Our methodology
Everything we claim is either a classical principle, a published constant, or a verified astronomical calculation. This page sets out all three, along with what we do not claim.
1. The astronomy: how the four pillars are computed
A BaZi chart has four pillars, and they do not all depend on the same thing. Two of them are solar-term dependent — a global instant in time. Two are local-day dependent — a function of where you were standing. We compute them separately.
| Pillar | Depends on | How we compute it |
|---|---|---|
| Year | The instant of 立春 | Birth instant converted to UTC+8, read against the exact 立春 boundary |
| Month | The instant of the governing jie | Birth instant converted to UTC+8, read against the exact solar term |
| Day | The local solar date | Apparent solar time at the birthplace, converted from Julian day number |
| Hour | The local solar two-hour block | Day stem of the local solar date, via the fixed 五鼠遁 derivation |
True solar time. Local clock time is adjusted by two corrections: the difference between the birthplace longitude and the timezone's standard meridian, at four minutes per degree; and the equation of time, which accounts for the ellipticity and tilt of Earth's orbit. Together they can move a birth by well over an hour. Full explanation.
Timezone handling. We store IANA timezone identifiers and resolve the correct UTC offset for your specific birth date, so daylight saving and historical rule changes are handled correctly rather than assumed away.
2. The boundaries we handle that others do not
Born outside Asia
Most BaZi calculators were written for a Chinese audience and read every birth as though it happened in Beijing. Someone born in New York on 4 July at 23:50 is converted to 5 July 11:50 Beijing time, and the calculator reports the wrong day pillar. We split the calculation so this cannot happen. Worked example.
Born between 23:00 and midnight
This is a genuine unresolved dispute in the tradition, not a technical question. We default to the reading that keeps the chart internally consistent with its own hour-stem derivation, and we give you a switch to follow the other school. Both sides explained.
Born in western China on Beijing time
China uses a single timezone across five geographic hours. Someone born in Lhasa at 13:30 by the clock was born at roughly 11:34 by the sun. Without the solar correction the hour pillar is simply wrong.
Born in mainland China between 1986 and 1991
China observed daylight saving time for six years and then abolished it. Births in that window were recorded on a clock one hour ahead of standard time for roughly five months of each year. Someone born in Beijing on 15 May 1990 at 12:00 by the clock was born at 10:49 by the sun — a different two-hour period and therefore a different hour pillar. We detect this automatically from the IANA timezone database and state it in the result. Most calculators do not, including several commercial Chinese ones.
Born in the first days of February
The Chinese year changes at 立春, not at Chinese New Year and not on 1 January. We use the exact instant of 立春, so births in the gap are handled correctly.
3. The elemental weighting model, in full
Every chart is scored two ways, and we show you both.
The surface count
The intuitive view: how many of your eight characters belong to each element, reading each Earthly Branch as its principal element. This is what most sites show you.
The weighted strength model
The view that actually drives the reading. Each component contributes as follows:
| Component | Weight | Reasoning |
|---|---|---|
| Each Heavenly Stem | 1.0 | A visible, unambiguous influence |
| Each Earthly Branch | 1.2, split across hidden stems | A branch contains one to three hidden stems; the principal one carries the most |
| Hidden stem split — one stem | 1.00 | e.g. 子 contains only 癸 |
| Hidden stem split — two stems | 0.70 / 0.30 | Principal then residual |
| Hidden stem split — three stems | 0.60 / 0.30 / 0.10 | Principal, middle, residual |
| The month branch | × 2.0 | The birth month carries the most authority in classical practice — 月令当权 |
| Resource element (印) | × 0.85 when assessing support | The generating element supports the Day Master, but less directly than its own element |
Hidden stems are taken from the standard 藏干 table: 子 癸; 丑 己癸辛; 寅 甲丙戊; 卯 乙; 辰 戊乙癸; 巳 丙庚戊; 午 丁己; 未 己丁乙; 申 庚壬戊; 酉 辛; 戌 戊辛丁; 亥 壬甲.
4. How we decide whether a Day Master is strong or weak
Support for the Day Master comes from its own element (比劫) and the element that generates it (印). Everything else draws on it: what it produces (食伤), what it controls (财), and what controls it (官杀).
support share = (own element + 0.85 × generating element) ÷ (support + output + wealth + officer)
A perfectly even distribution across all five elements produces a support share of about 0.381, because there are two supporting roles and three draining ones. We therefore express the result as a balance index: the support share divided by that even baseline, so 1.00 means perfectly even.
| Balance index | Reading | What it means practically |
|---|---|---|
| 1.45 and above | Strong (身强) | Well supported. The chart needs somewhere for its energy to go. |
| 1.2 and above | Slightly strong (偏强) | Comfortably supported. Output and structure are useful. |
| 0.85 and above | Balanced (中和) | Close to even. The elements in shortest supply are the useful ones. |
| 0.65 and above | Slightly weak (偏弱) | Slightly outnumbered. Nourishment and reinforcement are useful. |
| below 0.65 | Weak (身弱) | Drained by its surroundings. Support is the priority. |
The seasonal override
Strength arithmetic alone is not the whole story, and classical practice says so explicitly. A chart born in the cold months may be structurally adequate yet still need warmth; a chart born in high summer may need moisture even if the numbers do not demand it. We apply a 调候 layer on top of the strength result:
- Born in 亥 / 子 / 丑 with Fire below average → Fire is added to the useful elements
- Born in 巳 / 午 / 未 with Water below average → Water is added
- Born in a damp month with Water already high → Fire is added, for drying
When this layer applies, it is stated explicitly in your result rather than folded silently into the numbers.
5. How the chart becomes a body constitution
Each element that is above or below its even share contributes to specific constitution scores. The logic is stated in full so that a reader — or a practitioner — can argue with it:
| Constitution | Primary signals in the chart |
|---|---|
| Balanced | Element distribution is comparatively even — no single element dominates; Day Master strength sits in the middle band; No strong clash between the birth-month season and the day element |
| Qi Deficiency | Earth and Metal are the weakest elements — Spleen and Lung energy; Day Master is weak in a chart with more draining than supporting elements; Very little Resource (generating) element feeding the Day Master; Weakening season: born in a month that directly drains the Day Master |
| Yang Deficiency | Fire is the weakest element in the chart, with Water comparatively strong; Born in the Winter months (the branch is a strong undermining condition); Damp-cold branches dominate, adding cold and moisture together; Day Master weak and unsupported by a warming month |
| Yin Deficiency | Water is the weakest element, with Fire comparatively strong; Born in the Summer months — a hot month spending the coolant; An excess of Fire-type characters with little to moderate it; Day Master strong (and thus hot) with no draining element present |
| Phlegm-Damp | Earth is strong, often damp Earth branches dominate the branches; Water is also present, so the chart carries both moisture and volume; Day Master strong with heavy Earth, and little Wood to keep it moving; Born in the late-summer or damp months |
| Damp-Heat | Fire and Earth are both strong — heat with a medium to carry it; Born in Summer, particularly in a damp month; Damp Earth and Fire branches combining in the chart; Day Master strong and hot with no cooling Water present |
| Blood Stasis | Wood is strong and Metal is weak — the pushing force outruns the containing and moving force; Day Master is strong with excess Wood and no Metal to regulate it; A clash pattern in the branches, particularly involving the Day branch; Cold-dominant charts can also stagnate blood |
| Qi Stagnation | Wood is strong and Earth is weak — Wood over-controlling Earth; Several Wood characters with little Fire to express them outward; Day Master weak and outnumbered, producing a sense of being pushed around; Born in the Spring with excess Wood and no Metal to shape it |
| Special Constitution | Metal is weak — the defensive perimeter (Wei Qi) is under-resourced; Earth is weak as well, so the production of defensive energy is limited; Metal weak with strong Wood — the perimeter is breached by an active pushing force; An unbalanced chart in the transition months, spring and autumn |
We also detect the six Earthly Branch clashes (六冲) between pillars, which classical practice reads as motion and instability — a genuine signal for stagnation patterns.
The result is a ranked profile across all nine constitutions rather than a single label, because a single label implies a precision the underlying model does not have.
6. What we do not claim
This list is deliberate, and it is longer than most. We would rather be trusted for what we are than overpromise.
- We do not diagnose or treat anything. No chart or quiz result is a medical finding. If you have symptoms, see a clinician.
- We do not predict events. A chart describes patterns and timing tendencies, not what will happen on a date.
- We do not claim scientific validation of the five-element framework. It is a traditional system with a long observational history. Some of its practical advice converges with modern findings; much of it is unstudied. We distinguish the two where we can.
- We do not recommend herbs or supplements. Food and lifestyle only. Anything stronger belongs with a qualified practitioner who can see your whole health picture.
- We do not call our quantification classical. It is a modern formalisation of classical principles, and we say so on the tin.
- We do not claim certainty at the boundaries. Solar-term edges, midnight births and the late Zi hour dispute are genuinely ambiguous. We show the ambiguity.
7. Our verification record
Accuracy claims should be checkable. These are ours, as of the current release:
| Check | Method | Result |
|---|---|---|
| Day pillar | Independently recomputed from Julian day numbers for 300 randomly generated dates, compared against the reference implementation | 300 / 300 identical |
| Solar term instants | Compared against published astronomical almanac values for 2024–2026 across six terms | Agreement to the minute |
| Chinese New Year dates | Checked for 2020–2026 against the published lunar calendar | All correct, both directions |
| 立春 year boundary | Tested immediately either side of the 2026 立春 instant | Year pillar changes correctly |
| Full four-pillar agreement | 400 random birth moments compared against the reference implementation under identical settings | 400 / 400 identical |
| Timezone and daylight saving | Resolved offsets tested against known historical and modern values across 13 zones | All correct, including half-hour and quarter-hour offsets |
| Historical daylight saving | Standard-vs-daylight detection tested across eight zone and year combinations, including mainland China's 1986–1991 daylight saving period | All correct |
To be clear about what this proves: it demonstrates that our arithmetic matches the reference astronomical and calendrical data. It does not and cannot validate the interpretive framework itself. Those are different claims, and conflating them is the most common dishonesty in this field.
Where the place data comes from
The birthplace index behind the calculator's city field is built from GeoNames — every settlement above 50,000 people worldwide, plus every capital, each with its coordinates and IANA timezone — used under the Creative Commons Attribution 4.0 licence. We generate the index at build time, so what your browser searches is our own file, not a query sent to a third party. Your birth city is never transmitted anywhere.
Questions about our method
Is the elemental weighting model an authentic classical method?
No, and we will not pretend otherwise. Classical analysis is qualitative and depends on a practitioner's judgement. What we publish is a transparent modern quantification, built on classical principles — the month is given more weight, hidden stems are counted, and the generating and controlling cycles are respected. We publish the constants so you can judge the model rather than trust it.
Why do you publish your weights at all?
Because most calculators do not, and that makes them unfalsifiable. If we tell you the month branch is weighted twice, you can disagree with us specifically. That is a better relationship than asking you to trust a black box.
How accurate is the astronomical calculation?
We build on a library that computes solar terms to the minute against modern ephemerides. We verified it against the published astronomical almanac for multiple years, and we cross-checked the day pillar against an independent formula based on Julian day numbers across three hundred random dates with zero mismatches.
Can the same birth data produce two different readings?
Yes, at the margins — around the solar-term boundaries, around midnight, and around the late Zi hour dispute. That is not a bug in our tool; it is genuine ambiguity in the tradition. We surface it rather than hide it.