Skip to content
明 Ming Review

The question we get most

How accurate is Zi Wei Dou Shu? An honest answer, in two parts.

Every tradition gets asked this, and almost every tradition answers it dishonestly — with a testimonial or a shrug. The honest answer is that 'accurate' hides two different questions, and a serious journal owes you both answers separately.

By Lim Zhiming · Founding Editor · November 20, 2026

Part one: computational accuracy — solvable, and usually botched

A Zi Wei chart is a deterministic object. Given a birth moment, a rule set determines the palace grid, the star placements, the brightness grades, the decade floors — every field, no wiggle room. That means accuracy in this layer is a solvable engineering problem, and it means the popular tools fail it in known, nameable ways. The year boundary set at the festival instead of Start of Spring — wrong for every birth in the six-week gap. The clock read as civil time instead of solar time at your longitude — wrong whenever longitude correction pushes you across a two-hour pillar. A daylight-saving table that quietly stops covering the 1980s. Each of these errors is invisible to the reader and fatal to the chart. This layer is why our method reads like an engineering document: it is one, and its entire job is to make the first question — is this chart the right chart? — answerable with a yes you can verify.

Part two: interpretive accuracy — the honest kind of unknowable

The second question — does the tradition's reading of a chart describe a real life? — is a claim about symbolism, not arithmetic. No divinatory system can demonstrate it in the scientific sense, and the traditions that insist otherwise are selling certainty they do not possess. What a tradition can offer here is older and quieter: an internal consistency checkable against its own classical literature, a centuries-long record of readers finding the sentences useful, and a transparent register that states what the system does not know. Our reports carry classical citations on their face for exactly this reason — so the interpretive layer can be audited against its sources, not against our enthusiasm.

What we refuse to predict

Accuracy also lives in what a system declines to say. Our house rules exclude health predictions, death timing, and financial or legal instruction, in every register, without exception. The classical corpus contains passages that read as such forecasts; we quote them as historical claims with context, never as personal predictions. If a reading from any tradition tells you when you will fall ill or where to put your savings, the prediction says more about the reader's ethics than about your chart.

The audit trail, in your hands

Everything above is checkable, and we want it checked. Cast a free chart and read the footnotes: the dual-calendar year window, the solar-time correction, the late-hour handling each appear where they apply. Open a worked report and follow a citation to its classical source. Disagree with our reading of the rules — the method page names them one by one, and we would rather be corrected in public than silently wrong. A journal that cannot survive its own footnotes is not a journal, and accuracy begins with meaning it.

Keep reading

Frequently asked

Is Zi Wei Dou Shu scientific?

No, and we will not insult you by pretending otherwise. It is a symbolic tradition with a rigorous internal arithmetic — the computation of a chart is checkable in the way a horoscope column is not — but its interpretive claims are not scientific predictions. What science can fairly ask of any divinatory system is that it be honest about its inputs, consistent in its methods and transparent about its limits. That bar we meet on purpose, and it is higher than the bar most popular astrology clears by accident.

Why do different websites give me different charts for the same birth data?

Almost always one of four computational faults: the year boundary set at the festival instead of Start of Spring, the month boundaries set by lunar month instead of solar terms, civil clock time used where solar time at your longitude was required, or a daylight-saving table that is wrong for your decade. Any one of these silently moves a palace or a star. The interpretive tradition is old; the arithmetic is exact; the gap between them is where bad tools live.

Can it predict illness, death or disasters?

Our house rules: no. The classical literature contains passages that read as such predictions, and a responsible modern desk quotes them as historical claims with their context stated — never as personal forecasts. Health passages are written in a wellness register throughout: constitution and rhythm as tradition describes them, self-care as suggestion, and a clear valve to qualified practitioners wherever anything sounds clinical. Every reading we ship follows those rules, and the citation badges on its face show where each passage came from.

What does "accurate" even mean for a birth chart?

Two different things share the word, and the field is cleaner once you split them. Computational accuracy: does the chart reflect the birth data through the tradition's rules, exactly? That is a solvable problem, and it is where most sites fail. Interpretive accuracy: does the tradition's reading of the chart describe a real life? That is the symbolic claim, the one no system can prove and honest systems admit. We engineered the first so you are free to weigh the second.

How do I check your calculation for myself?

Three ways. Read the written method, which names every rule we apply and the order we apply them. Cast a free chart and look at the calendar footnotes — where the two legitimate year boundaries disagree, we print both and tell you which window you were born in. And compare our chart against a classical text you trust: the star placements and palace assignments should match exactly, because they are rule-determined, not vibes-determined.

Watch the method work before you pay for it

One letter a month walks a real chart through this exact pipeline — corrections, dual-calendar footnotes and all. Free, double opt-in, no fortune-telling inside.