ELIZA on steroids has been publishing structural analyses on accountability, politics, media, pandemic and technology since 2024. This page summarises how the series works — methodically, traceably, without illustrations. It is a response to recurring questions in comments and emails, and at the same time an explanation for everyone unfamiliar with the format.
This page is not a disclaimer. It is a methodology statement. Anyone reading the series should know what the posts claim, how they back it up, and what they do not claim.
What This Series Is#
Structural analyses are findings about architectures — institutional structures, personnel pipelines, funder networks, mandate concatenations. They work with documented substrates: released files, court rulings, peer-reviewed studies, sworn testimony, FOIA materials, SEC filings, parliamentary hearing minutes, official press releases.
They are not opinion pieces. They do not comment on personal extravagance, travel expenses or tabloid scandals. They analyse structures whose mechanism would remain invisible without the analysis.
What This Series Is Not#
This series is not a conspiracy narrative. It does not claim a “rigged game”, a “plan since 1999”, or covert steering. Where the substrate carries something concrete (personnel cross-linkage, funder pipeline, mandate architecture), the finding names the concrete. Where the substrate does not carry a further conclusion, the post does not assert one.
It is also not a tabloid format. The series does not judge persons over their lifestyle. It analyses functions — and names when functions invert (a supervisory body becomes a prosecutorial actor; an advisory committee becomes a ready-made ordinance template).
Frequent Questions#
1. Isn’t this all conspiracy theory?#
That is the classification question every serious reader asks first. The series works along a distinction we internally call Class A vs. Class B.
Class A are findings carried by direct substrates: files, court rulings, peer-reviewed studies, sworn testimony, FOIA materials, SEC filings, hearing minutes. The series works exclusively in Class A.
Class B are narratives suggested by editing, sound design or selective sequencing — typical of tabloid exposé formats or conspiracy channels. The series does not use these.
When a post documents personnel pipelines or mandate concatenations, the individual steps are separately verifiable (Person X was in exercise Y in 2001, moved to agency Z in 2009, was involved in exercise W in 2018 — all via public press releases, Wikipedia career pages, SEC filings). The chain follows from the individual steps.
2. But Event 201, SPARS and Atlantic Storm were just exercises?#
Yes. That is exactly the observation. An exercise trains. Whoever trains in an exercise builds a rehearsed response architecture. Whoever builds the response architecture is subsequently qualified for its implementation, because they know it most precisely.
The series does not claim that a pandemic was planned. It documents that the architecture of its response was rehearsed in advance across five exercise stages 2001-2019, and that the players of those exercises subsequently appeared as actors in the same architecture during the real COVID years. That is not plan. It is casting call.
See concretely: Dark Winter 2001, Atlantic Storm 2005, Event 201 2019.
3. Why always Johns Hopkins Center for Health Security?#
Because Tom Inglesby has been Director of the Center since 2013 and has been involved as co-designer in four of the five central pandemic tabletops since 2001 (Dark Winter, Atlantic Storm, SPARS — as director of the publishing center, Clade X, Event 201). Eighteen years, one person, one institutional constant. That is a documentable observation, not a personal accusation.
The series names Inglesby not because it marks him as guilty but because his function is the institutional constant of an architecture across two decades.
4. If everything is public, what’s the problem?#
That is exactly the series’ argument. Public does not mean processed. Substrates are documented; their concatenation is not. The JHU exercise microsites have been online for years; player lists were always public; SEC filings are public by law; Wikipedia career paths are public.
What the series provides is the concatenation of these public substrates into an institutional reading. This reading is not formulated in the substrates themselves; it emerges from their assembly. That is precisely the definition of structural analysis.
5. Do you have a political agenda?#
Structural analysis is not partisanship. The series criticises structures — across all political spectra. In the pandemic-architecture cluster, criticised equally: the Federal Interior Ministry under CSU leadership (shock paper 2020), the Health Ministry under CDU leadership (Spahn / Holtherm appointment 2020), the Leopoldina as an academic advisory body (statement 12/2020), the federal inquiry majority of CDU/CSU/SPD/Greens/Left (data and graph blockade 2026).
We criticise structures, not personal extravagance — see internal memo Verschwendungs-Aufdeck-Reflex vermeiden (avoiding the extravagance-exposure reflex). Travel expenses, sneaker brands, income-tabloid material are not Eliza fare.
6. Why do you cite Reitschuster, Tichy, Achgut, etc.?#
The series follows a simple source policy: every central factual claim has a mainstream, file-based, or court-derived backing. Secondary sources like Reitschuster, Tichy, Achgut are cited when they document a fact that mainstream sources under-cover — but only as additional evidence, not as primary source.
Example: the Stephan Kohn memorandum is documented in mainstream by Cicero coverage (primary source); the personal consequences for Kohn are additionally supported by Tichy/Achgut reporting. The function: mainstream plus secondary for completeness, never secondary instead of mainstream.
7. What about “both sides have a point”?#
Faux-balance is not objectivity. Objectivity is the willingness to follow the evidence to where it leads — even when that is politically uncomfortable. Mandatory genuflections of the type “both sides have a point” or “of course there are legitimate concerns” — when placed as disclaimers ahead of the actual critique and used to soften it — are not Eliza style.
The series has a position. It is at once factual AND has a position. The two are not mutually exclusive.
8. Aren’t these overstretched conclusions?#
Each post in the series formulates its conclusions only as far as the substrate carries them. When a post says “Four out of ten players moved into the agencies their exercise had recommended”, this statement is supported by concrete career paths. When a post says “Inglesby led four of the five tabletops”, this statement is verifiable via public exercise microsites.
Where a post goes further — e.g. to the structural-analytical reading of a “career pipeline” or “function inversion” — the argument is explicitly marked as structural analysis, not as causal claim.
9. Why the constant US lens?#
The series does not work from a US lens but documents a transatlantic architecture. WHO IHR 2005 was adopted in Geneva, four months after a transatlantic exercise in Washington (Atlantic Storm) with players from France (Kouchner), Germany (Hoyer), Norway (Brundtland), Poland (Buzek), Sweden (Eliasson). The Trusted News Initiative was co-founded by the BBC. The EU DSA architecture builds on SPARS language regulations.
The German application layer (Leopoldina statement 2020, Holtherm militarisation 2020, Kohn sanction 2020-2022, inquiry majority blockade 2026) gets its own posts, on equal footing with the international architecture.
10. What would prove you wrong?#
Structural analyses are falsifiable. Concrete falsification conditions for the pandemic-architecture series:
- If Tom Inglesby was not co-designer of the five tabletops 2001-2019 — wrong.
- If the player lists of Dark Winter, Atlantic Storm, Clade X and Event 201 show no personnel overlap — wrong.
- If Jane Halton did not become CEPI Chair in 2020, Avril Haines not DNI in 2021, Margaret Hamburg not FDA Commissioner 2009-2015 — wrong.
- If WHO IHR 2005 was not adopted in May 2005, four months after Atlantic Storm — wrong.
- If the Open Philanthropy Foundation did not fund Clade X or the Bill & Melinda Gates Foundation did not co-host Event 201 — wrong.
Each of these conditions is verifiable through public sources. Whoever refutes one of these conditions refutes a substantial part of the series. That is precisely the task of serious critique — and that is precisely why the substrates are explicitly linked in every post.
How We Operate#
Each structural-analysis post follows a fixed schema:
- Substrate inventory: which files, studies, statements, biographies exist? Which are primary (files/courts/sworn), which secondary (reports)?
- Mechanism description: what function connects the substrates? Career pipeline? Mandate inversion? Function reversal?
- Conclusion: what does the finding mean for the understanding of the structure under study? No mandatory punchline, no “the reader may decide” dilution — but clear naming of what the substrates carry.
- Source list: mainstream, file-based, court-derived and secondary sources with direct linking. Every central claim is linked.
- Cross-references: connection to earlier posts in the series when the finding is part of a larger structural pattern.
Before each commit an internal Self-Check runs: phrase scan for typical faux-balance reflexes (“of course…”, “this does not change…”), closing test for sharpness rather than dilution, cross-reference verification. The series’ reputation stands above any single post.
Who Writes This#
Alexander Renz publishes the structural analyses. The methodology has been developed jointly with ELIZA — a Claude-based writing and research tool. ELIZA operates under formalised discipline rules (SOUL.md, SELF_CHECK.md, several feedback memos) intended to compensate for the faux-balance bias reflex of language models. Final editorial responsibility lies with the human — not the model.
What We Do Not Do#
- No personally directed accusations without documented substrate.
- No speculation about individual motives without backing.
- No travel-cost, income or lifestyle analyses.
- No “who was the real string-puller?” narratives.
- No “rigged game” closing lines when the substrates do not support the claim.
- No shock sound design, no “CORE FACTOR” edits, no clickbait headlines.
If You Want To Disagree#
Write to kritik@elizaonsteroids.org with: (a) a specific post, (b) a specific substrate, (c) your counter-evidence. If your counter-evidence holds, we correct the post publicly and by name. That is methodological obligation, not friendly accommodation.
Past corrections — source attributions, personal names, factual claims, retractions — are documented in the Corrections archive, each with date, commit hash, and reasoning. The list is deliberately public, because a methodology obligation without visible correction behaviour is an empty gesture.
Or, more directly: refute one of the ten falsification conditions in Question 10. If one of them falls, a substantial part of the series falls.
Terms used in the series#
The structural analyses use a small set of recurring technical terms (apparatus, pipeline, echo machine, mandatelessness, amplifier / damper / explainer slot, function inversion, population bypass). They are defined briefly in the Glossary, each with one example post showing the term in use. Anyone meeting a term for the first time can look it up there.
This methodology page was first published on 27 April 2026 and is updated when methodology changes substantially. The state is rolling.
German version: Wie diese Reihe arbeitet