Is Jiang Xueqin Credible? The Track Record Says Mostly Yes

Yes — with an asterisk. As of August 27, 2026, 22 of the 382 predictions this tracker has extracted from Jiang Xueqin's YouTube lectures have resolved. Sixteen were confirmed and six were proved wrong — a 73% hit rate on settled claims. That is a strong number in geopolitical forecasting, a field where war-and-succession calls routinely humiliate the people who make them. But an honest answer to "is Jiang Xueqin credible" carries two caveats: the resolved sample is small, and it leans heavily on the one event — the US-Iran war — he spent two years preparing his audience for. Both sides of the ledger are below.

What he got right

The signature call was the war itself. In April 2024, Jiang predicted the United States would go to war with Iran within two to four years (P011). The war began in March 2026, inside his window — and once it started, the granular calls resolved his way in quick succession. The US bombed Iran's nuclear facilities (P012). A decapitation strike killed Ayatollah Khamenei around March 1, 2026 (P088). The Iranian navy was destroyed (P105). Iran closed the Strait of Hormuz (P026) and attacked GCC countries (P027), and strikes on civilian infrastructure escalated (P020).

He also called the war's shape, not just its outbreak: the shock-and-awe doctrine would fail (P015), the conflict would run from many weeks to possibly many years (P018), nuclear weapons would not be used (P016), and China would not intervene meaningfully on Iran's side (P071). Add his May 2024 call that Mojtaba Khamenei would succeed his father as Supreme Leader (P041), and the Iran file reads as a prepared forecast rather than a lucky headline.

The non-Iran record is thinner but real. He picked Trump to win the November 2024 election back in April 2024 (P001), flagged JD Vance as the running mate the Republican convention would land on (P002b), called a US attack on Venezuela on a short timeline in December 2025 (P082), and even landed a technology call — AI companies rolling out AI girlfriends and companions to drive adoption (P129).

What he got wrong

The misses stay on the page, permanently. The cleanest is his May 2024 prediction that Trump would pick Nikki Haley as his running mate (P002a); the convention chose Vance instead. Worth noting: Jiang had also flagged Vance that same month (P002b) — on personnel picks he hedged, and both halves get scored.

The more substantive miss is his May 2024 claim that Putin would declare a nuclear umbrella over Iran, deterring US nuclear use (P017). The umbrella never came. The outcome Jiang braced his audience for — no nuclear use in the war — did hold (P016), but the mechanism he named did not, so the prediction was scored wrong. Binary scoring judges what was actually claimed, not just how events felt. Every settled miss, with its sources, is listed openly on the wrong predictions page.

The small-denominator problem

Here is the statistical fine print. Only 22 of the 382 tracked predictions have resolved; 343 are still pending or parked as unverifiable. Most of what settles early settles because the Iran war fired roughly on his schedule, which concentrates the confirmed column in his strongest subject. If the pending majority breaks against him, the hit rate falls. A 73% figure on 22 observations is a running estimate with wide error bars, not a fixed character trait — anyone asking whether Predictive History, his lecture channel, is a credible source should treat the current score as a mid-experiment reading. That is why this page regenerates every few weeks as predictions resolve.

Why the misses stay on the page

Credibility in forecasting is not the absence of misses; it is the willingness to be scored on all of them. Is Jiang Xueqin reliable? On the current record: more reliable than not, wrong often enough to stay interesting, and documented tightly enough — every claim timestamped to a lecture, every resolution checked against fresh reporting — that you can verify him rather than take his word for it. The live score and the complete methodology sit on the accuracy page. Judge him there, not on anyone's summary. Including ours.

Browse all tracked predictions · Accuracy analysis · All blog posts