What a chart is made of, before it is a reading
A Zi Wei chart has three parts and nothing else. There is a ring of twelve palaces, a set of fourteen main stars placed into that ring, and a set of four transformations attached to stars by the year stem and by palace stems. Everything else on a printed chart — auxiliary stars, brightness grades, the twelve growth phases, the decade ranges — is a layer added on top of those three. The reason to separate them is that the first part is arithmetic and the last part is interpretation, and a chart read without that separation cannot tell you which one it is doing. A palace name is a position. It is not a description of a person.
The ring is fixed, and 命宫 only decides where it starts
The twelve palaces sit in one order: 命宫, 兄弟, 夫妻, 子女, 财帛, 疾厄, 迁移, 交友, 官禄, 田宅, 福德, 父母. Nothing reorders them. 命宫 is counted from 寅 by lunar month and then back by hour, and once it lands, the other eleven follow in sequence. Two inputs set 命宫 and neither is the year or the day, which is why two people born decades apart in the same lunar month and hour share a 命宫 palace even when every star in it differs.
The reading set is always four palaces, at four, eight and six steps
The palaces read together with 命宫 are found by fixed offsets: 官禄 at four steps, 财帛 at eight, 迁移 at six. Those offsets are not a convention to be memorised but a consequence of branch relations — four and eight land on the two branches that form a three-harmony combination with 命宫, and six lands on its clash partner. The set is closed at four because no fifth palace relates to 命宫 by the same two relations. One result follows immediately and is easy to check on any chart: 夫妻 sits two steps from 命宫 and therefore directly faces 官禄, in every chart ever drawn. 身宫 adds a second structural restriction — shifted from 命宫 by the birth hour, it can reach only six of the twelve palaces and never lands in one about a relative.
The star ring has twelve possible forms, and empty palaces are part of the arithmetic
The fourteen main stars are not placed one at a time. 紫微 is placed from the bureau number and the lunar day, 天府 is placed as its mirror across the 寅–申 axis, and the remaining twelve sit at constant offsets from those two anchors. Because one position generates the whole ring and that position has twelve values, there are exactly twelve main-star layouts. Fourteen stars into twelve palaces also cannot divide evenly, so every chart carries empty palaces: measured across 1,200 charts, 67 percent have two, 16.5 percent have none and 16.3 percent have four, while 命宫 itself is empty in 16.8 percent. An empty palace is a remainder, and the convention for reading it — borrow from the palace six steps away — is a reading convention, not a transfer of stars.
The bureau is a divisor, not a label for the person
五行局 is the 纳音 element of the 命宫 stem-branch pair, and its name carries a number: 水二, 木三, 金四, 土五, 火六. That number does two jobs. It is the divisor in the star-placement rule, and it is the age at which the first decade begins, which is why two people born on the same day can enter their first decade years apart. It is not a statement that a person is a water type or a fire type. Reading it as an element label for the personality skips the two mechanical roles it actually performs.
This system and BaZi do not share a calendar
BaZi turns the year at 立春 and the month at each solar term. Zi Wei turns the year at the lunar new year and takes the lunar month as its month unit. The two boundaries coincide only by accident, and in the gap they disagree: on 22 January 2023 Zi Wei read 癸卯 甲寅 while BaZi still read 壬寅 癸丑, and on 6 February 2024 the positions reversed, with BaZi on 甲辰 丙寅 and Zi Wei still on 癸卯 乙丑. 命宫 moves on the new moon, not on 立春. Both systems also advance the day pillar at 23:00, but Zi Wei names the hour that straddles the change instead of folding it silently into the next day.
Where to start, and where a reading stops
Start with the ring, not with the stars. If the birth date, hour and gender are ready, the chart tool prints the twelve palaces, the bureau, and the decade ranges, and the first thing worth checking is where 命宫 landed and whether it is empty. From there the structural articles go in order: the palace ring, the empty palaces, the placement arithmetic, and the calendar comparison with BaZi. What none of them support is a claim about a specific outcome. A palace and a star label describe a position in a system; they do not establish health, wealth, relationship quality, or a future event, and a reading that presents them as doing so has moved past what the arithmetic can carry.