How to read a BaZi case

What would count as evidence against the story?

A published case gives you the example an author chose to discuss. It rarely supplies every similar chart left out of the book. Before asking why the reported outcome happened, ask what comparisons are missing. Without that denominator, selected stories cannot provide a predictive success rate.

There is a second gap: when was the interpretation recorded? An explanation written after an event has different evidential value from a prediction preserved beforehand. If the source does not resolve that question, record the uncertainty. An elegant explanation does not supply its own date.

These questions do not prevent close reading. They distinguish learning an author's reasoning from testing whether that reasoning predicts lives. The two cases below are useful for the first task, with clear limits on the second.

Keep the two records intact

In the Direct Officer chapter of Ziping Zhenquan, two configurations illustrate an argument about Wealth and Resource. Here they are labelled in source order. Neither the people nor their reported ranks have been independently verified for this article.

Start the comparison by marking everything that differs. The Day Master changes from Yin Wood to Yang Earth; the month and other placements also change. This is not a controlled experiment in which only one combination varies. A difference in reported rank cannot be assigned solely to that combination on the strength of these two records.

Do not repair this problem by inventing two matched charts. That would produce a new teaching exercise, not improve the historical evidence. Likewise, an inconvenient stem must not be silently changed to fit an interpretation. Keep variant readings and editorial corrections separately, with their edition.

Separate the label from the explanation

For the Yin Wood Day Master in example one, month-stem Yang Water is Direct Resource and hour-stem Yang Earth is Direct Wealth. The source emphasizes Yin Wood between them and interprets the placement as avoiding interference. The copied positions, the Ten-God classifications and that interpretation are three distinct claims.

For the Yang Earth Day Master in example two, year-stem Yang Water is Indirect Wealth and month-stem Yin Fire is Direct Resource. The passage emphasizes the Yin Fire–Yang Water combination in its account of the arrangement. Identifying the traditional pairing does not establish transformation, nor prove a restriction on a person's career. See combination and transformation for that distinction.

This lets you state what the passage actually does: it uses relative position and a combination to explain its different assessments. You can reconstruct that argument while leaving its real-world conclusion unverified. Agreement with the vocabulary is not agreement with every resulting claim.

Run the checks that the supplied fields permit

The Hong Kong Observatory's month and hour tables support four local checks: a Yang Wood year gives Yang Water in the Monkey month; a Yang Water year gives Yin Fire in the Goat month; a Yin Wood day gives Yang Earth in the Tiger hour; and a Yang Earth day gives Yin Wood in the Rabbit hour. Both written records agree with those correspondences.

That result has a narrow scope. It does not identify a complete civil date, recover a birthplace, establish a person's identity or verify an official rank. Write “the listed year–month and day–hour pairs agree with the tables,” rather than “the case is verified.” Missing birth details remain missing even after a successful lookup.

Open the full-size diagram

Original BaZiFlow case-reading sequence. Arrows indicate review order, not an established predictive cause.

The diagram is a review sequence. Its arrows do not turn a successful transcription into outcome evidence. Each later claim needs support appropriate to that claim; the same calculation cannot do every job.

Preserve what you thought before reading the ending

For a first reading, cover the reported outcome and save a note about the supplied data and the argument you can reconstruct. After uncovering the ending, add a separate note. You can then see which explanations appeared only after you knew the result, instead of overwriting the earlier uncertainty.

Keep four categories in that record: source information, rule checks, interpretive claims and outcome evidence. For these examples, a defensible summary says that the passage compares Wealth and Resource arrangements, four local table checks agree, multiple features differ, and identity and rank remain independently unverified.

The guide to classical texts and editions explains how base text and later commentary can share a page. Use that distinction alongside the four categories. A memorable ending is worth examining, but it must not make the missing evidence disappear.

1. 1

   Write before seeing the outcome

   Save the pillars, edition and missing information with the reported ending covered.

2. 2

   Label the evidence

   Record each table check separately from the classification and the author’s explanation.

3. 3

   Add a second note afterwards

   Keep any explanations added after the ending visible; retain unverified identity and rank as open questions.