How to tell if a decision was good or you just got lucky

A decision and its outcome are two separate events connected by chance as well as by judgement, and the size of that gap is the thing almost nobody accounts for. You took the job that turned out badly — was the reasoning wrong, or was it a reasonable bet that happened to land on the wrong side? You skipped the argument and it blew up later anyway — was staying quiet the mistake, or would speaking up have blown up just as badly, only sooner? The honest answer requires separating two questions that feel like one: was this decided well, and did it turn out well. They are not the same question, and collapsing them is the single biggest reason people learn the wrong lesson from their own history.

Why the outcome is a bad teacher on its own

Three things make outcome, taken alone, an unreliable source of feedback.

1. Most real decisions involve chance you do not control

A well-reasoned choice can still lose to a bad break, and a careless one can still win because nothing happened to punish it this time. If you only grade the result, you will sometimes conclude that a reckless approach works — because it worked once — and that a sound one fails, because this particular instance did not land. Both conclusions are wrong, and both feel like hard-won wisdom while you are drawing them.

2. You only ever see the branch you took

The job you turned down, the conversation you had instead of the one you avoided — those branches never resolve, so you cannot compare your actual outcome to the one you would have gotten. Grading a decision purely on what happened next quietly assumes the alternative would have gone perfectly, which is almost never a fair comparison.

3. The story rewrites itself once you know how it ended

The moment you know a result, your memory of how confident you were beforehand adjusts to match it. A choice that felt like a genuine toss-up at the time gets remembered as obvious in hindsight, in either direction — "I always knew that would work" or "I should have seen it coming." This is not dishonesty; it is how memory behaves by default, and it means you cannot trust an unaided recollection of your own reasoning even a few weeks later.

The two scores, kept separate

Every decision has a process score — was the reasoning sound, given what you actually knew at the time — and an outcome score — did it turn out well. The skill worth building is holding both at once without letting one contaminate the other. A good process with a bad outcome is a loss you should not overcorrect from. A bad process with a good outcome is a win you should not repeat. The second one is the more dangerous case, because a good outcome feels like permission to stop examining the reasoning, and that is exactly the situation where a bad habit gets reinforced instead of corrected.

How to score the process, honestly

Ask these about the moment of the decision, not about what you know now:

• What did I actually know at the time, not what I know now? Write the honest list. If the information that would have changed your mind was not available to you then, the decision cannot fairly be graded against it.
• Did I consider the realistic alternatives, or just the first one that occurred to me? A rushed decision is not automatically wrong, but it is a different kind of decision than one that was actually weighed, and it deserves a different kind of review.
• Would I make the same call again, facing the same uncertainty, before knowing the result? This is the question that actually isolates process from outcome. Answer it by imagining yourself back at the moment of deciding, with the same missing information, not at the moment of finding out.
• Was I deciding in a state I would trust — or rushed, tired, angry, flattered, under pressure to just be done with it? A sound-looking decision made from a bad state is usually a bad process wearing a good outcome's clothes; the state matters as much as the logic.

A worked example

You take a job offer over a safer one because the team and the mission both look genuinely strong, and eighteen months later the company runs out of funding. The outcome score is bad — you are looking for work again. The process question is separate: given what you could actually verify at the time — the team's track record, the funding runway you were told about, the references you checked — was that a reasonable bet? If you did the checking a careful person would do and the company simply ran into something unknowable in advance, the process was sound and the lesson is not "never take a risk on a smaller company." It is closer to nothing at all, beyond one more data point in a distribution you already knew had some bad outcomes in it. If instead you skipped the reference checks because you were flattered by the offer and did not look closely at the runway, the process was weak regardless of how it turned out, and that is the actual thing worth fixing before the next offer.

Why a decision journal is the only practical way to run this check

None of this works from memory, for the reason already covered: your recollection of your own reasoning has already been edited by the outcome by the time you sit down to think about it. The only way to genuinely separate the two scores is to have written the process down before you knew the result — what you expected, your rough confidence, the actual reason, and what would have changed your mind. A decision journal entry written at the moment of deciding is what you compare the outcome against later; without it, "would I make the same call again" is a question you are answering with a memory you cannot fully trust, and the whole exercise quietly turns back into grading the outcome under a different name.

What this changes about a bad week

The same split applies to a values review, not just to a single high-stakes call. A week can go badly for reasons that had nothing to do with your judgement — a system outside your control failed, someone else changed their mind, an ordinary piece of bad luck landed on an ordinary day. Reviewing that week by asking only "how did it go" tends to produce the wrong verdict: either unearned confidence when a reckless call happened to work, or unearned doubt about a principle that was never actually tested by what went wrong. Asking "given what I knew and the state I was in, was the reasoning behind this sound" is a harder question, but it is the one that tells you whether to change something or simply let a bad week be a bad week.

CreedOS keeps a check-in for each day and a review that shows the trend behind it, which is a natural place to hold the state you were in and the reasoning at the time, separately from how a particular day turned out. It is free, with no in-app purchases.