Evidence before narrative
Data, sources and observations are separated from inference and recommendation.
Documented methodology
A method for converting incomplete information into traceable, conditional and reviewable decisions.
Principles
Data, sources and observations are separated from inference and recommendation.
A complete review may conclude that an option is weak; scoring does not reward form filling.
Scores from 1–5 are tied to observable descriptions to reduce arbitrariness.
Material recommendations state limits and conditions for start, withdrawal or review.
Actions close with an owner, date and acceptance criterion.
The greater the irreversibility, exposure or uncertainty, the stronger the independent verification required.
Process
Define the question, scope, sources and uncertainty.
Compare alternatives using explicit criteria and real constraints.
Set the option, thresholds, conditions and exit alternative.
Convert the decision into owners, milestones and control.
Clarify authority, conflicts, information and continuity.
Test outcomes and adapt when the evidence changes.
Scoring model
| Element | Function | Limitation |
|---|---|---|
| 1–5 | Forces a distinction between clearly unfavourable, intermediate and favourable conditions. | It does not eliminate judgement; it requires justification. |
| 0–100 | Makes dimensions and alternatives comparable. | It is not a scientific probability. |
| Completeness | Measures whether the review contains evidence, scores, conditions and ownership. | It does not indicate that the decision is good. |
| Guidance | Translates confirmed data into conditional guidance. | It requires specialist review where appropriate. |
When to use it
Suitable for structuring comparisons, documenting assumptions, detecting gaps and coordinating owners.
It must not replace audit, independent valuation, or legal, tax, medical, technical or regulatory advice.