Hi, I'm Bokki!

The birthplace field, and how often it changes the answer

This page's birth form asks where you were born, and it offers forty-four places — a field that does not appear at all until you have given a birth time, because the correction it feeds needs both. Picking one of them is not a preference; it is arithmetic that lands on your chart. A time zone keeps its clocks by a single line of longitude, and the earth turns one degree every four minutes, so the sun over your city reaches its highest point some minutes off what the clock said when you were born. Bokki corrects for that before it builds the hour pillar. The forty-four rows below sit on only thirteen zone meridians between them, and seventeen of the rows share a single one.

The longitude term is the only part of this that asks where you were born, and it runs the whole width of the table: Madrid at −74.8 minutes to Tokyo at +18.8, a spread of 93.6 minutes against a branch that is 120 minutes wide. Of the rows, thirty-four push the clock back and ten push it forward; the number sitting exactly on their own meridian is zero.

The second correction does not ask where you were born at all. The earth's orbit is an ellipse and its axis is tilted, so the sun runs a little fast in some seasons and a little slow in others, and across a year that term rounds to somewhere between −15 and +16 minutes — the same swing for every row here. Which is why the row with the smallest longitude term, London at −0.5 minutes, with the meridian it keeps time by running through the city itself, still moves the hour branch for 6.0% of birth minutes. Picking the place nearest its own meridian is not the same as picking nothing.

Walk all forty-four of them across every day of the year and the largest shift any place can produce is 89 minutes — Madrid, on February 3. A branch is 120 minutes wide. So the number of places in this table that could carry you two branches away from your clock time is zero: whatever is uncertain about your hour pillar, it is a question between two neighbouring branches and never three, and no birthplace on this list can widen it.

There is a second thing the correction does, and it reaches further into the chart than the hour ever does. When the shift crosses midnight the date itself rolls, and the day pillar rolls with it — the day master, which the saju readings here start from. It is much rarer than the hour effect: Madrid rolls the date for 5.19% of birth minutes, London for 0.50%. But when it happens, the whole reading starts from a different day master.

In thirty-nine of the forty-four, those are minutes the hour pillar never notices. The shift there never reaches far enough for both to happen at once, so a birth close to midnight lands on the other side of it — still inside the same two-hour branch, still drawing its stem from the same day — and the day pillar moves while the hour pillar stays exactly where it was. The places that never do both top out at 51 minutes; the five that do — Madrid, Singapore, Istanbul, Buenos Aires and Paris — start at 65.

Both corrections are applied only when a birth time and a birthplace are both given, so leaving this field blank is not the same as choosing the row closest to zero. Blank means your clock time is used exactly as written, with no shift at all, while even London moves by as much as 16 minutes on some days of the year. And each row carries one standard offset and nothing else — no daylight saving, no offsets that have changed since. A wrong correction is worse than none, which is why we would rather ask than guess your longitude.

The two terms, worked through in full