The pattern school 格局: the other way to read the same eight characters

The balance school asks a quantity question. Is the day master strong or weak, given the season it was born into and what supports it? Once you have the answer, the prescription follows: a weak day master wants the elements that feed and reinforce it, a strong one wants the elements that drain and restrain it. The output is an element and a direction, and it is arithmetic you can watch happen.

The pattern school asks a structure question. Look at the month command, which is the branch of the month pillar and the single most weighted position in the chart. Work out which of the ten gods its dominant hidden stem is, relative to your day master. That names your chart. A chart whose month command produces Direct Officer is a Direct Officer chart, and from there the questions are about the pattern rather than about you: is it intact, what is holding it up, what in the chart is undermining it.

Notice what has been demoted. In the pattern reading, whether the day master is strong or weak is not the opening question and not always the deciding one. A chart can carry a clean, well-supported pattern with a weak day master and be read as a good chart, which is precisely the case where the two schools return different answers.

Eight of the ten gods form standard patterns. The two Companion relations, which are the day master's own element, do not, because a chart cannot be named after itself. If the ten-god vocabulary is new, that table has its own page; this one assumes it.

The Ziping Zhenquan divides the eight in a way that is the heart of the method. Four are favorable and are handled by following them, meaning you support and protect what they do. Four are unfavorable and are handled by countering them, meaning you supply whatever holds them in check. The text puts it in one line: Wealth, Officer, Resource and Eating God are the useful gods that are good, and are used by following; Seven Killings, Hurting Officer, Rob Wealth and the Blade are the ones that are not good, and are used by countering.

Two more terms come with the method. The supporting god, xiang shen, is the character that makes the pattern work: the thing that protects a Direct Officer from the Hurting Officer that would attack it, or the thing that gives a Seven Killings somewhere to discharge. A pattern reading spends more of its attention there than on the day master. And a chart that fails to form any of the eight cleanly is an irregular pattern, of which the follower chart is the one that comes up most often and the one where the two schools diverge most violently, since in a follower chart the favorable elements invert. Our own limits section flags that case on a real public chart.

Here is the thing that makes English-language bazi so confusing, and it is not a translation problem. It is that the central term changed meaning in Chinese, inside living memory, and both meanings are still in circulation.

The term is yong shen, usually rendered as the useful god or the favorable element. In the Ziping Zhenquan, Shen Xiaozhan means by it the dominant qi of the month command: the thing that establishes which pattern you have. He is emphatic about it, and the sentence is worth having in full, because it is a direct shot at the practice most modern readers were taught. Anyone reading a chart, he writes, looks first at what category the useful god belongs to, and to seek a useful god anywhere but in the month command, picking one at will, is to grasp the false and lose the true.

On the modern English internet, and in our own calculator, the same term means something else: the element the chart is short of, chosen after judging the day master strong or weak. That is a balance-school object. Shen would not have called it a yong shen at all.

The shift is traceable to a specific book. Xu Leyi annotated the Ziping Zhenquan in the 1930s, and in the chapter on the useful god he divided its selection into four methods, strengthen-and-restrain among them. That annotation is how most twentieth-century readers met the text, and the four-method framing is what propagated. Chinese Wikipedia's entry on the book says plainly that the annotation misread the original on this point and that the misreading shaped later practice.

We are not in a position to be smug about that, because our calculator uses the post-Xu meaning. The method we implement descends from the same Republican-era lineage, by way of Wei Qianli's 1935 textbook, and our methodology page cites him. When our calculator says favorable element, it means the balance-school object, and a reader coming from a Qing text will find that the word does not point at the same thing. That seemed worth writing down on our own site rather than leaving for someone to discover.

Not because it is more correct. We have no way to establish that and neither does anybody else. The reason is narrower and it is about what can honestly be automated.

The balance reading is fully specifiable. Season weights, root strength, hidden-stem shares, a threshold, an affinity table: every step is a number, every number is published, and anyone can reproduce a year score by hand from the specification and tell us if we got it wrong. That is the only kind of divination tool this site is willing to run, because it is the only kind where the reader can audit the machine.

A pattern reading is not like that. Naming the pattern is easy. Everything that makes the reading worth having comes next: whether the pattern is intact, what is protecting it, what has broken it, whether the chart has slipped into an irregular pattern instead, and those judgments are a large body of rules on which the sources genuinely disagree. A tool that named your pattern and then rendered a verdict would be presenting an unresolved argument as an answer, which is the failure mode the methodology page exists to prevent.

So we compute what we can specify and we print the raw material for the rest. Your chart shows the month command, your day master, and the ten-god relation of every character in it, which is exactly where a pattern reader would start. What we will not do is hand you a pattern verdict our engine cannot defend.

If you want to see the other reading, the sequence is short enough to follow by hand from what our calculator already gives you.

Find the month branch and its dominant hidden stem. Read that stem's ten-god relation to your day stem: our chart labels it. That name is your pattern. Then ask which of the two groups it belongs to, following or countering, and look through the rest of the chart for the character that would support it and the character that would break it. That last step is the reading, and it is the step no page and no calculator can do for you in a sentence.

You may well find it agrees with what our engine told you, because on a great many charts the two schools converge. When it does not, that disagreement is real information about your chart rather than a fault in either method, and it is the sort of chart a practitioner would want to look at properly.

One more school is worth naming, since it comes up alongside this one. Blind school practice, mangpai, is a third lineage, transmitted orally among blind practitioners and working from image and technique rather than from either scoring or patterns. It is genuinely different again, we do not implement it either, and we know less about it than we would need to write this page's equivalent for it.