Saju and bazi — what is the same and what is not?

Two words, one count

A saju chart and a bazi chart are the same object counted two different ways. Four columns, two characters in each. Count the columns and the number is four; count the characters and the number is eight. Neither count leaves anything out that the other one includes.

This site's own result screen does both without remarking on it. It prints the four pillars as four columns, and a few lines further down it counts the five elements across the eight characters. Two labels, one result, and nothing between them that has to be converted.

What is already settled before either word applies

Ten stems and twelve branches run in fixed order and pair off into a cycle of sixty. A birth moment is placed into that cycle four times over — once for the year, once for the month, once for the day, once for the hour — and the eight characters are what the placing produces. None of that arithmetic asks which of the two words the reader arrived by.

The parts most likely to be held up as a difference are not new inputs either. Values read off a chart rather than entered into it — the void branches, the twelve stages, the ten gods, the sound element — were checked here against two different birth times that produce the same eight characters, and all four came out identical down to the character. Where the eight characters are the same, those values are the same, and there is no opening for a second tradition to put something different in.

This site does not print those four. What it prints is the eight characters, the element counts across them, and the rules it used to get there. The names those values go by, and how much weight each one is given, belong to the reading rather than to the chart.

Where results actually diverge, and it is not by country

Three things do change a result, and none of them is settled by the data. Which clock the solar term boundaries are read on. Where the day is taken to begin around 23:00. Whether a true solar time correction is applied. Every calculator has to pick all three, and calculators pick differently.

These are schools of practice rather than borders. The split at the day boundary shows up inside a single calendar library, which returns one day pillar when the day is taken to start at midnight and a different one when it is taken to start at 23:00. One library, both answers, and no country anywhere in it.

The clock question is the one that looks national and is not. The choices on this screen are the clock in force at the place of birth, a fixed Korean standard time frame, and a UTC+8 frame. A birth moment far from a solar term boundary comes out the same under all three. A birth moment close to one does not, and then the year and month columns are what move.

What this page does not settle

Whether the two traditions differ in their calculation rules is not settled here. The primary sources that would settle it were looked for and not found, and this page does not fill the gap with a guess in either direction: not that the two are identical, and not that either one came first.

The same limit is applied to this site's own three settings. They were picked as the least weakly supported options rather than verified, and the screen says exactly that beside the result. What can be shown is which rule put each column where it is, and what a different rule would have put there instead. That is the whole of what this page claims.