Research and evidence
| Area | Saved record | Meaning |
|---|---|---|
| Original source searches | 24 languages | Saved query and candidate receipts |
| Source traversal | 334 attempted; 189 opened; 145 failed | Failures remain recorded |
| Retained stories | 24 across 21 languages; 23 non-English | Direct original-source records |
| Research areas with retained stories | 34 of 75 | Empty areas disclosed |
| Historical price windows | 172 | Only complete saved observations |
| New predictions | 1 | Actual issuance records; separate experimental cohort |
What is still unknown
Opening a source does not make every research area complete. Retained stories cover 34 of 75 areas. Sources that failed to open and languages without a retained qualifying story remain visible in the source receipts.
3 older decisions lack an original endpoint method. Current Long Term reviews are partial where required traits remain unproven. Experimental calls do not establish a reliable forecasting edge.
Follow the complete record
- All retained headlines and original publishers
- Actual market charts, story markers and original predictions
- Every saved prediction and its original review evidence
- Dated Long Term item reviews
- Original independent research receipts
- Complete route, image, runtime and dataset hashes
Two identical complete builds establish repeatability of the saved outputs. Their visual and functional reviews establish whether the pages meet the publication contract.