Think of the last big decision that went badly. A job you took, a person you trusted, a house, a move, a yes you should have turned into a no.

Now notice what your mind says about it. If you're like most people, it says something like: I had a bad feeling. I should have known.

Here is the uncomfortable part. You may not have had that feeling at all. The feeling may have been written in afterward, by a memory that edits the past to fit the ending. You aren't only judging yourself harshly for a bad outcome. You're judging a version of yourself who never existed, one who had information the real you didn't.

The past you remember isn't the past you lived

In 1975, the psychologist Baruch Fischhoff published a now-classic set of experiments on what became known as hindsight bias. People were given descriptions of events and estimated how likely various outcomes were. When others were told how things actually turned out, they consistently believed that outcome had been more predictable than people without that knowledge judged it to be. Even when asked to recall what they would have predicted beforehand, their answers drifted toward the known result.

The feeling it produces is a quiet, persuasive certainty: of course it went that way. Once you know the ending, your mind builds a story in which the ending was always visible.

This matters because the real purpose of looking back at a decision is to learn from it. If your memory has already smoothed away the uncertainty you actually felt, you learn the wrong lessons. You conclude you should have trusted a gut feeling you never had. Or you decide you're simply bad at choosing, when you made a reasonable call with the information in front of you.

The second trap: judging the choice by the result

Hindsight bias has a cousin. The author and former poker player Annie Duke calls it "resulting," which means judging the quality of a decision by how it turned out.

Poker makes the problem easy to see. You can play a hand well and lose to a lucky card. You can play it badly and win. If you learn only from outcomes, you'll start repeating your lucky mistakes and abandoning your sound choices that happened to meet bad luck.

Life is full of this. You take the job with a reasonable risk and the company folds. You skip the risky investment and it soars. Neither result tells you much about whether the thinking was good. But without a record of the thinking, the outcome is all you've got, and it will always feel like the whole story.

What a decision journal actually is

A decision journal is a short note you write at the moment you decide, before you know how it goes. Its job is to freeze a snapshot of your mind: what you knew, what you wanted, what you feared, what you expected.

The idea has serious backing. Daniel Kahneman, who spent a career studying judgment, has endorsed the practice of recording decisions so they can be reviewed later, and it appears in the 2011 Harvard Business Review piece he co-wrote with Dan Lovallo and Olivier Sibony on how organizations can improve their decisions. It isn't magic, and nobody has proved it makes you wise. But it directly attacks the two distortions above, because a note written in the moment can't be edited by a memory that hasn't happened yet.

The point isn't to be right. The point is to be able to tell, later, why you were right or wrong.

What goes in an entry

Keep it small enough that you'll actually do it. A useful entry answers a handful of plain questions:

  • What am I deciding, and what are the real options? Include the one you're not seriously considering. It's often the revealing one.
  • What do I expect to happen? Be concrete. "The team will ship by March" can be checked later. "It'll probably go well" can't.
  • How confident am I? Put a rough number on it, even "about 70 percent." This is the part hindsight erases most completely.
  • What's pushing me? Fear, flattery, tiredness, a deadline, someone else's opinion. Naming your state at the time matters, because you'll want to know later whether a mood made the call.
  • What would change my mind? If you can't name anything, you may have decided already and be gathering permission.

One optional addition comes from the psychologist Gary Klein, who described the "premortem" in a 2007 Harvard Business Review article. Imagine it's a year from now and the decision has failed, then write down why. It's a surprisingly effective way to surface the risks you were politely ignoring.

Review it like a scientist, not a prosecutor

Pick a date when you write the entry, a few months out for most decisions, and look at it then. Read what you wrote before you read how things turned out in your own head.

Then ask two separate questions, and keep them apart:

  1. Was my information and reasoning sound, given what I knew?
  2. How much of the result was luck, good or bad?

Sometimes you'll find you reasoned well and got unlucky, which is a relief and a reason to stay the course. Sometimes you'll find a pattern, like always overestimating how fast things get done, or deciding best in the morning and worst at 11 p.m. Those patterns are the real prize. They're invisible without a record, because memory tidies them away.

There's also a gentler effect that people don't expect. Reading your own honest uncertainty from three months ago is oddly kind. You see someone doing their best in fog. It's hard to keep calling that person a fool.

Why most people stop after a week

The usual failure isn't forgetting. It's friction. The entry feels like homework at exactly the moment you're most stressed and least inclined to write an essay about your own head.

So make the barrier lower than the decision. Five sentences is plenty. Better still, don't write it at all in the first pass. Say it. Talking through a choice out loud tends to be easier than composing it on a keyboard, and it keeps you in the loose, honest register of explaining to a friend rather than the stiff register of writing for an imagined audience. You can tidy it afterward, or not.

Your next moves

  • Choose one live decision today. Not a trivial one and not the biggest of your life. Something with a real outcome you'll see within a few months: a hire, a purchase, a project commitment, a conversation you're about to have.
  • Write or speak five lines before you act: the options, what you expect, your confidence as a number, what's pushing you, and what would change your mind.
  • Add a one-sentence premortem. Finish this: "It's a year from now and this went wrong because..."
  • Set a review date now. Put it in your calendar with the note's location in the title, so future you can actually find it.
  • On review day, read the note first. Then answer two questions in writing: "Was my reasoning sound?" and "How much was luck?" Don't skip the second one.

Where a voice-first tool fits

I'll admit the practice stands or falls on how little effort it takes. For me, the version that stuck is the spoken one: a minute of talking through the decision, then letting the words land as text. That's the reason an app like Quill suits this habit. It's a dictation tool that works in any app and runs on-device, so a half-formed thought about a private decision stays on your phone or computer, and you can ask it to tidy the rambling into clean notes afterward. If you'd like to try capturing your reasoning that way, you can find it at quill.lumenlabs.works. But the notebook, the notes app, and the back of an envelope all work. What matters is that the note exists before the ending does.