JUDGMENT CARD / CALLABLE DECISION RULE
A good outcome does not automatically prove a high-quality decision
A good outcome is important evidence for a decision, but it does not automatically prove original decision quality. Review must separate judgment on then-available information, execution, external cycles and luck.
WHEN TO INVOKE
First test whether the card applies
Explaining the source of excess profit
Assessing a management or investment team
Designing incentives, review and the next decision
REQUIRED INPUTS
Without these inputs, do not return a Howard judgment
Decision V1 and facts available at the time
Key assumptions, alternatives and reasons for rejection
Execution and external changes
Final outcome, realisation and unresolved tail risk
DECISION LOGIC
Move from facts to a reviewable conclusion
Reconstruct the choice using then-available information rather than hindsight.
Decompose contributions from judgment, execution, cycle and luck.
Test sustainability and unresolved tail risk behind excess returns.
Update capability assessment with repeated outcomes and counterexamples.
ACTION OUTPUT
An attribution matrix, reusable capabilities, non-repeatable conditions, unresolved risk and revised next-decision rules.
RELATIONSHIPS
COUNTEREXAMPLES & REOPENING
A good outcome remains evidence; repeated success across cycles strengthens the capability assessment. Reopen on exit, tail-risk resolution or a new counterexample.
PRIMARY PUBLIC EVIDENCE
PROVENANCE
H: Howard's historical public position; structured by AI, not originated by AI.