Read the original decision context before evaluating the outcome — sequence prevents hindsight bias
When reviewing a past decision, read the original context record before evaluating the outcome, because evaluating outcome first allows hindsight bias to contaminate your assessment of whether the reasoning was sound.
Why This Is a Rule
The sequence of information exposure determines which biases activate. If you learn the outcome first — "the project failed" — and then read your original decision context, your assessment of the reasoning is contaminated. The failure outcome makes every risk factor in the original context feel like an obvious warning sign: "Look, I even noted the timeline risk!" Hindsight bias converts every noted uncertainty into a missed red flag.
Reading the context first — before knowing the outcome — forces you to evaluate the reasoning on its own merits. "Given what I knew at the time, with the constraints I documented and the confidence level I recorded, was this decision well-reasoned?" This evaluation is possible only if hindsight hasn't yet colored your reading.
The distinction matters because good decisions can produce bad outcomes (bad luck) and bad decisions can produce good outcomes (good luck). Learning from decisions requires evaluating reasoning quality independent of outcome — which is impossible once you know the outcome unless you've shielded the reasoning review from outcome knowledge.
When This Fires
- During scheduled decision reviews (triggered by the review date in Record five elements at the moment of every significant decision — before hindsight rewrites it)
- In post-mortems when evaluating whether the original decision was sound
- During annual reviews of significant decisions and their outcomes
- Any retrospective evaluation of past decisions
Common Failure Mode
Glancing at the outcome first: "Oh, the project succeeded — let me see what I was thinking." The glance is enough to contaminate the review. Once you know the outcome, every element of the context record is read through outcome-colored lenses. The sequence must be enforced: context first, outcome second, with no peeking.
The Protocol
When reviewing a past decision: (1) Open the decision context record (Record five elements at the moment of every significant decision — before hindsight rewrites it) WITHOUT first checking what happened. (2) Read the five elements: decision, forces, expected consequences, confidence, review date. (3) Evaluate: given this information, was the reasoning sound? Was the confidence appropriate? Were the expected consequences reasonable? (4) ONLY AFTER evaluating the reasoning: check the actual outcome. (5) Compare: did the outcome match expectations? If not, was the deviation due to flawed reasoning or unforeseen factors? This comparison, done in this order, produces genuine learning about decision quality.