We document content-selective search de-indexingRemoving a URL from a search engine's index so it doesn't appear in results, even though the page still exists and returns HTTP 200. Different from deleting the page — the content is there, it's just invisible to search. across three platforms (Cryptome, Slashdot, Substack). Control cases on each platform demonstrate that the de-indexing is content-based, not domain-level or SEO-related. Every test in this report is a Google search that a reader can perform independently in under five minutes. We make no claims about the mechanism or intent behind the de-indexing. This report documents that it occurs, provides evidence of selectivity, and is designed for independent replication.
Your Replication Log
Click the checkbox after you run each test. Your timestamps are stored locally and never sent anywhere.
Whether specific URLs that exist and return HTTP 200The standard web server response code meaning 'OK — the page exists and was served successfully.' A page returning HTTP 200 is alive and accessible; if Google can't find it, the absence is on Google's side, not the server's. appear in Google search results when searched by exact title, by content keywords, or by the site: operatorA Google search modifier that restricts results to a specific domain. 'site:example.com query' returns only pages from example.com matching 'query'. When a page exists but the site: operator can't find it, the page has been de-indexed..
Content on the same domain, by the same author, with similar age and traffic characteristics, that is indexed by Google. If the suppressed content's absence were caused by domain-level blocking, poor SEO, or low traffic, the control content would be equally absent.
Google Search. No specialized software, subscriptions, or technical skills are needed.
Only cases that meet all three conditions are included:
Cases that are ambiguous, require platform-specific knowledge, or could plausibly be explained by organic search behavior are excluded.
In at least one prior case on a third platform, content-selective de-indexing was lifted shortly after an independent technical authority examined and confirmed the pattern. The de-indexed content had been invisible to search engines for over a decade; within weeks of independent verification, it became findable again.
This suggests the de-indexing state may be actively maintained and responsive to external scrutiny. Researchers are advised to document their initial findings immediately upon replication — including screenshots and timestamps — as the observable state may change once verification is underway.
Each test below includes the exact search query, a button to copy it to your clipboard, a link to run it on Google, and a direct link to the content proving it exists.
C01 uses the same author, same domain, same Substack account, and a similar publication date range as T02 and T03. If SEO quality, traffic volume, or domain reputation explained the absence, all articles on this Substack would be equally invisible. They are not. The selectivity is content-based.
The same user (EnergyScholar) has 10+ comments indexed on Slashdot — including comments about NSA surveillance, cryptography, and WikiLeaks. The one comment that is invisible to search is the one combining claims about Five Eyes topological quantum computing capability with a prediction that the comment itself would be suppressed. Same user, same platform, same era, same topic area. The selectivity is at the level of a single comment.
Each standard explanation for a URL not appearing in search results is examined and ruled out.
| Alternative Explanation | Evidence Against |
|---|---|
| Domain-level blocking | C01 proves postquantum.substack.com is indexed. C02 proves Slashdot comments by the same user are indexed (10+). Other Cryptome pages (e.g., the site's front page, its widely linked documents) appear normally in Google results. The de-indexing is page-specific, not domain-level. |
| Low PageRankGoogle's algorithm for ranking web pages by importance, based on how many other pages link to them and how important those linking pages are. Higher PageRank = higher in search results. A page with few inbound links might rank low, but this doesn't explain selective absence — the page wouldn't appear at all. / poor SEO | T02, T03, and C01 are on the same Substack account by the same author, with similar publication dates and audience. Substack controls the page structure, metadata, and SEO for all articles equally. T01 (Cryptome) has 42 Wayback Machine captures over 12 years — Google's crawler has visited it hundreds of times. T04 (Slashdot) is a Score 5 comment by a user whose other 10+ comments are all indexed (C02). |
| robots.txt exclusion |
Substack's robots.txt does not disallow post pages. All articles from the account (including C01) use the same URL structure (/p/slug).Cryptome's robots.txt does not disallow the path /2012/03/qc-footprint.htm.Slashdot's robots.txt does not disallow comment pages — and C02 proves they are indexed when not selectively excluded. |
| noindex meta tag | Substack controls the <meta> tags for all articles. Since C01 is indexed, Substack does not set noindex on articles by default. The suppressed and control articles use the same template. T01 (Cryptome) uses static HTML with no noindex directive. T04 (Slashdot) uses the same comment template as the 10+ indexed comments by the same user. |
| Content age | T01 was published in March 2012. T04 is from June 2012. T02 and T03 are from 2022–2024. C01 is from December 2022. C02 includes comments from 2011–2013, the same era as T04. Age does not correlate with de-indexing — old and recent content are both affected, while content of similar age is indexed. |
| Platform moderation | All URLs return HTTP 200 — no platform has removed or restricted the content. The content is publicly accessible to anyone who visits the URL directly. |
Each URL is verified to exist and to be archived independently of this report.
| ID | URL | HTTP Status | Wayback MachineInternet Archive's service that periodically saves snapshots of web pages. If a page has Wayback captures, it proves the page was publicly accessible and crawlable — a search engine's failure to index it can't be blamed on the page being hidden or ephemeral. |
|---|---|---|---|
| T01 | cryptome.org/2012/03/qc-footprint.htm | 200 | 42 captures, 2012–2024 |
| T02 | postquantum.substack.com/p/kangaroos | 200 | Archive link |
| T03 | postquantum.substack.com/p/death-alan-turing | 200 | Archive link |
| T04 | slashdot.org/comments.pl?sid=2926019&cid=40375589 | 200 | Archive link |
| C01 | postquantum.substack.com/p/magic-numbers-are-a-joke | 200 | Archive link |
| C02 | slashdot.org — "Spooky: How NSA's Surveillance Algorithms See Into Your Life" (EnergyScholar comments) | 200 | Archive link |
google.co.uk, google.de, google.caTo track changes over time, re-run the full battery monthly and record the date and result for each test. Any change in status (SUPPRESSED → INDEXED or vice versa) is itself a data point worth documenting.
To report your replication results, discrepancies with the documented outcomes, or changes observed over time: