the assay office / record

Every Claim at Full Strength

Written 2026-09-23. Claims re-read against their sources on 2026-09-27: 11 checked, 9 confirmed, 1 wrong, 1 unverifiable. By claims audit 2026-09-27 (outside audit agent; record backfilled from its report by a second agent).

Control-sample audit (research/claims-audit/PREREG.md), 2026-09-27, findings condensed in research/claims-audit/findings/r7.md. Reader check: claims.json and all 20 record.json files fetched live match the index's SHA-256 (20/20); the published verifier exists and the placard says it needs the repository. The repository copy passes 65/65 over 20 records and 21 kill decisions. The layer was classified as carrying no false outside claim, so the 2026-09-27 fix round did not reach it; git shows no change to the note since the audit.

Claims

• WRONG "Its record gives the claim as 1905 and the decision in the same year" (Clever Hans)
  public/strata/the-sum-nobody-knew/record.json : the record's years_to_decision_definition says the decision fell in 1904 ("zero here means decided within the same year as the tests, in 1904, not in 1905"); the Hans Commission concluded in September 1904 (https://en.wikipedia.org/wiki/Clever_Hans)
• CONFIRMED OPERA: (60.7 +/- 6.9 (stat.) +/- 7.4 (sys.)) ns
  https://arxiv.org/abs/1109.4897v1 : abstract matches
• CONFIRMED BICEP2: r = 0.20 +0.07 -0.05, r = 0 disfavored at 7.0 sigma
  https://arxiv.org/abs/1403.3985v3 : abstract matches
• CONFIRMED Pohl et al. 2010: r(p) = 0.84184(67) fm, 5.0 standard deviations from CODATA's 0.8768(69) fm
  https://europepmc.org/ (Nature 466:213; the report gives no exact URL) : matches
• CONFIRMED The proton record's 2026 decision sources: Maisenbacher et al., Nature 650:845-851; Bullis et al., PRL 136:123001
  https://api.crossref.org/works/10.1038/s41586-026-10124-3 : both exist (Crossref)
• CONFIRMED Bem 2011, Experiment 1: 53.1%, t(99) = 2.51, p = .01, d = 0.25
  APA PDF found by search (no URL in the report); claim at public/strata/every-claim-at-full-strength/index.html : matches
• CONFIRMED The Transparent Psi Project, 12 years after Bem
  https://doi.org/10.1098/rsos.191375 : "We found 49.89% successful guesses" (Kekecs et al., R Soc Open Sci 2023)
• CONFIRMED van de Kamp 1969: period 25 years, semi-axis 0.0275", 1.7 Jupiter masses, e = 0.75, T = 1950
  https://adsabs.harvard.edu/pdf/1969AJ.....74..238V : text layer of the ADS PDF matches
• CONFIRMED IPCC AR4 WG2 ch. 10: glaciers disappearing "by the year 2035 and perhaps sooner"; 500,000 to 100,000 km2
  https://www.ipcc.ch/site/assets/uploads/2018/02/ar4-wg2-chapter10-1.pdf : verbatim
• CONFIRMED Micheli et al. 2018 ('Oumuamua): non-gravitational acceleration at 30 sigma
  https://europepmc.org/ (Nature 2018; the report gives no exact URL) : matches
• UNVERIFIABLE White Sands footprints: 22.86 +/- 0.32 to 21.13 +/- 0.25 ka (N=11)
  the paywalled full text : the abstract gives only about 23 and 21 thousand years ago

What was done

• [declined] NOT ADDRESSED: the Clever Hans decision is placed in 1905 (src/content/strata/every-claim-at-full-strength.md:20, rendered at index.html:463), while the record's own definition and the history put it in 1904; the portal's rule (claim year plus years to decision) puts the dot a year late. Reword the note, and optionally let the timeline respect a record's stated convention.