Predictive History Explained: Jiang Xueqin's Framework for Decline
Predictive history is the forecasting framework behind Jiang Xueqin's lectures. Its core claim is that geopolitics is less a chain of individual decisions than the visible output of structural pressure — fiscal exhaustion, elite competition, demographic squeeze, overextended militaries. Those forces move slowly enough to be read years ahead of the events they produce. History, in this telling, does not repeat. It rhymes, and rhyme is forecastable.
Three premises
Rhyme over repetition. Similar structural conditions produce similar families of events. Not identical events — the same pressures expressing themselves through whatever institutions happen to exist at the time. This is why a forecast can be right about the shape of a conflict and wrong about its cast.
Decline has a shape. Civilizations under strain pass through recognizable stages: a widening gap between elite expectations and available positions, the shift from productive investment to rent-seeking, the turn to external conflict as domestic legitimacy frays. Jiang treats these as diagnostic markers rather than moral verdicts. They can be dated, roughly, and once dated they constrain what comes next.
Structure over personality. Leaders are read as expressions of constraint, not authors of outcomes. A Khamenei or a Trump or a Putin occupies a position the structure created and operates inside limits the structure set. The method therefore spends little time on psychology and a great deal on fiscal capacity, alliance geometry, and the price of oil.
Why the calls arrive in clusters
If the structural read is correct, a lot follows at once. Jiang's April 2024 call that the United States would go to war with Iran (P011), inside a two-to-four-year window, came bundled with downstream claims: that Washington would bomb Iranian nuclear sites (P012), that a "Shock and Awe" opening would fail to force capitulation (P015), that Tehran would close the Strait of Hormuz (P026) and strike Gulf states (P027), that the fighting would run weeks or years rather than days (P018), and that nuclear weapons would not be used (P016). One structural judgment generated more than a dozen independently scoreable claims. That is a feature of the method — and a headache for anyone trying to grade it, since the claims are correlated rather than independent.
How the method is scored
As of 2026-10-08, the tracker holds 446 predictions drawn from 182 source lectures, 181 of them transcribed. Seventeen are confirmed, five are wrong, and 399 remain pending — a mix of long-horizon calls and claims too loosely specified to check against a news record. That leaves 22 resolved cases and a headline accuracy of 77%.
Treat that figure gently. On a 22-case sample, the standard error is roughly nine percentage points, so the defensible reading is "more right than wrong, with a wide interval" rather than a precise hit rate. The pending pile is the larger caveat: most predictive-history claims are dated years out, so the method's real test is still accumulating, and the number will move as it does.
Where structure beat personality — and where it didn't
The Iran cluster is the strongest current evidence. The prediction that a decapitation strike would kill Ayatollah Khamenei (P088) and that the Iranian navy would be destroyed (P105) both landed inside the 2026 war. Mojtaba Khamenei's elevation to Supreme Leader (P041), called in May 2024, matched. The call that China would stay out of the conflict (P071) held. Separately, the December 2025 forecast that America would attack Venezuela (P082) and the February 2025 expectation of Trump trade wars against Canada, Mexico and likely the European Union (P325) both resolved in Jiang's favor, as did Putin's visit to China shortly after Trump's trip there in May 2026 (P315).
The misses matter just as much. In May 2024 he named Nikki Haley as Trump's running mate (P002a). The pick was JD Vance — whom he had also flagged as a strong alternative (P002b). One of the two landed. A structural framework can identify the pressure toward an outsider-populist ticket without reliably naming the person, and the Iran record shows the inverse: right about escalation, right about the Strait, right about the succession, with much of the war's middle still pending.
How to read a predictive-history claim
Use the framework as a map of pressure, not a schedule. A dated call is a claim that conditions have matured to the point where an event family becomes likely — not that it will arrive on a particular Tuesday. Three practical habits help:
- Check the horizon. A two-to-four-year window is a genuine forecast; a "soon" is not.
- Check the specificity. "War with Iran" is easier to confirm than "the war goes badly for Washington."
- Check what would falsify it. Jiang's strongest calls, like the Strait of Hormuz closure (P026), carried clear defeat conditions.
The predictive history YouTube archive is organized by source lecture, so each claim can be traced back to the talk it came from and read in its original context — which is the only fair way to judge a method built on long arcs rather than headlines.
Browse all tracked predictions · Accuracy analysis · All blog posts