Jiang Xueqin Wrong Predictions: Nikki Haley VP and Other Misses
As of 2026-11-06, Jiang Xueqin has five wrong predictions in this tracker out of 463 total calls. The best-documented miss in the current extract is P002a, his May 2024 call that Donald Trump would choose Nikki Haley as his vice president. The live ledger of every wrong entry is at the wrong-predictions page.
The clearest miss: Nikki Haley as Trump's VP (P002a)
P002a was a near-term, binary call. Jiang predicted in May 2024 that Trump would pick Nikki Haley as his running mate, with the August 2024 Republican Convention as the timeline. That failed when Trump announced JD Vance as his vice-presidential pick in July 2024. The same cycle produced P002b, Jiang's May 2024 call that Vance was a strong alternative. That prediction was confirmed when Trump selected Vance. One VP call landed; the Haley call did not.
P002a is the cleanest miss in the sample: it named a person, a date, and an event, and the event contradicted it.
What five wrong calls do and do not mean
The scale matters. As of 2026-11-06, the tracker counts 17 confirmed predictions and 5 wrong predictions. That leaves 415 still pending or unverifiable. Only 22 predictions have resolved in either direction. Headline accuracy is 77% — 17 confirmed divided by those 22 resolved calls — but that number is provisional. The denominator is tiny, and most of Jiang's catalog remains open.
That caveat cuts both ways. Five misses are not enough to declare a pattern of failure. Seventeen confirmed calls are not proof of infallibility. The confirmed column includes P001, the call that Trump would win the 2024 election; P011, that the US would go to war with Iran; P012, that the US would bomb Iranian nuclear facilities; and P026, that Iran would close the Strait of Hormuz. Those predictions were confirmed when news reporting matched the events. They sit alongside P002a in the same ledger.
Why the wrong column is short
The wrong column is short because most predictions have not had time to fail or succeed. It is also short because a miss is counted only when an event contradicts the call. P002a specified a person and a date; the actual pick was someone else. That is a miss.
By contrast, P018 — that the US-Iran war will last many weeks, possibly many years — was not wrong just because the war continues. It was confirmed when the conflict extended beyond a short bombing campaign. The same logic applies to P020, on attacks against Iran's civilian infrastructure, and P027, on Iran attacking GCC countries when the war began. Those calls were confirmed when reporting described the events. The tracker does not retroactively call a prediction wrong because a later phase looks different.
This is where predictive history misses become useful. A predictive framework should be judged on calls made before events, not on ones rewritten after them. Jiang's method is broad and often long-range, so many predictions remain unresolved for years. That makes the small wrong column less impressive than it looks and less damning than critics might hope.
How to read Jiang Xueqin failed predictions
The live wrong-predictions ledger is the authoritative list. It currently contains five entries. The extract available for this post names P002a. The other four are cataloged there, and they should be read with the same standard: what exactly was predicted, what was the timeline, and what event contradicted it.
Readers should separate wrong predictions from pending ones. A failed prediction is not the same as a prediction that has not yet come true. The tracker lists 415 calls as still pending or unverifiable. Calling any of those wrong now would be fabrication, not analysis.
The purpose of tracking Jiang Xueqin wrong predictions is not to dunk on a lecturer. It is to keep score in public. A tracker that hides misses is a fan page. A tracker that lists them, dates them, and links to the full ledger is a measurement tool. As of 2026-11-06, that tool records five misses, 17 confirmed calls, and 415 open questions. The honest headline is not that Jiang is always wrong or always right. It is that the sample is still small, the misses are real, and the live ledger is where the next ones will appear.
Browse all tracked predictions · Accuracy analysis · All blog posts