Preregistrations, results and audits, as written.
A claim here is registered before the data is touched: hypotheses, measures and thresholds fixed in a dated document. It is then run as written and the result is published whether or not it survived. The audits review figures the platform had already published and record a verdict on each, including the ones withdrawn. 10 documents sets so far: 2 run: refuted, 2 run: not supported, 1 registered, not run, 5 audit.
Does the sovereign-default tilt replicate on an independent sample of defaults?
Refuted on this sample: primary ratio 0.99, p = 0.59; no row in the registered direction.
Are conflict onsets more frequent on the descending limbs of the roughly eighty-year Gleissberg envelope?
Refuted at this scale on the Brecke catalog: onsets per year are lower on descending limbs than ascending (ratio 0.91, bootstrap 95% interval 0.83 to 1.00).
Do financial and political crises cluster in the declining phase of the eleven-year solar cycle?
Under section 6 of the registration, H1 is not supported.
Do sovereign-default onset years tilt toward the declining solar phase?
H1 not supported on the primary test: the tilt is present in direction and size but does not reach the declared threshold (p = 0.053).
Does the harmonic vocabulary of charting popular music narrow over the period, measured with rarefaction so that distinct-item counts do not rise with sample size?
Registered 2026-09-27 before any chord annotation was read. Not yet run; revisions require a new dated document.
Which of the published FRED correlations survive when the series are differenced instead of correlated on levels?
Several published correlations collapse on changes; the home-prices-to-confidence figure goes from -0.89 on levels to -0.04 on changes (p = 0.66). Supersedes the FRED table in the 2025 validation log.
Do the correlations the platform once stated hold against public data from FRED, NOAA and USGS?
Solar activity to VIX: claimed 0.68, measured 0.075 (p = 0.118), refuted. Tectonic to gold: claimed 0.95, measured 0.131, refuted. Solar cycles to the dollar index: claimed about 0.70, measured 0.234, refuted as weak.
What does the archive actually hold for the figures most often repeated from it?
Verdict table per figure; the solar-to-VIX 0.68 is recorded as tested twice and refuted, with the test files named.
Were the encyclopedia cards accurate, and did every entry have the image and sources it claimed?
All 251 card hooks read; 42 corrected and 6 verified against live sources; 34 files that were saved error pages found and replaced; the 20 flagged claims verified with sources on 2026-09-18.
What did four models with different training say about the same encyclopedia claims, and where did they disagree?
A dated record of four models’ verdicts from 2026-02-11, kept authoritative as a record; the consensus-row corrections were checked by hand before being applied.