My Q and QAnon tweets as a historical corpus

1,749 posts, nearly seven years, and an invitation to independent analysis

I am releasing a searchable collection of my Twitter/X posts about Q and QAnon. It contains 1,749 posts, beginning on 22 December 2017 and ending on 8 December 2024.

My aim is to make the primary material available for anyone who wants to examine it: researchers, journalists, historians, supporters, critics, or simply curious readers. You can trace the language, arguments, predictions, interpretations, and personal experiences recorded at the time.

The wording has been preserved. These are historical statements whose claims can be assessed against other evidence.

Martin Q Final Extract Corrected

572KB ∙ XLSX file

Download

Download

The source and the selection

The underlying source was a Twitter/X backup generated on 1 January 2025. Its three main tweet files contained 105,937 unique records, spanning 23 December 2007 to 1 January 2025. That larger collection includes original posts, replies, and retweets.

Posts that quote somebody else can still qualify if they contain a substantive contribution from me. The quoted material remains in place for context and should retain its original attribution.

The gap needs explaining. I was deplatformed from Twitter between 2021 and 2023. The underlying archive’s surviving activity jumps from 9 January 2021 to 1 June 2023. The first selected Q-related post after that interruption is dated 12 June 2023.

Anyone studying posting frequency or changes in emphasis should account for that break in access. The Twitter record does not provide continuous coverage of those years, or a complete record of what I was publishing elsewhere.

How the messages were selected

The extraction used AI assistance, text searches, automated integrity checks, and targeted editorial review. The working criterion was straightforward:

Martin said something substantive about Q or QAnon.

That includes arguments, interpretations, predictions, criticisms, observations about research and communication, and personal accounts of participation and its consequences.

The process had several stages:

1. Find possible references. Searches covered explicit Q/QAnon wording, standalone “Q,” and related terms such as Q drops, Q proofs, QArmy, and Qisms.

2. Remove unrelated uses. My earlier professional work involved the DeltaQ (∆Q) network-performance framework. Those references, along with ordinary questions, product names, and other unrelated uses of the letter Q, could otherwise contaminate the results.

3. Separate my contribution from material I circulated. Retweets were excluded. So were quotation-only posts, AI-generated passages without substantive commentary from me, routine sales notices, bare links, and brief reactions that added little analysable content.

4. Review overly restrictive exclusions. A final #QAnon hashtag can identify the subject of a substantive statement. Short posts can also contain meaningful claims. The review restored posts rejected by earlier hashtag and length rules—including the December 2017 opening post.

5. Recover overlooked material. A fresh search of the raw backup found 68 Q-related records missing from the initial candidate index. Of these, 55 were retained as substantive non-retweets. Seven usable long posts were also recovered from the backup’s separate long-post file.

6. Check the resulting records. Checks covered unique post IDs, chronological ordering, dates, permalink consistency, and text fidelity. Where fuller long-post text was recovered, the XLSX file retains the original excerpt separately and records the source of the expansion.

These checks matter because an extract can copy every selected message perfectly while still missing relevant messages. The validation found examples of both search omissions and overly restrictive filtering.

What this release can support

The corpus allows readers to examine how my public commentary developed across time. Possible questions include:

• What did I claim, predict, or leave uncertain at different stages?

• How did I describe research, evidence, authority, and independent judgment?

• How did expectations about timing change?

• What roles did community, professional experience, personal cost, and spiritual language play?

• How did my descriptions of Q relate to my criticism of established media?

Those questions need not produce the same answers for every reader. Publishing the records makes the discussion more inspectable.

What this release cannot support

There are limits:

• This is an editorial selection from a surviving archive, rather than a certified inventory of every Q-related post I ever published.

• The 2020 and 2026 backups have not yet been fully compared against the validated 2025 baseline.

• Deleted material, implicit references, missing media, and incomplete thread context may leave further gaps.

• The entire retained collection has not received an independent line-by-line second review.

Final word

Distinct post IDs are preserved even where the same wording was published more than once; those repetitions are marked. All dates are UTC, and the original post IDs and permalinks are supplied.

If you analyse the corpus, please identify the release as Martin Geddes Q and QAnon tweet corpus, 20 October 2026, explain any additional filtering you apply, and cite the original dated posts where possible.

I would welcome analyses that illuminate the material, and corrections that identify genuine omissions or attribution problems. The purpose of making this available is to let others work with the historical record directly.

Walking the same path I did in Covid with the same camera playing the same music.