Good Outcome. Bad Decision? Why Results Cannot Validate Judgment Alone

Carmen Fries · 2026-10-16

The investment succeeded. Revenue rose, the asset recovered or the venture finally reached profitability. It is natural to look backward and say: the decision was right.

That conclusion may be premature. A favorable result can follow a poorly founded decision. An unfavorable result can follow a careful and reasonable one. Judgment quality must be evaluated using what the decision-maker could reasonably know at the time, not merely what became visible afterward.

The outcome bias problem

Psychological research by Jonathan Baron and John C. Hershey (1988) documented outcome bias: evaluators allowed the eventual result of a decision to influence their assessment of the quality of the decision process, even when judging information available at the time. This does not mean outcomes are irrelevant. They matter enormously for learning and performance. But they are not a substitute for evaluating the original reasoning.

In investment settings, hindsight can be particularly seductive. A company that later prospers may make earlier red flags seem trivial. A position that later fails may make a defensible decision look reckless. Neither judgment is reliable without reconstructing the original decision context.

Reconstruct the decision before the outcome

Consider an illustrative investment committee that continued to fund a struggling venture. Later, the venture succeeded. To assess the original decision, a review would have to establish:

- Which forecasts, operating evidence and downside scenarios were available at the funding date? - Which uncertainties were acknowledged and which were ignored? - Were alternative uses of capital seriously considered? - Who could challenge the proposal, and was dissent recorded? - What would have triggered a stop, review or change in course?

A useful diagnostic exercise is to examine dated records before reviewing the eventual outcome, where possible. This helps reduce the risk of rewriting past reasoning to fit known results. It cannot remove hindsight completely; missing records must be recorded as such.

Persistence or a failure loop?

Persisting after setbacks is not inherently irrational. A venture can fail at execution while the underlying opportunity remains viable. But repeated funding can also reflect sunk-cost attachment, informal loyalty or concentrated decision power.

The crucial distinction is evidentiary: did new information improve the investment thesis, or were the same assumptions repeated despite contrary evidence? What would have falsified confidence? What were the explicit limits of continued support?

Those are diagnostic questions, not verdicts about the competence or motives of a particular investor.

Better future decisions

Reality Recognition can examine discrepancies between a stated investment discipline and the way decisions were actually made under pressure. It can also identify where the historical record is too incomplete to support a finding.

The object is not to declare that a successful investment was secretly wrong. It is to protect future judgment from being confused with past luck.

A good result deserves attention. A good decision deserves its own evidence.

SOURCES

Baron, J. & Hershey, J. C. (1988), “Outcome bias in decision evaluation,” Journal of Personality and Social Psychology, 54(4), 569–579, DOI: 10.1037/0022-3514.54.4.569
Reality Recognition, Methodology & Operating Manual, v1.5, §§8–10, 20, 26, 34.6. This article discusses established outcome-bias research and RRG's proposed diagnostic application
it is not an efficacy claim.