The Prophet and the Record

Chain of custody. This edition stands on a source document assembled the same day, which merges two bodies of material: eight auto-captioned transcripts of Jiang Xueqin’s 2026 sessions, and a verification layer built against primary documents, wire reporting and scholarship. The transcripts were first synthesized into a themed knowledge base under a strict no-outside-facts rule; only afterwards was the record consulted. That ordering is deliberate and it is the same discipline the adjudication asks of Jiang: fix what was said before deciding whether it was true.

What was checked. The roughly twenty-five claims the framework rests on, every dated prediction, and every named source and text — eighty claims in all. Occult and metaphysical claims were deliberately excluded from adjudication rather than rated badly; they carry a tag and are left to the reader. Where the record confirms Jiang — which happens more often than a reader of the raw corpus would expect — that is stated with the same prominence as where it does not. This is neither an exposé nor a defence.

Transcription. The captions are corrupt in specific, recoverable ways. “World War Three” is rendered “World War II” almost every time and is silently corrected. Garbled proper nouns are normalized where the referent is unambiguous — Hormuz, Malacca, Gibraltar, Chabad-Lubavitch, Schneerson, Thucydides, Hegseth, Spengler, Kharg. Dates Jiang gives inconsistently are flagged rather than fixed.

How to read this. Six parts, in order. Threaded through are thirteen numbered Patterns — reasoning moves stated explicitly so you can carry them to problems that have nothing to do with geopolitics — and after each part a Mental Model built to outlast the details. Every figure ships its numbers as a table, so nothing depends on seeing a colour. The charts are static SVG; the page reads with scripting off.

Three rules govern every figure here. They are declared once and never restated, and they are specific to what this document is about.

1 · Amber is his. Blue is the record.

The entire exercise is one comparison — what Jiang said against what can be established. Amber is always a quantity or claim Jiang produced. Blue is always the record. Nothing else on the page is either colour. Where a verdict ramp appears, it runs from blue through grey to red: blue because the record supports the claim, red because the record runs against it. That reuses blue with its meaning intact rather than against it.

2 · Solid is verified. An open outline is his.

A filled mark is a quantity some source measured or some document states. An open outline, drawn at full size and never faded, is a quantity that Jiang produced and no source confirms. It is not a lesser version of the number; it is a different kind of number. That one rule catches his sixty thousand troops, his two billion people, his fifth pillar of a four-pillar document, and every modelled oil price in Figure 6 — because those are modelled too, and the rule does not care whose arithmetic it was.

3 · Evidence tags travel with the mark.

Every claim carries a tag for how it was obtained, and the tag stays attached even when it is inconvenient. The tags do real work later: when Jiang and the record disagree, which side is built from F and which from S is the whole argument.

Retrieved fact

Dated and sourced. Tiered from T1, a primary document or official release, down to T4, aggregator or secondary analysis.

Estimate

A quantity with assumptions, stated or reconstructable.

Inference

Derived by Jiang from other elements of his own base.

Speculative

Rumour, association chain, unnamed sourcing. A fact that has forgotten how it was obtained is a rumour with tenure.

Not in evidence

The claim exists in the discourse and in no primary source. It is refused rather than rated.

Six verdicts are used. CONFIRMED, the record supports it. CORE / INFERENCE, the factual predicate is verified and the conclusion drawn from it is not — this one does most of the work. PARTIAL, materially right and wrong in separable parts. CONTRADICTED, the record runs against it. UNRESOLVED, future-dated or genuinely open. NOT IN EVIDENCE, refused.

In May 2024, in a Beijing classroom, Jiang Xueqin made three predictions on camera. Trump would win in November. Trump would start a war with Iran. And the United States would lose it. He sketched a scenario he called Operation Iranian Freedom, said conquering Iran would need three to four million soldiers, and warned that deployed American troops would become “hostages, not soldiers.”

On 28 February 2026 the United States and Israel launched Operation Epic Fury. Within three days the video had earned him a hundred thousand subscribers, an epithet — China’s Nostradamus — and an audience that now numbers in the millions. By July he had left teaching and was touring: Tokyo, Los Angeles, New York, holding three-hour sessions in packed rooms.

The standard reaction to this is one of two things, and both are useless. The first is that he is a prophet. The second is the line a commenter left under the Newsweek profile: “Even a broken watch is right twice a day.” Neither is a finding. Both are ways of not doing the work.

The work is available, because Jiang made it available. He treats his predictions as the test of his model and says so repeatedly. He names three events that would break it outright. He publishes his sources. He tells you which of his claims are intuition and which are documents. Whatever else is true of him, he has built something that can be scored — which is more than most of the people who dismiss him have done.

So it was scored. Eighty claims, one at a time, against the record. What comes back is not a verdict on a man. It is a shape.

He is perfect on when. He is good on what a document says. He is a coin-flip on how much. And on why — which is nearly half of everything he says — he is right about one time in seven.

That shape is more useful than either the prophecy or the dismissal, because it tells you what to do with him. It is also, as it turns out, a shape that a great many confident analysts have, and almost none of them have been measured for it.

Part I builds the instrument: how a body of prophecy becomes a rated ledger, and what licence Jiang himself gives for doing it.

A body of commentary is not a claim. It is a few hundred claims of five different kinds, delivered in one voice, and its truth value is not a single number. The first move is therefore mechanical: break the corpus into separable assertions, tag each by how it was obtained, check it, and rate it.

Eighty claims survived that filter as load-bearing — meaning the framework changes shape if they fail. Occult and metaphysical material was set aside rather than rated, on the principle that a document should refuse where it cannot judge. What remains splits cleanly into five types, and the split turns out to matter more than anything else in this study.

The second move is to check whether the subject consents to being scored this way. Jiang does, explicitly. Asked in a livestream what would change his mind, he names three events without hedging: if Trump dies, resigns or loses in 2028, “the materialist viewpoint would be much more valid”; if Russia and Ukraine sign a real peace, “I would have to start from scratch”; if China invades Taiwan, “I’m completely wrong about China.”

That is not a small thing. Most commentators of his reach never state a condition under which they would be wrong, and the ones who do rarely pick conditions that could plausibly occur. All three of his remain open. None has yet triggered. The adjudication that follows is fair rather than hostile precisely because he set the terms.

There is one structural weakness in his method and he states it himself, in a passage about how to retain information. Asked how he holds so much material, he describes a memory-palace technique:

“Rather than focus on the facts, focus on the story… the grand structure. Don’t worry also if this grand structure is factually correct or not, but try to construct the paradigm, try to construct the narrative. And then you organize the facts to fit the narrative.”

He offers this as pedagogy, and as pedagogy it is defensible — narrative scaffolding really does aid recall. But it is also, read plainly, a described procedure for fitting evidence to a pre-built frame. He reinforces it a second time: “You always want to solve the macro, because the macro gives you the framework, and the micro are just specific instances that confirm the macro.” Instances that confirm. Not instances that test.

Against that sits a genuinely good practice he also describes: he reads Russian and Iranian state media against Western coverage and treats convergence between hostile sources as a reliability signal. He estimates ninety per cent of what circulates on Twitter is false. He opened one session by naming a reader who had emailed to correct his factual errors, accepted the corrections on air, and reflected that “sometimes I prefer the sound bite… I need to be more nuanced, more balanced, more scholarly.”

Part II is the finding. It is a single shape, and everything after it is an attempt to explain and then to test that shape.

Split accuracy by what kind of claim is being made, and it does not stay flat. It falls, in order, and the fall is steep.

The bottom row is the finding. Not because fourteen per cent is low — on a hard forecasting problem it might be respectable — but because of how much sits in that row. Thirty-nine of eighty claims are causal attributions. Nearly half of everything load-bearing he says is an answer to why, and that is the category where the record most often supplies a different answer.

And the answer it supplies is almost always more boring. A court ruling. A statute. An attrition curve. A bureaucratic process that nobody chose and nobody is running.

The descent is not evenly distributed across subject matter. Sort the same eighty claims by domain and the pattern sharpens into something closer to a diagnosis.

Grand strategy and the dollar system are his strongest domains and they have something in common: both are about published documents and observable plumbing. The National Defense Strategy exists and says what he says it says. The yen carry trade exists and transmits the way he says it transmits. Neither requires him to know what anyone intended.

Collapse mechanics is his weakest, and it has the opposite property. Every claim in it is an analogy imported from a discipline he has read about rather than in — Bronze Age archaeology, rodent ethology, agronomy — and in each case the import has gone wrong in the same specific way, which Section 10 takes apart.

Before the argument goes further it is worth looking at the raw material, because the single most dangerous property of this corpus is not that it contains errors. It is the ratio.

An analyst is entitled to inference. The whole value of analysis is inference. What an analyst is not entitled to is the transfer of confidence — earning credibility on the thirty per cent that is retrieved fact, and then spending it on the seventy per cent that is not, without ever marking the transition.

Jiang does mark it, sometimes. He labels his association chains as association chains: “I don’t know what’s going on. I’m just trying to draw these points of convergence.” He distinguishes logic from intuition aloud. But the marking is inconsistent, and it is absent at exactly the moments it matters most — which the next three sections demonstrate on his three strongest cases, and the three after that on his three weakest.

Part III takes his three strongest cases at full strength, because a verdict that has not met the best version of the argument is not worth having.

Start with the thing he gets most right, which is also the thing least discussed about him: he is a competent macro-financial analyst. His account of the dollar system is standard economic history told accurately — Bretton Woods, the 1971 closing of the gold window, the petrodollar and the opening of China as replacement anchors — and the diagnosis he attaches is one many orthodox economists would recognize.

Then he does something more specific. He argues that sanctioning Russia in 2022 destroyed the political neutrality that reserve-currency status depended on, and that the observable consequence would be a rotation out of Treasuries and into gold. The record: the dollar’s share of global foreign-exchange reserves has fallen below 57% for the first time since 1995, and central banks are buying gold on the order of a thousand tonnes a year.

And he is right about the scale of what closing the Strait of Hormuz does, in a way most commentary was not.

He is also right, and unusually early, about a channel the reference macro model leaves out. The Dallas Fed analysis states plainly that its model “does not account for… the disruption of natural gas and fertilizer exports from the Persian Gulf.” Jiang spends more time on fertilizer than on oil. His number for what fertilizer failure would do is wrong — badly, as Section 10 shows — but his identification of the gap is not.

Jiang’s most repeated rhetorical move is to tell his audience to stop speculating and go download the document. “Anyone can go there and download it for free and read it themselves.” He does this about the National Defense Strategy, about the Clean Break memo of 1996, and about a policy brief on Ceuta. In all three cases the document exists and says substantially what he says it says.

The National Defense Strategy was released on 23 January 2026 and is public. Its four pillars are homeland defence, deterring China, burden-sharing with allies, and the defence industrial base — in that order. When his interviewer paraphrased the fourth as “supercharge the US defense industrial base,” Jiang said “that’s exactly correct.” It is: that is the document’s own phrase, word for word. The related National Security Strategy commits to enforcing the Monroe Doctrine “with a new Trump Corollary” — which is what his joke name, the “Donroe Doctrine,” is pointing at. The NDS names the Arctic, South America, Greenland, the Panama Canal and the Gulf of Mexico as key terrain, which is most of his chokepoint list.

And then he adds a fifth pillar.

The two supporting instruments he names for the chokepoint programme also check out, and he dates both correctly: a US–Indonesia defence partnership announced 13 April 2026, and a US–Morocco ten-year defence roadmap the same month. What those agreements are for is his inference. That they exist, and when, is not.

The strongest single passage in the corpus is not a prediction. It is a mechanism, described in New York in early August, that required no coordinating actor to be right.

The argument runs: Japan is the world’s marginal supplier of liquidity through the yen carry trade — borrow at near zero domestically, buy US Treasuries at several points higher, take the spread. Japan also imports most of its energy from the Gulf. So a Gulf disruption raises Japan’s import bill, weakens the yen, and eventually forces the Bank of Japan to recall liquidity from the world to defend it — which pulls the support out from under exactly the American asset prices that are carrying the economy.

This is worth stating plainly because it cuts against the shape of the rest of this document. On the one mechanism in his framework that was near-dated, specific and falsifiable within weeks, he was right — and right for structural reasons that involve no secret societies, no eschatology and no coordinating actor. It is the part of his work that most resembles ordinary good analysis, and it is the part his audience discusses least.

Part IV takes the three weakest cases. The first is the most instructive document in the corpus, because it contains his best and worst work on the same event, seventy-two hours apart.

On roughly 31 July, hours after tens of thousands of people crossed from Morocco into the Spanish enclave of Ceuta, Jiang convened a three-hour emergency session. What he produced there was, in part, genuinely good journalism.

He had found a policy brief published by the American Enterprise Institute on 16 March 2026, in which Michael Rubin argued that “Moroccans should gather, send bulldozers to the border, and then enter Ceuta and Melilla unarmed to raise the flag” — predicting that “Sánchez and the Spanish press bleat, but they have no grounds to act,” and arguing that NATO’s Article 5 does not cover the enclaves geographically. Every word of that is in the brief. He had the congressional report describing the enclaves as Moroccan territory. He had the ten-year US–Morocco defence roadmap. He knew that Spain’s offence was refusing the United States access to its airbases — Rota and Morón — and that Morocco had waved migrants through once before, in 2021, to punish Madrid over Western Sahara.

Almost none of that was in the Western coverage of the week. He was reading things other people were not.

And then, in the same session, he asserted that Moroccan police had loaded migrants onto government trucks and directed them across; that the United States and Israel had orchestrated it; and that Ceuta was therefore lost, with “absolutely nothing the Europeans can do about this.”

The proximate cause was legal and domestic. A Spanish Supreme Court precedent had established that migrants arriving by swimming do not cross a “border containment element,” which removed the legal basis for summary return; and a second ruling, STS 868/2026 of 8 July, struck down provisions weakening protections for unaccompanied minors. Together they voided the mechanism Spain had used to manage the border. The surge followed within three weeks.

None of that appears anywhere in the three-hour session. And while Jiang was explaining that the enclave was irrecoverable, roughly forty-eight thousand of the sixty thousand had already voluntarily returned to Morocco.

The weakest domain in the ledger is collapse mechanics, and its failures share a signature. In each case Jiang imports a mechanism from a field he has read about, reproduces the framework accurately, and then leans his conclusion on the one component that field has specifically discarded.

The Bronze Age. His account of the collapse is faithful to Eric Cline’s: a perfect storm of climate change, drought and famine, earthquakes, invasions and internal rebellions, in which no single cause suffices and interconnection converts local failure into systemic failure. That is right, and the systems-collapse framing is a real contribution to how his audience thinks about fragility.

Then he makes the Sea Peoples the agent — mass migrations of the hungry, organized into pirate armies, overwhelming civilizations — and maps them onto a coming wave of climate and famine refugees into Europe. But Cline’s argument exists in large part to demote exactly that explanation. Inland empires — Kassite Babylonia, Elam, Assyria — collapsed without ever encountering the Sea Peoples. They were a symptom, not a cause. Jiang has borrowed the frame and hung his forecast on the piece the frame removed.

The fertilizer arithmetic. He says that without fertilizer the planet supports two billion people.

The rat utopia. The third import is the most compressed error in the corpus. Jiang describes experiments by a researcher he calls “James Poon,” in which rat colonies given unlimited food and no predators collapsed into warfare — and applies it to young men in a gerontocracy who will, he argues, eventually form groups and produce urban violence.

Three failures, one signature. Name it.

Where a structural, legal or bureaucratic cause is available, he reaches past it for a coordinating actor.

At Ceuta the available cause was two rulings of the Spanish Supreme Court; he reached for Washington and Tel Aviv. On the draft, the available cause was a registration provision in an appropriations bill; he reached for a war plan requiring bodies. On the National Defense Strategy, four pillars were on the page; he added a fifth describing an intention. On Iranian missile attrition — launches falling from 480 a day to 40 — the available cause was that they were being destroyed; he reached for a deliberate feint, Iran playing possum to lure an invasion.

That last one is instructive because it may be right. The record does not adjudicate between attrition and strategy; both fit. What is diagnostic is not that he chose the strategic reading but that he chose it without noting the other, in a corpus where he is otherwise careful to enumerate possibilities.

The substitution has a structural cause of its own, visible in his own analytical template.

Part V puts both cases at their strongest before ruling. A verdict that has not met the best version of the opposing argument is not a verdict.

For Jiang. In May 2024, with no institutional position and no access, he made three predictions on camera. Two have landed. He said Trump would win; Trump won. He said Trump would start a war with Iran; the war began on 28 February 2026. The third — that America would lose it — remains open, and the current force posture is not obviously against him: roughly forty to fifty thousand American troops across the entire theatre, and named defence analysts saying plainly that “there simply aren’t enough troops” for an Iraq-style invasion.

He read the National Defense Strategy and reported its pillars accurately, including one word for word. He found the Clean Break memo, dated it to 1996, attributed it to a Perle-led study group with Feith in it, and described its recommendations correctly — all of which is right, and all of which most commentary on the Middle East gets wrong or never mentions. He found an AEI brief on Ceuta that the Western press had almost entirely missed, and quoted it verbatim. He correctly dated two defence agreements. He described the tanker seizures of 7 January with precision, down to the mid-chase reflagging.

He predicted, in advance, that Trump and Xi would meet in Beijing in mid-May. They met on 14–15 May. And he laid out a chain from the Strait of Hormuz through Japanese liquidity to American asset prices which was validated three weeks later by a Treasury intervention that had only two precedents this century.

He also understated the AI capex figure. When the truth is more alarming than the alarmist’s number, the alarmist is not simply catastrophizing.

Put at full strength, the case is this: a man with no institutional access, working from open sources in a second language, has repeatedly identified real documents, real mechanisms and real dates that credentialed commentary missed — and has done so far enough in advance to be checked.

Against Jiang. He is wrong about causes six times out of seven, and causes are nearly half of what he says. He asserted that Moroccan police bussed migrants to a border where they were in fact deploying water cannon. He declared an enclave permanently lost while four-fifths of the crossers were walking back. He cited an experiment by the wrong researcher, on the wrong species, with the wrong outcome, to support a forecast the actual finding contradicts. He is off by a factor of two on the carrying-capacity arithmetic underpinning his entire theory of collapse.

He described sixty thousand troops staged for an invasion when there were forty to fifty thousand in the whole theatre. He built a legal argument for a third presidential term that collapsed when a participant in his own livestream read the Twelfth Amendment aloud, and he has since been contradicted by the subject himself, who says it is “pretty clear” he cannot run and is publicly weighing a Vance–Rubio ticket — while Jiang predicted Vance would be discarded.

The two great-power entries his World War III thesis requires have both failed to occur: Russia condemned the strikes and suspended its own nuclear construction at Bushehr; China rejected Trump’s call for support and is, by CSIS’s assessment, shielded from the energy shock by stockpiles, coal and renewables — the opposite of the existential exposure he describes.

And there is the deeper problem, which is not any single error. It is the register. Seventy per cent of his load-bearing claims are inference, estimate or speculation, and they arrive indistinguishable from the thirty per cent that is retrieved fact. He tells his audience to go read the National Defense Strategy and they should. He tells them in the same voice about elite ritual sacrifice.

There is one more entry for the case against, and it is the one Jiang himself supplies. Before the World Cup final he set out a conditional: if Argentina won, and the officials visibly favoured them, and Messi scored the winner, then Messi would become president of Argentina and the country would become a hub of transnational capital. He hedged the antecedent aloud — “my logic tells me it’s Argentina; my intuition tells me it’s Spain” — which is more honest than most forecasting.

Spain won 1–0 in extra time. The prediction never fired.

What he did next is the single most important passage in the corpus:

“The first possibility of course is that I’m just paranoid… this is not rigged and the world is not as deterministic as we once believed. So the first explanation is that I’m just completely wrong and my entire worldview framework is just completely off base — and that is a very strong possibility.”

He then offered two rescuing explanations without choosing between them, and elsewhere converted the loss into evidence for the framework, generalizing the mechanism of Argentine hubris to Israel and Pax Judaica, and into grounds for hope: “Zionism is not invincible. It’s not inevitable. History is not inevitable.”

Leading with self-refutation is rare and should be credited. But a framework that absorbs the fixed outcome and its opposite as confirmation has stopped being at risk from the world.

Twenty-two dated predictions. Five have resolved.

One more thing had to be dated before any of this could be scored, and it is worth showing because it is the only place where the record repaid the corpus rather than the other way round. None of the eight transcripts carries a date. All eight are now fixed to the calendar by the events they discuss.

The ruling. H2. The observation is sound and the causal attribution is not.

He is a reliable instrument for locating documents, dating sequences and describing mechanisms, and an unreliable one for explaining why anything happened. Both halves are firm. Neither cancels the other. The practical consequence is that he should be read the way one reads a good scout report and not the way one reads a verdict — and that the seventy per cent of his output which is inference should be treated as inference even when, especially when, it is delivered in the voice of the thirty per cent that is not.

Part VI is the retention layer: the one move worth carrying, twelve claims worth knowing cold, a monitoring card, and the pattern index.

If nothing else survives six months, let it be this.

Split the claim from the register it arrives in. Then split the predicate from the inference.

Almost everything that goes wrong in reading a source like this — in either direction, credulous or dismissive — comes from treating each session as one object with one truth value. It is not. It is eighty separable claims of five evidentiary kinds, and they do not stand or fall together.

The Clean Break memo is real whether or not the Kabbalah reading is. The National Defense Strategy says what he says it says whether or not there is a fifth pillar. The draft-registration statute exists whether or not conscription follows. The Hormuz closure is twenty per cent of world oil whether or not the Sea Peoples are coming.

A reader who performs the split gets something genuinely useful: an early-warning system with a known failure mode, which is worth considerably more than a source with no failure mode you have identified. A reader who does not gets either a prophet or a crank, and both are unusable — not because the judgement is wrong, but because neither tells you what to do on Tuesday.

The move costs about four seconds a claim. How was this obtained? Then, if the answer is good: and does the conclusion he draws follow from it? Those two questions, in that order, are the entire apparatus of this document.

Verified, dated, and useful independent of anything Jiang says about them.

1. The 2026 Iran war began 28 February 2026 and the Strait of Hormuz has been effectively closed since — removing about 20% of global oil supply, three to five times any prior disruption on record. About 80% of that oil was Asia-bound.
2. The 2026 National Defense Strategy is public, was released 23 January 2026, and its four pillars are: defend the homeland; deter China through strength, not confrontation; increase burden-sharing with allies; and “supercharge the US defense industrial base.”
3. The accompanying National Security Strategy commits to enforcing the Monroe Doctrine “with a new Trump Corollary.” The NDS names the Arctic, South America, Greenland, the Panama Canal and the Gulf of Mexico as key terrain.
4. A Clean Break is real: 1 June 1996, Institute for Advanced Strategic and Political Studies, prepared for Benjamin Netanyahu by a study group led by Richard Perle and including Douglas Feith. It recommended breaking from Oslo, striking Syrian targets, and removing Saddam Hussein.
5. Michael Rubin’s AEI brief of 16 March 2026 proposed that Morocco send bulldozers and enter Ceuta and Melilla unarmed to raise the flag, predicted Spain would have “no grounds to act,” and argued NATO’s Article 5 does not cover the enclaves geographically.
6. The proximate cause of the Ceuta surge was two Spanish Supreme Court rulings — the precedent that migrants arriving by swimming cannot be summarily returned, and STS 868/2026 of 8 July restoring protections for unaccompanied minors. About 48,000 of the 60,000 returned within roughly a day.
7. On 3 August 2026 the United States joined Japan in intervening to defend the yen — only its third such participation this century, after euros in 2000 and yen in 2011. The Federal Reserve did not participate, and the intervention failed within nine days.
8. AI infrastructure is 0.8% of US GDP and 74–75% of US GDP growth (Q1 2026). Both are true; they answer different questions.
9. Automatic Selective Service registration begins December 2026 under the FY2026 NDAA, covering men 18–25 via DMV, Social Security and Census records — and explicitly does not constitute a step toward conscription, which would require separate authorization.
10. Without synthetic nitrogen fertilizer, best estimates put sustainable world population near 4 billion, not 2 billion. About 48% of people alive depend on it. The dependence is on Haber-Bosch nitrogen globally, not on Gulf exports specifically.
11. Eric Cline’s account of the Late Bronze Age collapse makes the Sea Peoples a symptom: inland empires including Kassite Babylonia, Elam and Assyria fell without ever encountering them.
12. John B. Calhoun’s Universe 25 (NIMH, published 1973) ended in reproductive collapse and withdrawal — not warfare — and the generalization to human societies is rejected by the field.

What would move the verdict, and in which direction. Six of these are events Jiang himself named as decisive; three are his own falsification tests.

Two cautions on using this card. First, several rows would confirm Jiang’s conclusion while leaving his reasoning untouched — a bubble that breaks for ordinary reasons is not evidence of engineered timing, and the card marks that difference. Second, the three decisive rows are his, not ours; a reader inclined to dismiss him should notice that he chose conditions that could plausibly occur, which is not the behaviour of someone protecting an unfalsifiable position.

Thirteen reasoning moves, stated so they travel. Nine appeared in the text above; four are consolidated here. None of them is about geopolitics.

№1 · Split the claim from the register it arrives in.

Confidence is a stylistic property, not an epistemic one. Track provenance, not tone.

№2 · Count the self-corrections.

Retraction behaviour is evidence about good faith, which is a separate question from accuracy and is constantly confused with it.

№3 · Grade citations and quantities separately.

They are different skills, they fail independently, and a single accuracy score hides the difference.

№4 · Ask what the model excludes.

The caveats paragraph is where the unpriced risk lives, and it is the cheapest available edge for a non-specialist.

№5 · Separate the document from the reading of it.

A real citation and a false inference travel together comfortably. Check them as two operations.

№6 · Read the instrument, not the headline.

Registration is not conscription. A threat is not a measure. An invitation is not a visit.

№7 · Check whether the mechanism you borrowed is the part the field kept.

Faithfully reproducing a framework while centring the element it exists to demote produces a conclusion the source does not support, carrying the source’s authority.

№8 · When a structural cause is available, check it first.

Courts, statutes, attrition curves and procurement cycles produce most of what looks like strategy. Reach for the actor only after the process is eliminated.

№9 · Score conditionals on the antecedent.

“If A then B” is void when A never happens. Record it as void.

№10 · Base rates before impressiveness.

“A Republican administration confronts Iran” and “a summit in Beijing in mid-May” are not the same achievement, and a scoreboard that treats them alike is flattering someone.

№11 · The audience is part of the evidence.

Three of the sharpest objections in this corpus came from participants, not analysts — a Russian on Putin, a reader on the Weber thesis, a listener reading the Twelfth Amendment aloud. Record the challenge and the answer; how a source handles being corrected in public is data.

№12 · A verified predicate does not transfer its status to the conclusion.

“The law passed in December” and “therefore young men must fight” are two claims, and the second does not inherit the first’s evidence.

№13 · When someone describes a method for fitting facts to a narrative, believe them.

“Don’t worry also if this grand structure is factually correct or not… then you organize the facts to fit the narrative.” That is not a throwaway line. It is the model’s licence and its limit, stated by its author, and it predicts the shape of every finding in this document.