Historical predictions usually fail for one of two reasons: the facts were wrong, or the model was too simple.
In Professor Jiang Xueqin’s Predictive History framework, the second problem matters most. People can see a military buildup, an economic crisis, or a political movement and still misunderstand what it means. They identify the event but miss the incentive.
Good analysis asks a harder question: what would each actor be rewarded for doing next?
That question matters because leaders do not make decisions in a vacuum. They respond to domestic pressure, alliance commitments, resource limits, institutional habits, ideology, and fear of losing credibility.
Bad predictions often come from treating history like a straight line. A country did something before, so it will do it again. A leader said something publicly, so that must be the real objective. A trend has continued for five years, so it must continue for five more.
History rarely behaves that cleanly.
The better Predictive History approach is to separate signal from noise:
- Signals come from incentives, capabilities, geography, and institutional constraints.
- Noise comes from speeches, temporary sentiment, and dramatic but isolated events.
- The strongest warnings appear when multiple signals point in the same direction.
Prediction is not certainty. It is disciplined uncertainty: a way to judge which outcomes become more likely when history, institutions, and incentives point in the same direction.
The point is to improve the quality of the question before the answer becomes obvious. That is why History Predicted treats Jiang’s lectures as a corpus to study, not just content to repost.