---
name: regret-minimization
description: Use when asked to decide a personal or career-scale choice where numbers do not settle it, by imagining oneself at a distant future point (for example age 80, or ten years on) and asking which option would be regretted least — covering regret of action versus inaction, regret of omission, and what the decision reveals about values — as distinct from a [[weighted-decision-matrix]] (quantified criteria), [[premortem]] (assuming a plan failed), and general [[journaling]].
---

# Regret minimization

Regret minimization asks one question: *looking back from a distant
future, which choice will I wish I had made?* It was popularized by
Jeff Bezos as the way he decided to leave a secure job and start
Amazon. It works best for choices that are significant, hard to reduce
to numbers, and likely to be distorted by fear of the short term. Short-term
fear tends to inflate the cost of acting; research on regret suggests
that over long horizons people regret inactions and missed chances more
than actions that failed.

## Input required

- **The decision and the options**, including staying put.
- **The time horizon** to look back from — age 80, end of career, or
  ten years from now.
- **What is at stake** — money, relationships, health, identity,
  freedom — and what is *not* at stake.
- **Reversibility** — what can be undone and what cannot.

If details are missing, state assumptions or ask the questions below once, then proceed.

## Questions to ask first

Ask in **one batch of at most five questions**, skipping any already
answered, then proceed:

1. What are the options, and which one feels like the "safe" default?
2. What are you most afraid of if you act? If you don't?
3. What would you be proud or ashamed to tell your future self about
   each?
4. What can be undone, and what is permanent or has a window that
   closes?
5. Whose expectations (family, employer, peers) are mixed into this,
   separate from yours?

## Method

1. **State the options plainly**, including the inaction option as a
   real choice with real consequences.
2. **Pick the vantage point** — a specific future age or date — and
   write a short note from that person's point of view for each option.
3. **For each option, answer two things:**
   - If it works, what do I gain?
   - If it fails, what do I regret *about having tried*, and is that
     regret one I could live with, learn from, or recover from?
4. **Then answer the inverse:** if I do nothing, what will I wonder
   about? Name the "what if" in a sentence.
5. **Separate kinds of regret.**
   - *Regret of action* — pain from a bad outcome I caused. Usually
     fades and teaches.
   - *Regret of inaction* — an unanswered "what if." Usually lingers.
   - *Regret of harm* — damage to others or to commitments. Weigh
     separately and heavily.
6. **Check for fear and borrowed values.** For each reason against
   acting, ask whether it is a real constraint (a mortgage, a
   dependent) or a feeling.
7. **Limit the downside.** If action wins but the downside is
   frightening, look for a smaller version: a trial period, a leave
   of absence, a side project, a savings runway.
8. **Decide and state the reason** in one sentence, in values terms.

## Output format

- **Decision** — one sentence.
- **Per option** — the regret note from the future vantage point (3–5
  sentences) and the kind of regret it would be.
- **Fears sorted** — real constraints versus feelings.
- **Leading choice** and the **smaller, safer version** if there is
  one.
- **Condition for revisiting** — what new fact would change the
  answer.

## Ground rules

- **Be honest, not noble.** The future self is the actual person, not a
  heroic version.
- **Constraints are real.** Do not use this to talk someone out of
  obligations to dependents or into financial ruin; keep catastrophic,
  irreversible downside in view.
- **The heuristic informs; it does not decide.** Pair it with facts.
- **Not for clinical decisions.** If the choice involves self-harm or a
  crisis, point to professional or emergency help rather than
  analyzing it.

## Self-check before answering

- Was inaction treated as a real option with its own regrets?
- Are real constraints separated from fear?
- Is there a way to act while limiting the downside?
- Does the answer follow the person's values, not the assistant's?

## Common pitfalls

- **Treating "stay" as free** — the default has costs that accumulate
  silently.
- **Recklessness dressed as courage** — ignoring a downside that
  cannot be recovered from.
- **Using someone else's future self** — imagining the regret of
  parents or peers, not one's own.
- **Too short a horizon** — a one-year view mostly measures anxiety.
- **Skipping the smaller version** — a staged step often gets most of
  the upside with less of the risk.

## Example prompt

> I've been offered a role abroad, but my current job is stable and
> my family is here. Use regret minimization: ask me what you need,
> then tell me which choice I'm least likely to regret at 70, and
> whether there's a lower-risk way to try it.

## Learn more

- [[two-way-door-decisions]] for judging how reversible the choice is.
- [[weighted-decision-matrix]] for the quantifiable part of the choice.
- [[premortem]] to stress-test the option you lean toward.
- [[decision-journal]] to record the reasoning and revisit it later.
- [[journaling]] for reflective writing.