Idle Lab / How we work
How a result reaches this site
The lab is an ordinary home computer running scheduled analyses of public scientific data. Making a measurement is the easy part. This page explains what happens before we call anything a finding.
From signal to submission
- Measurement. An experiment downloads authoritative data, records the exact files and their hashes, and computes measurements with uncertainties.
- Automated checks. Sanity checks on completeness, physical plausibility and data integrity. Any candidate claim must also be reproduced in an independent dataset.
- Human review and approval. A person reads the evidence, writes the plain-language summary and caveats, and approves publication. That approval is a recorded commit.
- External review. Where a result could be useful, we send the evidence to the organisation or researcher best placed to assess it, and record their response here, whatever it is.
What the statuses mean
| Awaiting independent review | Passed the lab's automated checks and a human read-through. No outside scientist has assessed it yet. |
| Under review | An outside reviewer or organisation is assessing it. |
| Supported by review | At least one independent reviewer found the evidence sound. This is not the same as peer-reviewed publication. |
| Revised | Corrected after review. The history below records what changed. |
| Retracted | Found to be wrong or unsupported. Kept here, with the reason, so the record stays complete. |
What we don't claim
Nothing here is peer-reviewed research unless the review record says so. "Passed all validation gates" means the lab's own automated checks passed and a second dataset agreed. It does not mean an expert has confirmed the result. Most unusual signals turn out to be data or processing errors, and when we find one, we say so.