The passage contrasts modest winners with self-congratulating forecasters, discusses hindsight and warns about complacency while ahead. It also connects setbacks with renewed effort.
Audit the favorable case
It is tempting to review losses carefully and leave wins unexamined. A favorable outcome may conceal a broken rule, a lucky event or a risk not taken into account. Reviewing both helps separate what should be repeated from what merely happened to work.
Worked example
A fictional method violates its definition but receives a favorable result. Counting that outcome as evidence for the original method would mix luck and a different decision.
Case connection
Earlier success is not a fresh safety test. Use the case to ask whether good outcomes have hidden a fragile process.
Challenger: review the decision chain
Source-grounded facts
The Rogers Commission found serious flaws in the launch decision process, including communication failures around safety concerns.
Context
Before Challenger’s January 1986 launch, engineers raised concerns about the solid rocket booster joints and cold conditions. The Rogers Commission later reconstructed how those concerns moved through the organisation.
Outcome
The Commission found the launch decision flawed, citing communication failures and conflict between engineering evidence and management judgment. This was an organisational decision failure, not merely a bad final outcome.
- Thiokol engineers initially recommended against launch under the forecast cold conditions.
- After internal discussion, management changed the recommendation.
- Crucial information about previous O-ring damage and the disagreement did not reach key senior launch officials.
Case analysis
Good previous outcomes can hide unresolved weaknesses in a process. Review a favourable case using the same questions used after a loss: what assumptions were made, what evidence was available and where could the process have failed? The launch review is a serious example of why a final outcome alone cannot substitute for examining how concerns were evaluated.
Try it
Pick a good outcome and identify one decision worth repeating, one uncertainty and one mistake the result concealed.