Outcome Bias

What outcome bias means

Outcome bias is the tendency to evaluate the quality of a decision based on its result rather than on the information and reasoning available when the decision was made. If things turn out well, we assume the decision was good. If things turn out badly, we assume the decision was poor. The actual logic behind the choice - whether it was well-reasoned, well-informed, and appropriate given the uncertainty at the time - gets overwritten by what happened next.

This bias was formally described by psychologists Jonathan Baron and John Hershey in their 1988 research on decision evaluation, which showed that people consistently rated identical decisions differently depending on whether the outcome was positive or negative. The decision itself hadn’t changed. Only the result had. And yet the judgement of the decision-maker shifted dramatically.

Outcome bias matters because it confuses luck with skill, punishes good reasoning that encounters bad luck, and rewards reckless decisions that happen to succeed. Over time, it distorts how organisations learn, how individuals are evaluated, and how societies decide what works.

How outcome bias works

Collapsing process into result

When we evaluate a decision after the fact, we naturally have access to information that wasn’t available at the time. We know what happened. The challenge is to set that knowledge aside and judge the decision using only what the decision-maker knew when they made it. This is extraordinarily difficult for the human brain to do.

Hindsight bias compounds the problem. Once we know the outcome, it feels like it should have been predictable all along. If the outcome was bad, the decision-maker should have seen it coming. If the outcome was good, they obviously made the right call. The uncertainty that made the decision genuinely difficult at the time becomes invisible in retrospect.

The result is that we collapse the entire messy, uncertain process of decision-making into a single data point: did it work or didn’t it? This is like judging a poker player entirely by whether they won a single hand, ignoring whether their bet was mathematically sound.

Why good decisions can fail

Every decision made under uncertainty carries a distribution of possible outcomes. A well-reasoned decision is one that gives you the best probability of a good result given what you know - but probability is not certainty. A doctor who correctly prescribes the treatment most likely to work will still see some patients for whom it doesn’t. A fund manager who makes a well-researched investment will still sometimes lose money. A general who devises a sound strategy will still sometimes lose the battle.

Outcome bias ignores this entirely. It treats the single realised outcome as definitive evidence of the decision’s quality. The doctor whose patient deteriorated is judged more harshly than the doctor whose patient recovered, even if both made identical clinical decisions. The only difference is luck - but outcome bias reads luck as competence.

The reverse problem: bad decisions that succeed

Equally dangerous is the way outcome bias rewards poor reasoning that happens to produce good results. A business that takes a reckless gamble and wins is celebrated for its boldness. A driver who overtakes dangerously and arrives safely is seen as skilled. A politician whose ill-considered policy happens to coincide with an economic upturn is credited with wisdom.

This creates terrible incentive structures. If outcomes are all that matter, then the rational strategy is to take big risks - because if they pay off, you’ll be rewarded, and if they don’t, well, the same outcome bias that would have praised you will find a scapegoat elsewhere. The connection between this and the self-serving bias is direct: we take credit for successes and blame circumstances for failures, and outcome bias ensures that onlookers do the same.

Outcome bias in everyday life

In the workplace

Outcome bias is one of the most destructive forces in organisational life. Performance reviews overwhelmingly focus on results rather than process. The salesperson who landed a major deal through a lucky referral is promoted. The salesperson who built a rigorous pipeline and followed best practices but hit a quiet quarter is put on a performance plan. Neither outcome tells you much about who is the better salesperson. But outcome bias ensures that only one of them is rewarded.

This discourages the very behaviours organisations claim to value: careful analysis, risk management, honest uncertainty, and learning from near-misses. If only results count, then people learn to optimise for visible wins rather than sound reasoning. They take credit for good luck and hide bad luck. They avoid decisions with uncertain outcomes, even when those decisions are clearly the right ones to make.

In medicine

Medical malpractice law is heavily influenced by outcome bias. Juries judge clinical decisions by whether the patient recovered, not by whether the decision was consistent with best practice at the time. A surgeon who follows correct procedure but encounters a rare complication is more likely to be found negligent than a surgeon who deviates from protocol but whose patient recovers. The outcome becomes the evidence, even when it shouldn’t be.

This has real consequences. Doctors practise “defensive medicine” - ordering unnecessary tests, avoiding risky but beneficial procedures, and making conservative choices designed to protect them from blame rather than to optimise patient care. Outcome bias in the courtroom shapes medical practice on the ward.

In sport

Sports analysis is saturated with outcome bias. A football manager who substitutes a player in the 70th minute is a genius if the replacement scores and a fool if the team concedes. The substitution was either reasonable or unreasonable at the time it was made - the subsequent events don’t change that. But commentary, punditry, and fan reaction are almost entirely outcome-driven.

This extends to player evaluation. A striker who misses an open goal is criticised for the miss. A striker who scores from an improbable angle is praised for brilliance. The actual difficulty and decision-making involved in each situation is secondary to what happened. Survivorship bias amplifies this: we study the tactics of winning teams and assume they worked because they won, ignoring losing teams that used identical approaches.

In public policy

Governments and voters routinely evaluate policies by their outcomes rather than by the quality of the reasoning behind them. A public health campaign that coincides with a decline in illness is deemed successful, even if the decline was caused by unrelated factors. An economic policy introduced before a recession is blamed for the downturn, even if the recession was global and inevitable.

This makes evidence-based policymaking extremely difficult. If voters judge leaders by outcomes they can’t control, then leaders have little incentive to make well-reasoned decisions and every incentive to make popular ones - or to time announcements to coincide with favourable trends they had nothing to do with.

How to counter outcome bias

Separate the decision from the outcome

The core discipline is to evaluate decisions at the point they were made, using only the information available at that time. Ask: given what they knew, was this a reasonable choice? Was the reasoning sound? Were the risks properly weighed? If the answer is yes, then a bad outcome doesn’t make it a bad decision. It makes it an unlucky one.

Think in probabilities

Probabilistic thinking is the natural antidote to outcome bias. If you understand that a well-reasoned decision gives you a 70% chance of a good outcome, then you also understand that 30% of the time it won’t work - and that doesn’t mean the decision was wrong. Judging decisions across many instances rather than by single outcomes reveals their true quality.

Learn from process, not results

The best organisations and individuals build feedback loops around process rather than outcomes. Was the analysis thorough? Were the assumptions tested? Were alternative scenarios considered? Were risks identified and managed? These questions produce genuine learning. “Did it work?” produces only hindsight narratives.

Reward good reasoning explicitly

If you manage people, evaluate their decision-making process alongside their results. Praise a well-reasoned decision that had a bad outcome. Question a poorly reasoned decision that had a good one. Over time, this builds a culture that values thinking over luck - which, counterintuitively, produces better outcomes in the long run.

Outcome bias is difficult to eliminate because our brains are wired to learn from consequences. That instinct served us well in simpler environments. But in a world of complex, uncertain, interconnected decisions, the result is often a poor guide to the reasoning - and confusing the two leads us to reward the wrong behaviours and punish the right ones.