Choices Updated September 2026

How to make a hard decision without waiting for certainty

Start by asking why the decision is hard. Are you missing information that could change the choice, or does each option cost you a version of yourself? The first problem calls for facts. The second calls for choosing without waiting for a version of the decision that carries no loss.

Two kinds of hard
Information

A concrete fact is missing: price, timing, eligibility, a medical result, another person's answer. Get it if you realistically can.

Identity

You know enough to choose, but one option means becoming someone who left, declined, committed, changed direction, or disappointed an expectation.

When more analysis still helps

Some hard choices really are information problems. If a contract term, inspection, salary number, or deadline could materially change the decision, gathering that information is not avoidance. It is part of deciding well.

A useful test is simple: name the missing fact and say what you would do differently depending on the answer. If you can do that, go get the fact.

If you cannot, the next spreadsheet may make you busier without making the choice clearer.

When the difficulty is what the choice would mean

Ruth Chang's work on hard choices makes a useful point: some options are not neatly better, worse, or equal on a single scale. Choosing can help create the reasons by which you live.

That is why a career move can stay hard after salary, commute, title, and risk have all been compared. One option may say, “I am still the person who stays.” The other may say, “I am willing to become the person who leaves.”

The difficulty is not evidence that there must be one hidden correct answer. Sometimes it is the cost of closing a real alternative.

Four things to do before you decide

A short protocol
  1. Name the choice in one sentence. Remove the background story.
  2. List the facts that could still change it. Get the ones you can actually obtain.
  3. Write the identity cost of each option. What would choosing it force you to admit, give up, or become?
  4. Save your present lean and state. A weak lean counts. No lean counts too. Date it before the outcome can make the choice look obvious.

Amazon's 2015 shareholder letter popularized the distinction between reversible and irreversible decisions. It is useful because reversibility changes how much evidence you need. A choice you can cheaply undo does not deserve the same threshold as a choice that closes a door for years.

But reversibility does not remove identity cost. You can leave a job and later change course, while still feeling that leaving says something enormous about who you are.

Example
Question
Should I take the smaller company offer?
Facts left
Equity terms; expected travel.
Identity cost
Leaving the safe path I said I wanted.
Lean
Take it, if travel stays under one week a month.
State
Excited and afraid of looking reckless.
Have the real question?

Give it something to react to.

Shadow OS gives a real question an outside signal. Agreeing, resisting, or wishing it had said something else can add another observation before the outcome is known.

Your reading arrives by email. Free.

Ask your question →

Do not let the outcome grade the decision

A good decision can produce a bad outcome. A careless one can get lucky. That is why judging the choice only after you know what happened is so tempting and so misleading.

The useful record is the one made before the result: what you knew, what you did not know, what you were leaning toward, and what the choice seemed to cost at the time.

Later, the look-back can ask a fairer question: given what was actually available then, did I choose in a way I still understand?

If the problem is not one unusually hard choice but a repeated inability to close any option, read why indecision keeps taking the same shape. If you cannot see a lean at all, start with what to do when you don't know what to do.

Bring the decision you are actually making.

Do not ask Shadow OS to remove the cost. Use the signal to notice your reaction, then keep your own choice on record.

Ask your question →

Shadow OS Editorial Research, sourcing, AI assistance, and human review are described in our editorial methodology.
Sources
  1. Chang, Ruth. “Making Comparisons Count.” Routledge Encyclopedia of Philosophy work on hard choices; see also her 2014 TED talk, “How to Make Hard Choices.”
  2. Amazon. 2015 Letter to Shareholders, on reversible and irreversible decisions.
  3. Kahneman, D., & Tversky, A. Work on prospect theory and loss aversion.