Jiang Xueqin Accuracy: More Right Than Wrong So Far
Professor Jiang Xueqin's predictive record is mostly confirmed. As of 2026-07-06, the claims extracted from his @PredictiveHistory lectures that have resolved break 15 confirmed to 5 wrong — a headline accuracy of 75%. That lead rests heavily on a cluster of unusually specific US-Iran war calls that came true in early 2026. The honest caveat is the denominator: only 20 of 340 tracked predictions have settled yet, so this is an early read on a live record, not a final grade.
The headline number
The full denominator is larger than the resolved count suggests. This tracker holds 340 predictions drawn from 171 of Jiang's lectures (170 transcribed), but only 20 have reached a binary verdict. The other 304 sit in pending or unverifiable buckets, awaiting either a deadline to pass or conclusive reporting to land. Headline accuracy counts only the settled calls — confirmed divided by confirmed-plus-wrong — which currently stands at 75%. The live accuracy dashboard updates each time a prediction resolves.
Where he has been right
A disproportionate share of the confirmed column comes from Jiang's US-Iran war calls, which resolved rapidly once the conflict opened in 2026. His April 2024 prediction (P011) that the United States would go to war with Iran inside a two-to-four-year window is the centerpiece — logged when the prospect still looked remote, confirmed when strikes began early the following year. Around it sits a tight ring of specific calls: the bombing of Iranian nuclear facilities (P012), the failure of Shock and Awe (P015), Iran closing the Strait of Hormuz (P026), attacks on Gulf states like Dubai and Bahrain at the war's start (P027), and the decapitation strike that killed Ayatollah Khamenei around March 2026 (P088). Each was timestamped before the event; each was later matched to on-the-ground reporting.
The 2024 election calls show the ledger doesn't tilt automatically toward confirmation. Jiang correctly called a Trump victory (P001), predicted in April 2024, and flagged JD Vance as a strong VP alternative (P002b) a month later, confirmed at the Republican convention that August.
Where he has been wrong
The same election cycle produced a clean miss: Jiang also called Nikki Haley as Trump's running mate (P002a), made in the same May 2024 window as the Vance call. Same topic, same prediction window, one right and one wrong, scored identically. That pairing is the clearest evidence the methodology applies evenly.
Three of the five wrong calls on file are worth naming. Beyond Haley, Jiang predicted that Putin would declare a nuclear umbrella over Iran (P017) — which did not happen — and that Muhammad Mokhber would win Iran's June 2024 presidential election (P040), which he also did not. These sit in the wrong column outright, not reframed as near-misses.
Is Jiang Xueqin reliable?
The base rate so far says more right than wrong, but it rests on a thin resolved base — 20 calls out of 340, with the confirmed column weighted toward one event that resolved in a hurry. Judging predictive history accuracy fairly means waiting for the open ledger to close: a forecaster's true calibration only emerges across hundreds of independent calls and many years, not a single quarter dominated by one war.
The remaining 304 predictions — on China's role in the Iran conflict, on Venezuela, on Iran's succession — are where the record will be built or broken. A few are already confirmed, including China staying on the sidelines of the Iran war (P071), a US attack on Venezuela (P082), and Mojtaba Khamenei rising to Supreme Leader (P041). The rest are still pending. Browse the full prediction ledger to see what's open and what's settled, because the question of whether Jiang Xueqin is right or wrong more often will ultimately be answered there.
Browse all tracked predictions · Accuracy analysis · All blog posts