note and add a dated update. You should be able to distinguish a prediction from a later observation.

Do not collect information simply because it is collectible. Ask what a new answer could change. If neither answer would affect your available options, record it as interesting rather than essential. Conversely, a missing permission can matter even if it takes only one message to resolve.

For ordinary low-stakes choices, make a short check list before browsing: what must I know, where could I check it and when will I stop? This is an editorial stopping rule, not a scientifically optimal number of searches. The appropriate amount of checking depends on the possible harm, uncertainty and reversibility.

End with three lines you can defend: what I know; what I am assuming; what would change the choice. If all three say probably, your next action may be a question rather than a commitment.

An option needs a next action and a cost. Draw on Saturday sounds plausible. Draw when life is calmer is not comparable until it says what will happen and when. Include doing nothing for now if that is a real option, but state what waiting preserves and what it postpones.

Nadia compares Tuesday at seven with Saturday at ten. Both are thirty-minute practice sessions at home. Tuesday competes with an existing household task. Saturday needs an agreed table slot. She is not choosing between discipline and failure. She is choosing between arrangements with different constraints.

Use a few criteria that matter to the stated purpose. For this example: actual availability, a place to work and a finish small enough for the time. A hundred weighted scores would not make the unconfirmed table more available. If you use numbers, show what they represent and avoid presenting subjective scores as measured probabilities.

Amazon's 2015 shareholder letter distinguishes reversible from difficult-to-reverse business choices. [S5]

The letter is a corporate example, not psychological proof or a permission slip for haste. In this handbook, reversibility is a question to examine concretely: what can be changed, by whom, at what cost and with whose agreement? A cancel button does not necessarily restore time, money, relationships or disclosed information.

Nadia can change her next practice time. She cannot recover a missed family commitment by deleting a calendar entry. She therefore excludes conflicting times rather than counting them as cheap experiments.

Choose an option only after identifying what makes it feasible. Then write why the unchosen option remains unchosen today. This gives your future review something more useful than I just felt it was right, while leaving room for feelings to be named as feelings rather than disguised as facts.

A trial has an edge: a beginning, a limit and an exit. Without those, try it for a while can become an arrangement nobody remembers agreeing to. An edge is especially useful when other people's time is involved.

Nadia chooses two Saturday sessions of thirty minutes. Her purpose is two small sketches, not a polished portfolio. Her total planned drawing time is sixty minutes. She adds ten minutes for setting up and ten for a review, making eighty minutes of planned commitment across the trial. These are fictional amounts, not a recommended dose of practice.

Her exit is concrete. If the shared space becomes unavailable, she will pause and ask about another arrangement rather than silently moving someone else's belongings. If the sessions are consistently unsuitable, she may change the time or stop this trial. She need not keep going merely to protect a sentence she wrote last week.

Separate the review date from an urgent stop condition. A minor inconvenience can wait for review; a missing permission or a risk outside the agreed scope should not be endured until the calendar says it is time to think. State the condition in observable terms. I feel bad about this is worth noticing, but I cannot access the materials without using someone else's private account is a specific boundary.

You may discover that the supposedly small trial is not small. That discovery can be the useful result of planning. Reduce the scope, obtain the necessary agreement or choose another option. Do not hide a large commitment inside a short title.

Before starting, a person who has not followed your thinking should be able to read the card and understand what you will do, when you will review it and what would make you stop sooner.