How the score is computed
- Exposure
- 20% of known substantive inputs. Public observations retain their rubric range, evidence references, and as-of time; missing is unknown, never 50.
- Adaptation
- 20% of known substantive inputs. Public observations retain their rubric range, evidence references, and as-of time; missing is unknown, never 50.
- Monetization
- 15% of known substantive inputs. Public observations retain their rubric range, evidence references, and as-of time; missing is unknown, never 50.
- Defensibility
- 15% of known substantive inputs. Public observations retain their rubric range, evidence references, and as-of time; missing is unknown, never 50.
- Management execution
- 15% of known substantive inputs. Public observations retain their rubric range, evidence references, and as-of time; missing is unknown, never 50.
- Financial resilience
- 15% of known substantive inputs. Public observations retain their rubric range, evidence references, and as-of time; missing is unknown, never 50.
- Evidence strength
- Moves the substantive estimate toward 50 when evidence is sparse or weak; it does not add durability points.
- Publication gates
- An aggregate requires at least 70% weighted coverage, four known substantive dimensions, exposure and adaptation, and evidence strength of at least 40.
- Sensitivity
- Publishes the range supported by adjacent rubric anchors, evidence-strength bounds, sector fallback, and one-at-a-time dimension changes.
Missing dimensions remain unknown and can withhold the aggregate. They are never filled with zero or 50.
Known biases
- Coverage
- Ratings reflect sourced public evidence we track; private or unavailable activity may be absent.
- Survivorship
- Companies and outcomes that stop generating evidence can be underrepresented.
- Publication
- Only policy-qualified public evidence enters a public profile, which favors verifiable claims.
- Sector classification
- Sector context depends on a versioned cohort and falls back to absolute evidence when the cohort is thin.
- Financial comparability
- Reported financial periods, currencies, restatements, and accounting choices can limit comparison.
Changelog
Every published score names its methodology version. A version bump appears here — silent re-scoring is the credibility-killer.
cw-v2.2 current 2026-07
Exposure reads the structural capability relation, not sentiment alone.
- Substitution and margin-compression assessments now lower exposure below what impact direction alone can express.
- A helping relation raises nothing on its own: positive positioning still has to be earned by impacts.
- An unresolved relation widens the published exposure range instead of asserting a neutral score.
- Ratings computed under cw-v2.1 are retained unchanged; the two versions are not comparable.
cw-v2.1 superseded 2026-07
Evidence-weighted multidimensional durability profile with explicit abstention and sensitivity.
- Separated exposure, adaptation, monetization, defensibility, management execution, and financial resilience.
- Made missing inputs unknown and withheld the aggregate unless coverage gates pass.
- Added bounded sector context, evidence-strength attenuation, and public sensitivity ranges.
- Scored exposure from impact direction only, so displacement and mild bad news landed alike.
cw-v1 superseded 2026-07
Legacy scalar impact rollup retained for reproducibility and comparison.
- Combined live impact direction, magnitude, and confidence around a neutral baseline of 50.
- Included a founded-before-2008 proxy that cw-v2 removes.
FAQ
- Is the rating a prediction?
- No. It is a speculative assessment of sourced evidence, not a forecast, valuation, or investment recommendation.
- Why can a profile have no overall number?
- Century Worth withholds the aggregate when coverage or evidence gates fail. Known dimensions remain visible without filling missing inputs with 50.
- Why show a range?
- Rubric anchors are judgments, not physical measurements. The range shows how adjacent defensible inputs and sector context could change the aggregate.
- What happened to cw-v1?
- It remains stored and reproducible as a legacy comparison. cw-v2 never overwrites or relabels it.
- Can I rely on this for investing?
- No. Century Worth records correlation and assessment from public evidence; it is not financial advice or a buy, sell, or hold signal.