Jiang Xueqin Accuracy: More Right Than Wrong, but the Sample Is Small
Jiang Xueqin has been right far more often than wrong on the predictions that have resolved, but only a small slice of his forecasts has resolved. As of 2026-10-25, 17 of his predictions are confirmed and 5 are wrong, a headline accuracy of 77% (confirmed out of confirmed plus wrong). That is a promising record built on 22 resolved calls, out of 450 tracked.
The live tally
Here is where the numbers stand as of 2026-10-25, drawn from 183 source lectures (182 transcribed):
- Total predictions tracked: 450
- Confirmed: 17
- Wrong: 5
- Still pending (not yet resolved or unverifiable): 403
- Resolved so far: 22
The 77% figure is the resolved base rate. It is not a measure of all 450 predictions, and it should not be read that way. About 90% of the tracked set has no verdict yet. Some of those calls have long time horizons. Others are too vague to check.
Anyone asking whether Jiang Xueqin is reliable should hold two facts together: the resolved record is good, and the resolved record is short.
Three confirmed calls worth examining
The 2024 election. In April 2024, Jiang predicted Donald Trump would win the presidential election (P001). The prediction was confirmed when the November 2024 result was reported. In May 2024 he also called JD Vance a strong alternative vice-presidential pick (P002b), which was confirmed at the Republican Convention in August 2024.
War with Iran. In April 2024, Jiang predicted the United States would go to war with Iran, on a timeline of two to four years (P011). That was confirmed by the 2026 conflict, well inside his window. Several related calls resolved in the same direction:
- The US would bomb Iranian nuclear facilities (P012).
- Iran would close the Strait of Hormuz (P026), a call he made in summer 2025.
- Iran would attack Gulf states when war began (P027).
- Mojtaba Khamenei would become Supreme Leader (P041), a May 2024 call.
Trade wars. In February 2025, Jiang predicted that Trump would launch trade wars against Canada, Mexico and likely the European Union (P325). It was confirmed once those tariff actions were reported.
The misses
The record includes five wrong calls. The clearest example in this dataset is his May 2024 prediction that Trump would pick Nikki Haley as his running mate (P002a). Trump did not. The same lecture period produced the correct Vance call (P002b), which shows how a single forecaster can land on both sides of one question.
Misses matter because they are what separate a track record from a highlight reel. This page counts them in the same ratio as the hits.
Why the 77% needs caveats
Small sample. With 22 resolved calls, one or two outcomes moving would shift the percentage noticeably. A 77% rate on 22 cases carries wide uncertainty. It does not establish a long-run hit rate.
Clustering. Many confirmed calls belong to one story, the US-Iran war. Predictions such as P011, P012, P015, P018 and P020 are not independent. If the underlying thesis was right, many calls resolve together. If it had been wrong, many would have failed together. Fifteen confirmations from a handful of themes are weaker evidence than fifteen unrelated successes.
Some calls are easier to verify than others. Several confirmed items are statements about what would not happen, such as nuclear weapons not being used in the Iran war (P016) and China not significantly participating (P071). These were confirmed as the conflict unfolded, but they are lower-risk calls than a specific, dated event.
Some entries were logged after the fact. A few confirmed items, such as the decapitation strike on Ayatollah Khamenei (P088) and the destruction of the Iranian navy (P105), carry no recorded prediction date. They are tracked as consistent with his framework, and readers should weigh them accordingly.
Pending is not wrong. The 403 unresolved predictions could move the number in either direction. Predictive history accuracy is only meaningful once a much larger share has resolved.
What this means for judging Jiang
On the evidence so far, the answer to "Jiang Xueqin right or wrong?" is mostly right on what has resolved, with the strongest results on US-Iran war questions and the 2024 election. Several of those calls were made years before the events, which is harder than forecasting weeks ahead.
That is not proof of a reliable forecasting method. A fair reading is that the early results are encouraging, the sample is small, and the next few dozen resolutions will say more than the first 22.
You can see every call, its date and its current status on the full predictions list, and the methodology and running totals on the accuracy page. This post is updated every three weeks as more predictions resolve.
Browse all tracked predictions · Accuracy analysis · All blog posts