PsychologicalDUAL-USE

Outcome Bias

What it is

Judging the quality of a decision by how it turned out rather than by whether it was sound given what was known when it was made.

How it works

Baron and Hershey (1988) showed people the same medical decision — surgery with stated odds — and found it rated as a better decision, and the surgeon as more competent, when the patient happened to survive. Unlike hindsight bias, which distorts what we think we knew, outcome bias operates even when the probabilities are stated plainly: a good result feels like proof of a good choice and a bad result like proof of a bad one. The lever is that outcomes are concrete and process is abstract; results are easy to count and reasoning is not. Poker players call the error “resulting”. Persuaders exploit it by showing you winners. Fund managers advertise the years they beat the market, gurus present the students who got rich, and pump-and-dump promoters circulate screenshots of gains — all inviting you to infer a good process from a good outcome without seeing the denominator. Politically, incumbents are rewarded and punished for economic timing they did not control, and reckless decisions that happened to work are retold as bold leadership.

Real-world examples

  • •Baron and Hershey (1988): identical surgical decisions with identical stated risks were judged better and the physician more competent when the operation succeeded.
  • •Annie Duke's account of a poker player who folds a strong hand correctly, loses to a lucky draw, and is told he “played it wrong” — the everyday form of resulting.
  • •Cryptocurrency and meme-stock promoters post gains screenshots as evidence of skill; the far larger number of identical bets that lost is not posted.
  • •Presidents of both parties have been credited or blamed for the business cycle — Carter for 1970s stagflation, Clinton for the 1990s boom — largely on timing rather than on any evaluated decision.
  • •A company promotes the manager whose gamble paid off and fires the one whose careful decision lost to bad luck, teaching the whole organization to gamble.

Ethical guidelines

Where the line is

Citing a good outcome is honest when the audience can also see how many similar decisions were made, how many failed, and what the decision-maker actually knew; it becomes manipulation when a favorable result is offered as proof of a sound process while the losers and the odds are hidden.

  • ●When presenting results, present the process and the base rate too: how many tried, how many failed, and what the decision looked like before the outcome was known.
  • ●Do not market a lucky outcome as evidence of skill; if you cannot show a track record across many decisions, say the sample is one.
  • ●Judge and reward people you manage on the quality of their reasoning given the information available, not only on results.
  • ●Using a success story to illustrate a genuinely repeatable method is legitimate; using it to imply repeatability that the data do not show is not.

How to defend against it

  • ►Separate the two questions explicitly: was the decision reasonable given what was known, and did it work out? Answer the first before you look at the second.
  • ►Ask for the denominator. For any success story, ask how many people made the same bet and what happened to the others.
  • ►Look for a track record of many decisions with confidence levels, not a highlight reel; ask for the misses.
  • ►When you evaluate your own past choices, write down what you knew at the time and grade the process — otherwise every lucky win teaches you to take the same risk again.
  • ►Treat “it worked for me” as a report of one outcome, not a demonstration of a method.

From the Defense Playbook

Decide Metrics Before Testingminutes

Before an A/B test or pilot starts, write down the primary metric, the harm metrics that must not get worse, the sample size or stopping rule, and the planned analysis, so that the result can tell you something you did not already want to hear.

Every playbook entry states how strong its evidence is and when not to use it. Browse the full playbook.

References

  1. Baron, J., & Hershey, J. C. (1988). Outcome bias in decision evaluation. Journal of Personality and Social Psychology, 54(4), 569-579 · link
    The founding study: identical decisions judged better when the outcome was good, even with probabilities stated.
  2. Duke, A. (2018). Thinking in Bets: Making Smarter Decisions When You Don't Have All the Facts. Portfolio / Penguin
    The concept of “resulting” and the practice of grading decisions separately from outcomes.
  3. Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux
    Discussion of outcome and hindsight bias in evaluating leaders, physicians, and financial decisions.
Last reviewed
Suggest a correction

Detect Outcome Bias in any text

Paste any message, email, or article into our free Manipulation Detector to see if Outcome Bias or other techniques are being used on you.

Try the Detector FreeCreate Free Account