Prediction records
Twelve prediction records stood here. Eleven are withdrawn pending audit. The twelfth, pred-3, was retired on 6 September 2026 because the result it cited predates it; the corrections register carries the entry.
Each record carried a public date of 1 October 2024 with no immutable source for that date, and several cited evidence that post-dates the forecast it was said to support. A record whose own date cannot be checked independently is not a prediction, whatever it is labelled.
They are withdrawn rather than deleted. The withdrawal is the finding. Deleting the page would remove the correction while leaving the original claim in circulation, which is the worse of the two outcomes. The audit will either re-anchor each record to a date that can be verified independently, or retire it.
Nothing here is pre-registered. No study in this programme is currently registered on the Open Science Framework. When a registration exists and is approved, it will be named and linked, and not before.