Metatron will learn more reliably when forecasts, evidence, interventions, decisions and outcomes share stable identifiers and append-only histories than when reports are stored as disconnected documents.

Counterevidence & uncertainty

A richer graph can create process cost, false precision and surveillance risk; linkage does not itself improve decisions.

What would change the reading

Pilot against document-only control workflows. Reject or simplify if the linked design does not improve audit completion, calibration, time-to-correct, decision reversibility or measured outcome learning net of operating cost.

Primary routes

1. Gneiting, T. & Raftery, A. E. (2007), “Strictly Proper Scoring Rules, Prediction, and Estimation.”https://doi.org/10.1198/016214506000001437Open source record →
2. HM Treasury & Evaluation Task Force (2026), Magenta Book: Central Government guidance on evaluationhttps://www.gov.uk/government/publications/the-magenta-book/magenta-book-central-government-guidance-on-evaluation-htmlOpen source record →
3. HM Treasury (2026), The Green Bookhttps://www.gov.uk/government/publications/the-green-book-appraisal-and-evaluation-in-central-government/the-green-book-2026Open source record →
4. W3C (2013), PROV-DM: The PROV Data Modelhttps://www.w3.org/TR/prov-dm/Open source record →
5. W3C (2013), PROV-O: The PROV Ontologyhttps://www.w3.org/TR/prov-o/Open source record →
6. UN Statistics Division, Fundamental Principles of Official Statisticshttps://unstats.un.org/fposOpen source record →
7. UN Statistics Division (2019), UN National Quality Assurance Frameworks Manualhttps://unstats.un.org/unsd/methodology/dataquality/un-nqaf/Open source record →
8. UK Statistics Authority (2025), Code of Practice for Statistics 3.0https://code.statisticsauthority.gov.uk/wp-content/uploads/2025/10/Code-of-Practice-for-Statistics-3.0.pdfOpen source record →
9. UK Statistics Authority, Standards for Official Statisticshttps://code.statisticsauthority.gov.uk/standards-of-the-code-of-practice/standards-for-official-statistics/Open source record →
10. NIST (2023), AI Risk Management Framework 1.0https://doi.org/10.6028/NIST.AI.100-1Open source record →
11. NIST AIRC, AI RMF Core / Measurehttps://airc.nist.gov/airmf-resources/airmf/5-sec-core/Open source record →
12. NIST (2024), Generative AI Profile, NIST AI 600-1https://doi.org/10.6028/NIST.AI.600-1Open source record →
13. OECD, Strategic Foresighthttps://www.oecd.org/en/about/programmes/strategic-foresight.htmlOpen source record →
14. OECD (2025), Strategic Foresight Toolkit for Resilient Public Policyhttps://doi.org/10.1787/bcdd9304-enOpen source record →
15. UK Government Office for Science (2024), Futures Toolkithttps://www.gov.uk/government/publications/futures-toolkit-for-policy-makers-and-analysts/the-futures-toolkit-htmlOpen source record →
16. UK Government Office for Science (2024), Futures Toolkit resource collectionhttps://www.gov.uk/government/collections/futures-toolkit-resources-for-government-officialsOpen source record →