Decision quality depends on process not outcome alone

Decision quality depends on process not outcome alone

A favorable outcome can come from a weak decision. A sound decision can produce an unfavorable result. Decision quality therefore depends on what was known, assumed, considered, constrained, authorized, and documented before the outcome.

The outcome still matters. It becomes new evidence for the next decision.

A result is not the whole explanation

Uncertain decisions combine judgment with factors no one completely controls. A single result may reflect the decision, implementation, another person's action, an external event, or chance.

Treating success as proof encourages weak processes to survive. Treating failure as proof punishes reasonable risk-taking and encourages people to hide uncertainty.

The review should keep process quality and outcome quality separate long enough to understand each one.

Process quality has concrete parts

A strong process begins with a clear objective and a known decision owner. It considers credible alternatives, uses evidence appropriate to the stakes, records important assumptions, and identifies uncertainty rather than hiding it.

It also respects authority and controls. A carefully reasoned choice is not sound if it exceeds permission, ignores a required review, omits an affected stakeholder, or accepts foreseeable harm without the proper response.

Documentation matters because memory changes after the outcome is known.

Outcomes update the process

A bad outcome can reveal an omitted risk, stale input, faulty assumption, implementation gap, or poor estimate. A good outcome can reveal that a useful method deserves repetition.

It can also reveal nothing decisive by itself. One observation cannot validate a probability or distinguish luck from skill with confidence.

Learning improves when the review asks what was surprising, what was reasonably knowable, what changed after the decision, and what should be different next time.

Patterns deserve more weight than stories

People naturally remember dramatic wins and losses. A decision system should also examine repeated comparable cases.

If an outcome described as rare happens frequently, the estimate or process needs attention. If a process consistently performs well across changing conditions, that pattern is stronger evidence than one memorable success.

Good review connects the individual story to the wider record without erasing the details that made the case different.

Feedback needs consent and purpose

Analytical feedback is not appropriate in every conversation. A person may want to vent, receive support, explore options, or ask for a formal critique.

The purpose should be clear before a truth-seeking review begins. In a consequential setting, the required investigation, audit, legal, compliance, safety, clinical, employment, or regulatory process takes precedence over an informal conversation.

Connections

E015 develops this thesis through Annie Duke's Thinking in Bets, Dalton's personal examples, and the distinction between support and truth-seeking. [[How to Review a Decision Without Letting the Outcome Rewrite It]] provides the working procedure.

Use this note as a general learning principle. It does not replace a validated model, required control, professional judgment, or accountable authority in a specific domain.