How We Calculate

How we calculate your saju

Saju turns the moment you were born into four "pillars": year, month, day, and hour. Each pillar has two characters, one of ten Heavenly Stems and one of twelve Earthly Branches. That makes eight characters in all, which is where the name saju palja ("four pillars, eight characters") comes from. The stem of your day pillar is your Day Master, the character that stands for you. This page explains the choices our calculators make, and which ones you can change.

1. Your birth moment, in UTC

You enter your birth date, hour, and minute as the clock showed them where you were born, and choose your birth city (or time zone). Using the international time zone database (IANA), we look up what that clock meant on that exact day, including daylight saving time, and convert your birth to a single moment in Coordinated Universal Time (UTC).

This matters for Korea, whose clocks have changed over the years. From 1954 to 1961, Korea's standard time was UTC+8:30, not UTC+9, and Korea used daylight saving time in several years, including 1987 and 1988. The time zone database includes these changes, so a clock time from those years is converted correctly.

If your clock time didn't exist (the hour skipped when daylight saving time began) or happened twice (when it ended), we tell you which reading we used.

2. The year starts at Ipchun; months follow the solar terms

In saju, the new year doesn't begin on January 1 or on Lunar New Year. It begins at Ipchun (the start of spring, around February 4), one of the 24 solar terms. Each saju month also begins at a solar term, not on the 1st of a calendar month.

We convert both your birth time and each solar term's start time to UTC and compare them to the minute, so a birth in Seoul, New York, or anywhere else is compared with the same astronomical moment. If you were born between January 1 and Ipchun, your saju year, and your zodiac animal, may be the previous year's.

Your Korean zodiac animal follows the same rule: we switch animals at Ipchun, not at Lunar New Year (Seollal). Some calendars and websites switch at Lunar New Year instead, so if you were born between those two dates, your animal may differ from what you've seen elsewhere.

3. Your day and hour pillars: local solar time

Saju hours follow the sun, so for your day and hour pillars we adjust your clock time to the average sun time of your birth city. First, we take out daylight saving time, because it is a clock rule, not a change in the sun. Then, by default, we apply a longitude correction (local mean solar time):

• correction in minutes = (your city's longitude − your time zone's standard meridian) × 4
• The standard meridian is the longitude your time zone is based on. For UTC+9 it is 135°E; for UTC+8:30 (Korea, 1954 to 1961) it is 127.5°E.
• Example: Seoul is at about 127°E, so a Seoul birth is read about 32 minutes earlier than the clock showed. New York is about 1° east of its meridian (75°W), so a New York birth is read about 4 minutes later.

Each city's longitude comes from its published coordinates on Wikipedia and Wikidata. Our list covers 93 cities in 13 countries, including Korea, the United States, Canada, the United Kingdom, Australia, and Japan. Each city uses one central point, so the far edges of a large city can differ by a minute or two. If your birth city isn't on our list, choose "Somewhere else" and pick your time zone: we then use your time zone's standard time without the longitude adjustment, and your result says so. For a closer result, choose the nearest listed city.

The correction changes only your day and hour pillars. Your year and month pillars and your luck cycles come from the UTC comparison above, so the correction never changes them. You can choose one of three options:

• Yes (recommended): the longitude correction above. This is the default.
• True solar time (reference): also adds the equation of time (the sun running a little fast or slow through the year). We offer this for comparison; it is not our default.
• No, use my clock time: your standard clock time, with daylight saving time taken out.

Both public calculators we checked our results against offer a longitude correction in their default mode. One also adds the equation of time by default; the other does not, judging by its results. We offer the equation of time as an option, not as our default.

Why the equation of time is off by default: through the year the sun runs up to about 16 minutes ahead of or behind a steady clock. Some Korean saju calendars correct only for birthplace longitude (local mean solar time), and we follow that convention. It matters only if you were born within about 16 minutes of a two-hour boundary. If you want it included, choose "Use true solar time"; some schools prefer true solar time, and neither approach is the only correct one.

Because of this correction, a birth close to the start of an hour can land in the previous hour. Your result shows the corrected time we used.

4. The late-night Rat hour (corrected time 11:00–11:59 PM): you can choose

The Rat hour (jasi) runs from 11 PM to 1 AM. Traditions differ on whether its first hour belongs to the next day. If your birth time after the birthplace correction falls between 11:00 PM and 11:59 PM, you'll see a choice:

• Count my birth as the next day (default): your day pillar is the next day's, and your hour pillar is the Rat hour, with its stem worked out from the next day's day stem.
• Keep my birth on the same date: your day pillar stays on your birth date, and your hour pillar is still the Rat hour, with its stem worked out from the next day's day stem. This follows the yajasi ("late-night Rat hour") view, one common approach; schools and calculators differ. For a birth just after midnight (12:00–12:31 AM in Seoul), "Keep" gives the previous calendar day's pillar, because the corrected time is still before midnight.

Because the choice uses your corrected time, a Seoul birth at 11:10 PM (corrected to about 10:38 PM) won't see it, while a birth just after midnight, such as 12:20 AM in Seoul (corrected to about 11:48 PM the day before), will.

Neither method is the only correct one; both are used in practice. The two public calculators we checked our results against each handled this hour differently from our default, and from each other, when we tested in October 2026. Another site may therefore show a different day pillar for the same birth. By default, we treat the whole Rat hour (11 PM to 1 AM) as the start of the new day.

5. "Male / Female" and the direction of your luck cycles

Your chart also has 10-year "luck cycles" (Daeun), ten-year chapters that each carry their own pair of characters. In the Full Saju Reading you'll see a field called Male / Female, with two choices: Male or Female. In our calculator, this choice is used only to set the direction of your luck cycles, and it doesn't rate anyone. It doesn't change your four pillars or your Day Master; it changes only your luck cycles and what is built on them: their starting age, your most active decades, and your life stages.

A yang year stem (Gap, Byeong, Mu, Gyeong, Im) with Male, or a yin year stem (Eul, Jeong, Gi, Shin, Gye) with Female, means your luck cycles move forward. The other two combinations move backward. Your year stem follows the Ipchun-based year.

6. When your luck cycles begin

We measure the time, to the minute and in UTC, from your birth to the next solar term that starts a month (forward) or back to the previous one (backward). Then we convert it with the traditional ratio: 3 days = 1 year, so 1 day = 4 months and 2 hours = 10 days. We show the starting age as your age in completed years (not Korean age), rounded: 0.5 and above rounds up. Some traditional calendars count differently, so results may differ by a year. Each later cycle begins 10 years after the one before.

Example: a gap of 7 days and 12 hours is 7.5 ÷ 3 = 2.5 years, shown as age 3.

7. Your most active decades and life stages

In the Full Saju Reading, we mark the 10-year luck cycles that fit your chart most smoothly as Most active in the next 30 years (our score). We score each cycle by whether its two characters bring the elements your chart is short on, and we add or subtract a little when a cycle combines or clashes with key branches in your chart; seasonal balance counts only when it points the same way. This is our own scoring, not a traditional rule, and the scores themselves aren't shown on screen.

We only highlight cycles that end within 30 years of your current age, including the one you're in now, so the reading stays useful for the years ahead. If your best-fitting cycle comes later than that, we simply say it comes later in life. Your full cycle table also marks Highest of all decades (our score) across your whole life, without numbers. We may also note one nearby Growing (our score) decade, a cycle that asks for steadier effort.

Most active in the next 30 years looks only at the next 30 years. Highest of all decades looks at your whole life, so it can point to a past or far-off decade.

We also describe your early, middle, and later years. These stages follow your own luck cycles, not fixed ages: we split them at the cycle start closest to age 30 and the one closest to age 60 (if two are equally close, we use the later one). This is our simplified three-stage split, not a traditional division of life.

8. If you don't know your birth time

You can still get a reading. Without a birth time we can't set your hour pillar, so we use six characters instead of eight and don't show an hour pillar. We calculate the other pillars as if you were born at noon.

If you were born on a day when a solar term begins, your month (or even your year) can depend on the hour. In that case, we show both results: one if you were born before the solar term began, and one if after. Your luck cycles and their starting age are also less precise without a birth time.

9. How we judge chart strength

Many saju readings say how much support your Day Master gets from the rest of the chart. Schools do this differently. Our system gives each chart a support score from 0 to 100:

• We look at the seven other characters. A character supports your Day Master if it is the same element or the element that feeds it in the Five Elements cycle.
• Your month branch counts most. Traditional saju puts the birth month first: the classic text Japyeong Jinjeon (Ziping Zhenquan) treats the birth month (wollyeong) as the key to reading a chart.
• Roots count. Each branch holds hidden stems. If they share your Day Master's element, your Day Master has a "root" there (tonggeun). Branches carry more weight than stems.
• The score places your chart in one of three groups: Strong self-energy, Balanced, or Leans on support. If your birth time is unknown, we score the six characters we have.

The principle, putting the month first and counting roots, comes from traditional saju. The exact weights and cut-off points are our own settings. This is a simple weighted tally: it doesn't adjust for combinations, clashes, or seasonal temperature, which other schools weigh differently, so another site may label the same chart differently. No group is better or worse than another; they're simply different chart patterns.

10. Lunar birthdays

If you enter a lunar-calendar birthday (including a leap month), we first convert it to the solar calendar with an open-source lunar calendar library, then calculate as above. We use the lunar-javascript library by 6tail (MIT License, Copyright (c) 2018 6tail) for this conversion. That library follows the Chinese lunar calendar, and in some years the Korean and Chinese calendars start a lunar month one day apart. For every year from 1900 to 2026 where the two differ, we use Korean month start dates instead. We took these dates from the open-source manseryeok library (MIT License), which states that its data comes from the Korea Astronomy and Space Science Institute. We are not affiliated with or endorsed by that institute.

11. How we check our results

We checked our results against two public calculators on 150 birth records, including births within an hour of a solar term, births around midnight, and births during Korea's UTC+8:30 and daylight-saving years. Year and month pillars matched in all 150. Day and hour pillars matched in 144 records (96%) with one calculator and 117 (78%) with the other. Apart from one case we could not explain, every difference we examined traced to rule choices (how 11 PM is counted, equation-of-time correction, daylight-saving history). None of them looked like an error in our own calculation, but we can't rule one out in that unexplained case. These are agreement rates, not accuracy rates. The records were chosen to include hard cases, not drawn at random. The records are all Korean births (Seoul, Busan, Gwangju, Daegu, Daejeon, Incheon) from 1940 to 2010; we did not test births within one minute of a solar term or births outside Korea. We plan to re-check whenever we update the calculator.

We also ran 500 randomly generated charts (birth years 1940 to 2010, Seoul) through our previous method and the current one to see what changes. About 2 in 3 kept the same strength label; of those that changed, about 90% moved one step and some moved two. In a separate simulation of 3,000 Seoul births, the year pillar never changed and the month pillar changed in about 1 in 1,000 charts. These are simulated charts, not our visitors' charts.

12. Compatibility and Bond Circle

Our compatibility tools compare the elements of two Day Masters (whether one feeds the other, one checks the other, or both share an element) and adjust for how your month branches meet. Bond Circle's six star types (North Star, Daybreak, Binary Star, Satellite, Meteor, and Solstice) are inspired by the Five Elements: our own system, not a traditional saju category.

Changelog

• October 6, 2026: Page published.
• October 6, 2026: More precise charts: birth time converted to UTC for your birthplace and compared with solar terms to the minute; day and hour pillars adjusted for birth-city longitude; births with an adjusted time of 11:00–11:59 PM can choose how that hour is counted; strength gives the birth month the most weight and includes hidden-stem roots; 10-year luck cycles. In a separate simulation of 3,000 Seoul births, the year pillar never changed and the month pillar changed in about 1 in 1,000 charts. Charts for births during past daylight-saving periods or near a solar-term change may also differ.
• October 11, 2026: The luck-cycle field is now Male / Female with two choices; the third option that showed both directions was removed.

Please remember

For entertainment only. See our Disclaimer.