Naming Good Thinking Before Outcomes Appear

Category: Safety Enables Signal Calling out decision quality early prevents outcome bias from distorting learning.

Two people on the same finance team make two bets in the same quarter.

The first, a treasury analyst, is sitting on a large foreign-currency exposure that will land in three months. She does the work. She lays out the range of plausible moves, prices the cost of hedging against the cost of being caught wrong, weighs the firm’s actual tolerance for a bad month, takes a second opinion, and recommends hedging most of the exposure. She writes down what she’s uncertain about and what would change her mind. The recommendation is taken. Then the currency moves the rare way, the unhedged tail goes against everyone, and the hedge — which was the right decision — happens to cost more than doing nothing would have, because the disaster it insured against didn’t arrive. On paper, she lost money.

The second, a colleague, has a similar exposure and does almost none of this. He has a feeling. He leaves the whole thing open because hedging is “expensive” and he’s “got a sense” the rate will hold. He writes nothing down. The rate holds. He’s up.

Three months later there is a review. And if the review judges these two bets by how they turned out — which is the natural, gravitational thing to do — it will praise the gambler and quietly mark down the analyst. The careful one lost; the reckless one won; the numbers say what the numbers say. Everyone watching will draw the obvious conclusion about which behaviour the firm rewards, and they will be right to.

In a small number of environments, that is not what happens. In those, somebody named the quality of the analyst’s thinking the day she made the call — before the currency moved, before anyone knew — and put it on record. So when the review arrives, there is already a fixed point that the outcome cannot move: this was a well-made decision. The loss is discussed honestly, but it does not get to rewrite the verdict on the judgement, because the verdict was written first. And the gambler’s win is enjoyed, and then gently examined, and found to be what it was.

That sequence — praising the thinking before the result is known — is not a nicety. It is one of the few things that stops an organisation teaching itself exactly the wrong lesson.

The Principle

Outcome bias makes us judge a decision by how it turned out rather than how well it was made — so good thinking that gets unlucky is punished and reckless thinking that gets lucky is praised. The correction is to name and record decision quality at the moment it happens, before the outcome is known, so that learning is anchored to the judgement and not hostage to the luck.

The deep error is collapsing two different things into one. A decision is the thinking done with the information available at the time. An outcome is what the world then does, much of which was never in anyone’s control. These come apart constantly, because luck is real: a good decision can lose and a bad one can win, and over any short run they frequently do. Yet once the outcome is known, it floods backward over the decision and colours it entirely — the winning call looks wise and the losing call looks foolish, regardless of how either was actually made. Naming the thinking early is simply the act of fixing the decision’s quality in place before the outcome arrives to distort it. You are taking the one assessment that can be made cleanly — was this well-reasoned? — and making it while it can still be made cleanly.

Why It Is Inevitable

This isn’t an optional refinement that thoughtful organisations happen to add. It’s a correction they’re forced into, because without it the default is not neutral — it actively teaches the wrong thing, and it does so with confidence.

Outcome bias is inevitable because the outcome is loud and the reasoning is quiet. When the review happens, the result is sitting right there — a number, a shipped thing, a loss, a win. The quality of the thinking that produced it is invisible by then; it has to be reconstructed, and reconstruction always happens in the glare of knowing how it ended. Hindsight makes the loser’s reasoning look full of holes that were not visible at the time and makes the winner’s luck look like foresight. Nobody has to be lazy or unfair for this to happen. It is the natural optical effect of standing downstream of an outcome and looking back.

It compounds because outcomes are what get measured and remembered. The currency bet shows up in the figures; the careful analysis behind it does not. The win is celebrated at the meeting; the reasoning that should have been celebrated was never recorded, so there is nothing to celebrate it with. An organisation that only ever has outcomes in front of it will, with complete sincerity, reward and punish on outcomes — because that is the only data it kept.

And it is self-reinforcing, which is the dangerous part. Once people see that good losing decisions get marked down and lucky reckless ones get praised, they learn the real lesson: don’t be the careful one who can be blamed when the dice go wrong. Stop writing down your uncertainty, because it’s only ever used against you. Take the punt that will look brilliant if it lands and shrug if it doesn’t. The bias doesn’t just misjudge the past; it reshapes the future toward exactly the thinking you least want. Any environment that wants good judgement to survive contact with bad luck has to install something that holds the line — and praising the thinking early is the cheapest thing that does.

How It Shows Up

• The quality of a decision is named and recorded at the time it’s made — “this is a well-reasoned call, whatever happens” — and not left to be inferred later from the result.
• Reviews open with what was knowable when the decision was made, before anyone is allowed to talk about how it turned out.
• A loss that followed good reasoning is described as exactly that — a good decision that got unlucky — and the reasoning is still held up as the standard.
• A win that followed poor reasoning is enjoyed and then honestly examined, and the luck in it is named rather than relabelled as skill.
• People are praised, on the spot, for flagging their own uncertainty, taking a second opinion, or writing down what would change their mind — the visible marks of good thinking — independent of where the bet lands.
• “That was a good decision” and “that worked out” are kept as two separate sentences, and everyone knows they can come apart.

Why It Causes Benefit

When decision quality is named before the outcome is known, an organisation gets something quietly rare: it can learn from results without being deceived by them.

The first benefit is that good judgement stops being punished out of existence. The analyst who reasoned well and lost keeps her standing, because her standing was fixed to the quality of her thinking, not to the toss of the currency. Everyone watching learns the lesson you actually want them to learn — that careful, honest, well-hedged reasoning is what gets rewarded here, even on the quarters it loses. That is the only message that keeps people reasoning carefully when the stakes are real and the luck is uncertain, which is precisely when you most need them to.

The second is that luck stops being mistaken for skill. The gambler’s win, examined honestly, is recorded as a win that good process did not produce — so nobody copies the process, because there wasn’t one. This matters more than it first appears. The most expensive failures in many organisations are committed by people repeating a reckless move that happened to work once and was applauded as brilliance. Naming the thinking early is what lets you applaud the right thing and decline to applaud the wrong one, even when the wrong one made money.

And the third benefit is the compounding one: the signals that let an organisation correct itself stay switched on. Because flagging uncertainty and writing down your reasoning are praised in the moment rather than weaponised in hindsight, people keep doing them. The uncertainty stays visible, the assumptions stay on record, the honest “I’m not sure, here’s why” keeps getting said. An organisation that only judges outcomes trains everyone to hide the very reasoning it would need to improve. One that names good thinking early keeps that reasoning in the open, which means it can actually tell, next time, a good decision from a lucky one — and that ability is the whole basis of getting better rather than merely getting outcomes.

How To Cultivate It

• Record the verdict on the thinking before the outcome lands. The decisive move is timing: a note made the day the call is taken — “well-reasoned, here’s why, the uncertainty is X” — is a fixed point the result cannot move. The same judgement attempted after the result is already contaminated by it.
• Open reviews with what was knowable at the time, and hold the outcome back until that’s been established. If the first thing on the table is how it turned out, every reconstruction of the reasoning will bend to fit it. Reconstruct the decision in the light it was actually made in.
• Say the two sentences separately and out loud: “this was a good decision” and “this worked out.” Keep them apart on purpose, especially when they disagree — a good decision that lost, and a bad decision that won, are the two cases the whole practice exists to handle.
• Praise the visible marks of good thinking — flagged uncertainty, a second opinion sought, an assumption written down, a tail risk hedged — on the spot and regardless of result. These are the things you can see at the time and the things you actually want more of.
• When a reckless bet wins, name the luck. Enjoy the result, then say plainly that the process wasn’t sound and shouldn’t be repeated. Declining to canonise a lucky win is as important as protecting an unlucky good call, and far harder, because the money is real and it feels ungrateful.
• Make all of this safe. None of it survives without psychological safety, because praising good thinking that lost, and questioning bad thinking that won, both require people to be honest about decisions in front of others. If honesty about a call costs you standing, everyone reverts to defending the outcome — and the practice quietly dies.

What Good Looks Like

An environment where the quality of a decision is named when the decision is made, and that naming holds even after the world has had its say. Where a loss that came from good reasoning leaves the reasoning’s reputation intact, and a win that came from a reckless punt is enjoyed but never mistaken for skill. Where reviews begin with what was knowable at the time and only then turn to the result, so hindsight is kept from quietly rewriting the past. Where “good decision” and “good outcome” are two separate assessments that everyone knows can disagree — and the disagreement, when it comes, is treated as the most interesting thing in the room rather than an embarrassment to be smoothed over.

In a place like that, people keep flagging their uncertainty and writing down their reasoning, because doing so is what gets praised and never what gets them blamed. The careful analyst stays careful. The gambler’s luck is seen for what it is. Decision quality and outcome quality are tracked as the two different things they are, which means the organisation can be genuinely lucky and genuinely unlucky without either fooling it about how well it is actually thinking. Over time it gets better at the only thing that compounds — making good calls — because it never let a run of outcomes talk it out of recognising them. That is not optimism about luck. It is simply what learning looks like once it has been protected from the outcome that would otherwise distort it.

A Reflective Question

Think of the last decision in your environment that got praised or criticised. Was the judgement actually about how well the call was made with what was knowable at the time — or was it about how the call happened to turn out? And if those two came apart, which one did your environment have the safety to say out loud?