Standard 2.2 (draft): scores for this category will appear once its release checks are complete. Any scores shown above this section were calculated under an earlier version of the Standard and will be replaced.
Cocoa supply chain: how the Throne Standard is calculated
Pack: Cocoa 1.0.0 · Standard 2.2.0 (draft) · Engine: Pillar · Status: Experimental, scores not yet published
Cocoa is a separate supply-chain layer covering traceability to farm, child-labour monitoring and remediation, living-income premiums, deforestation compliance and the certification chain.
Criteria, weights and evidence levels| Criterion | Weight | Level 2 | Level 3 | Level 4 |
|---|
| Traceability to farm | 25 | to be written | to be written | to be written |
|---|
| Child-labour monitoring and remediation | 25 | to be written | to be written | to be written |
|---|
| Living-income premium | 20 | to be written | to be written | to be written |
|---|
| Deforestation compliance | 20 | to be written | to be written | to be written |
|---|
| Certification chain | 10 | to be written | to be written | to be written |
|---|
Level 0 is dated contrary evidence, or a source we read that shows the practice is absent. Level 1 is a claim with no supporting evidence. A criterion whose source was not found, not read or blocked is unknown, which is different from 0. A certificate is checked for holder, scope, validity and product coverage; "free from" wording is level 1 until a test report or scoped certificate backs it.
The calculation
V = 25 × S ÷ W, where S = sum of (weight × level) and W = sum of evidenced weights
- Each level earns a quarter of the criterion's credit per step: level 4 is full credit, level 1 a quarter.
- Coverage C = 100 × W ÷ total applicable weight. Below 70% the item is unresolved; below 85% it is capped at 89.99 (3/5).
- If less than half of the total applicable weight is at level 2 or above, the value is capped at 59.99 (2/5), so a page of claims cannot score well.
- Key criterion: child-labour monitoring and remediation. A 5/5 needs it at level 3 or above, otherwise the value is capped at 99.99.
- Exact integer band tests: 5/5 when S = 4W; 4/5 when 5S ≥ 18W; 3/5 when 5S ≥ 12W; 2/5 when 5S ≥ 8W; otherwise 1/5.
Bands (the same in every category)| Value | Band |
|---|
| 100 | 5/5 |
|---|
| 90 to below 100 | 4/5 |
|---|
| 60 to below 90 | 3/5 |
|---|
| 40 to below 60 | 2/5 |
|---|
| 0 to below 40 | 1/5 |
|---|
All arithmetic is exact (fractions, never rounded floating point). Nothing is rounded before the band test, and each band includes its lower bound. Values are shown cut to two decimals, never rounded up, so 89.996 shows as 89.99 and is 3/5.
Worked example: illustration
No cocoa supply chain item has been scored yet, so these evidence levels are illustrative. Every number below is computed by the scoring engine. The evidence levels for this category are still being written, so no cocoa supply chain can be scored until they are published.
Illustration: a chocolate brand| Criterion | Weight | Evidence level | Weight × level |
|---|
| Traceability to farm | 25 | 3 | 25 × 3 = 75 |
|---|
| Child-labour monitoring and remediation | 25 | 3 | 25 × 3 = 75 |
|---|
| Living-income premium | 20 | 2 | 20 × 2 = 40 |
|---|
| Deforestation compliance | 20 | 3 | 20 × 3 = 60 |
|---|
| Certification chain | 10 | 2 | 10 × 2 = 20 |
|---|
Calculation| Step | Result |
|---|
| Evidenced weight W | 100 |
|---|
| Weighted sum S | 270 |
|---|
| Coverage C = 100 × W ÷ applicable weight | 100 × 100 ÷ 100 = 100.00% |
|---|
| Value V = 25 × S ÷ W | 25 × 270 ÷ 100 = 67.50 |
|---|
| Result | value 67.50, band 3/5 |
|---|
Results
Each brand gets its own result. Documented findings (for example unremediated forced labour, or captive wildlife as a core activity) cap the final band after the calculation, and are always disclosed. Unknown is never a fail.
Evidence we use
Supply-chain disclosures, monitoring reports, premium payments and certification records. Every record keeps its source address, exact locator, capture date and a SHA-256 hash of what we read. Evidence older than two years is re-checked before it supports a high level.
Why scores are not shown yet
Before any score in this category is published, the evidence levels 2 to 4 for each criterion are not yet written; no verified product has yet been confirmed at 5/5 with published evidence; the calibration run has not been reviewed and signed off; we have not yet shown that scores separate stronger from weaker products. Until then we show the evidence we have found, not a score.