# Why do two BaZi calculators disagree? Separate input, convention and output definition

When two BaZi charts differ, stop comparing interpretations. This guide narrows the check to two often-missed points: whether an element diagram counts visible characters or applies weights, and how the late-Zi day-boundary rule alone can change the day pillar. Time zone and solar terms are pointers only.

Updated: 2026-11-10 · Wenbu AI-assisted editorial

When two BaZi charts disagree, do not start with interpretation. Decide which layer the difference sits in. If the input record (Gregorian date, recorded local clock time and its precision) and the historical time zone already match, the remaining disagreement usually comes from calculation conventions or output definitions. On conventions, the year and month pillars follow solar-term boundaries, the day pillar depends on a midnight or 23:00 day boundary, and the hour pillar is also affected by the late-Zi hour-stem convention and the solar-time switch. On output, two "element results" may not be the same quantity at all: Wenbu's element diagram counts each visible stem and branch once, with no hidden-stem weights or seasonal influence, so counts are not element strength. Change one setting at a time and note which pillar moves; if the birth time is unknown, omit the hour pillar, and label any provisional year or month pillar as uncertain. Only after the difference is located does it make sense to discuss interpretation.

[Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/)

- Decide whether the difference is input, convention or output definition before discussing interpretation.

- Different element numbers often mean one result counts visible characters and the other applies weights; they are not the same quantity.

- Near late Zi, the day-boundary rule alone changes the day pillar; change one setting at a time and record which pillar moves.

## From "two different results" to one checkable sentence

1. **Do input and time zone match?** — Same Gregorian date, clock-time precision, birthplace IANA zone and offset for that date? If all four pillars differ, stop at this layer first.

2. **Which pillar differs?** — Year or month pillar: check solar-term boundaries. Only the day pillar, near 23:00 or midnight: check the day-boundary rule. Hour pillar: check solar time and the hour boundary.

3. **Is the output the same quantity?** — Are both element results visible-character counts, or does one include hidden stems or seasonal weights? Different definitions cannot be compared numerically.

4. **Write one checkable sentence** — For example, "date and time zone agree; the day pillar differs because of the day-boundary rule," and note which information remains unknown.

Decide in order: are input and time zone aligned, which pillar differs, and do both outputs share a definition? If any step contains unknown information, mark it unknown and stop inferring downward rather than filling the gap with a guess.

[Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/)

## First confirm both results measure the same thing

Many "different chart" arguments actually happen at the output layer. Both results say "five elements," but one counts each visible stem and branch once, while the other may fold hidden stems or seasonal factors into weights. The first answers "how often does each element appear among the eight characters"; the second answers "what score does some weighted model give." Different numbers do not show that either is wrong. Wenbu's element diagram is the first kind: it counts each visible stem and branch once, adds no hidden-stem weights or seasonal influence, and therefore cannot establish element strength, structure or a favourable element from counts alone.

When checking, read the column headings and layers first. Some charts place the year pillar on the right, so the Day Master is not always in the same visual position; visible characters, hidden stems and Ten Gods are separate layers and should not be mixed when counting. Ask what each output defines before asking why the numbers differ. A calculation can be reproduced under its conventions without its interpretation being validated, and that holds for any two results you compare.

[Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/)

## Compare two counting methods on one fixed example

Take the teaching example in Wenbu's documentation: 23 December 2005 at 08:37, Asia/Shanghai, civil clock, midnight day boundary, solar time off. The four pillars are Yi You, Wu Zi, Xin Si and Ren Chen, with Xin Metal as the Day Master. Classify each visible character: Yi is Wood, You is Metal, Wu is Earth, Zi is Water, Xin is Metal, Si is Fire, Ren is Water, Chen is Earth. That gives Wood 1, Fire 1, Earth 2, Metal 2, Water 2, and the total is exactly eight.

If another result reports different element numbers, do not rush to call it wrong. Ask whether it also counts hidden stems or weights elements by season. A visible-character count and a weighted model answer different questions: the first can be checked character by character, while the second can only be checked if its weighting rules are published. The value of this example is that it gives a baseline you can verify item by item; any deviation should be explainable as a different input, a different convention or a different output definition, not a vague "the algorithm differs."

[Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/)

## Near late Zi, the day-boundary rule alone changes the day pillar

With input and time zone already matching, the easiest convention to overlook is the day boundary. Wenbu defaults to a midnight boundary and also offers a 23:00 Zi boundary; under the default rule, the late-Zi hour stem still follows the library's next-day convention. So a time after 23:00 but before midnight can yield different day pillars under the two rules, while the year and month pillars may stay identical. Here the difference has a single source, and there is no need to recheck time zone, solar terms and solar time all at once.

The test is direct: first see whether the two results agree on the year and month pillars. If they agree but the day pillar differs, and the birth time is near 23:00 or midnight, compare the day-boundary rules first. If the hour pillar also differs, then check whether solar time is enabled and whether the corrected time crosses an hour boundary. When the record says only something like "a little after eleven at night," the boundary itself cannot be settled, so write that down as unknown rather than picking whichever result looks smoother.

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md)

## Time zone and solar terms: confirm, then look further

Time zone and solar-term boundaries are two further layers; this guide does not repeat them in full, only the order of checking. A clock time means something only with the time zone in effect on that date, and today's offset is not necessarily the offset at birth. Mainland China observed daylight saving in some years, and overseas birthplaces can have clock times that were skipped or repeated at a transition. Wenbu rejects such ambiguous inputs instead of silently choosing an instant. If all four pillars differ, return to this layer before suspecting the day-boundary rule.

The year and month pillars follow absolute solar-term boundaries rather than Lunar New Year; Wenbu compares solar terms using a fixed UTC+08:00 standard time without historical daylight saving, while the day and hour pillars use the selected local clock. Enabling solar time corrects only the day and hour pillars, leaving the year and month pillars on the original instant. The full treatment is in Wenbu's birth-time and solar-time guide; here these are prerequisites on the checklist, because until this layer is aligned the later comparisons mean little.

[Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md)

## Locating the difference: from output definition to day-boundary rule

| What you observe | Most likely layer | What to check |
| --- | --- | --- |
| All four pillars match but element numbers differ | Output definition | Confirm one is a visible-character count and whether the other includes hidden stems or seasonal weights |
| Only the day pillar differs, time near 23:00 or midnight | Day-boundary rule | Compare the midnight and 23:00 settings and record the day-pillar change |
| Only the hour pillar differs | Solar time or hour boundary | Confirm birth-time precision and whether the corrected time crosses an hour boundary |
| All four pillars differ | Input record or historical time zone | Go back to the Gregorian date, clock time and birthplace zone item by item |

[Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md)

## Hypothetical case: same four pillars, two counting methods

The following is a teaching scenario, not any real person's birth data. Suppose two readers use the same inputs — 23 December 2005 at 08:37, Asia/Shanghai, midnight day boundary, solar time off — and both get the pillars Yi You, Wu Zi, Xin Si and Ren Chen, but their element numbers differ.

1. First, confirm the inputs match: date, clock time, time zone and day-boundary setting are identical, and so are the four pillars, so the difference is neither in the input layer nor in the convention layer.

2. Second, count visible characters one by one: Wood 1, Fire 1, Earth 2, Metal 2, Water 2, total eight, matching the eight characters and verifiable item by item.

3. Third, ask what the other result defines: if it counts hidden stems or weights elements by season, it answers a different question, and different numbers are expected rather than wrong.

4. Fourth, write the conclusion: "Both results share the same four pillars; the element numbers differ because one counts visible-character occurrences and the other uses a weighted model." To compare strength, first align on one definition.

The lesson is a concrete action: when you receive two element results, count the visible characters yourself, check whether the total is eight, then ask whether the other result introduces hidden stems or weights. The version you can verify character by character is the baseline; the one you cannot verify needs its rules published. Do not treat the hypothetical case as a real chart, and do not assume the more "balanced" numbers are more accurate.

[Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/)

## FAQ

### The element numbers differ. Does that mean one chart is wrong?

Not necessarily. First check what each output defines. Wenbu's element diagram counts only visible stems and branches, without hidden-stem weights or seasonal influence; if another result includes hidden stems or weighting, the two are not the same quantity. Align the output definition before judging the numbers.

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md)

### Only the day pillar differs. What is the most likely cause?

If input and time zone match and the birth time is near 23:00 or midnight, the most likely cause is a different day-boundary rule: one uses midnight, the other a 23:00 Zi boundary. First compare whether the day pillar changes under those two settings, then consider other layers.

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md)

### The birth time is remembered only as "a little after eleven at night." What should I do?

Write it down as unknown rather than picking one of the two results. Near the day boundary, different rules give different day pillars, and the record itself cannot distinguish them; if even the hour is uncertain, the safer move is to omit the hour pillar and note the uncertainty of any provisional parts.

[Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md)

### Should I recheck time zone and solar terms as well?

Only when the four pillars differ widely or the date and time zone are themselves in doubt. If the year and month pillars agree, the solar-term boundary and absolute instant are probably handled the same way, so focus on the day pillar, hour pillar and output definition. The full time-zone and solar-time procedure is in Wenbu's related guide.

[Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/)

## Glossary

- **Visible-character count**: A method that counts each stem and branch among the eight characters once, always totalling eight; it shows occurrences, not element strength. [Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md)

- **Day boundary**: The rule deciding when the day pillar advances, commonly midnight or 23:00 Zi; different rules change the day pillar near the boundary. [Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md)

- **Late Zi**: The interval between 23:00 and midnight; under Wenbu's default rule the day pillar stays on the civil day while the hour stem follows the library's next-day convention. [Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md)

- **Apparent solar time**: An approximate correction to local clock time using longitude and the equation of time, applied only to the day and hour pillars; it is a calculation convention and does not by itself show which interpretation is more valid. [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/)

## Evidence notes

- A BaZi chart should be checked against the birth record and calculation settings before interpretation; eight characters alone are not a reproducible record. — “A screenshot of eight characters alone is not a reproducible record.” (editorial); [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md)

- Wenbu's element diagram counts each visible stem and branch once, without hidden-stem weights or seasonal influence, so counts alone cannot establish strength. — “五行图把每个可见天干和地支各计一次，表示出现次数。” (fact); [Wenbu · calculation conventions](https://wenbu.app/methodology/)

- A clock time identifies an instant only with its date and historical time zone; today's offset may differ from the offset at birth. — “Today’s UTC offset may differ from the historical offset at birth.” (fact); [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md)

## Sources

- [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) — Existing editorial and reproducible chart conventions; fetched 2026-11-10T03:00:34.325Z; complete response

- [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) — Existing editorial and boundary examples; fetched 2026-11-10T03:00:34.347Z; complete response

- [lunar-typescript · official source](https://raw.githubusercontent.com/6tail/lunar-typescript/master/README.md) — Calculation implementation and conventions; fetched 2026-11-10T03:00:34.332Z; complete response

- [Wenbu · calculation conventions](https://wenbu.app/methodology/) — First-party implementation; not evidence of predictive accuracy; fetched 2026-11-10T03:00:34.163Z; complete response

[Generate one chart from the same inputs and count the visible characters first](https://wenbu.app/en/bazi/)

# Why the same instant can give different BaZi day and hour pillars

Check the UTC timestamp for three fixed inputs, then see which pillars depend on the selected local clock.

Updated: 2026-11-12 · Wenbu AI-assisted editorial

Two timestamps can name the same instant and still produce different BaZi day and hour pillars. Wenbu calculates year and month against absolute solar-term boundaries. By default it uses the selected local clock for day and hour. To repeat a chart made for the birthplace, keep its local date, time, time zone and calculation settings. Save UTC separately as a check.

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Current complete en canonical article](https://wenbu.app/knowledge/en/why-bazi-charts-differ.md)

- June 18 at 00:30 in Shanghai is June 17 at 16:30 UTC.

- Changing the zone and keeping the digits instead moves the instant, even if the chart happens to match.

- Keep the birthplace clock record to repeat the chart. Compare UTC to check whether the instants match.

## Two comparisons that answer different questions

1. **Copy the input** — Keep the local date, time and zone together. Record the day boundary and solar-time setting.

2. **Compare UTC** — Convert the date as well as the time. Check for a change of day.

3. **Compare both pairs of pillars** — Check year and month, then day and hour. Note which input changed.

Check the timestamp first, then compare the two pairs of pillars.

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Current complete en canonical article](https://wenbu.app/knowledge/en/why-bazi-charts-differ.md)

## Check the instant before comparing the clock inputs

Is the UTC timestamp a conversion of the birthplace clock? Here we subtract eight hours from Shanghai time. The date moves back a day. It is the same instant, expressed on a different clock.

Copy each date, time and zone. Convert to UTC, or compare calendar.instant in the calculation JSON. Matching pillars cannot prove that two records refer to the same instant.

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Current complete en canonical article](https://wenbu.app/knowledge/en/why-bazi-charts-differ.md)

## Change one part of the record at a time

Keep the midnight boundary and turn solar time off. A uses the Shanghai clock; B uses its UTC equivalent. Both give Xin Si and Jia Wu for year and month. Day and hour differ.

C keeps A’s date and digits but changes the zone. It is eight hours later, even though all four pillars happen to match. To repeat A, keep its local date, time and zone.

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md)

## Three inputs with midnight boundary and solar time off

| Date, time and zone entered | UTC timestamp | Year and month; day and hour |
| --- | --- | --- |
| A: 2001-06-18 00:30, Asia/Shanghai | 2001-06-17 16:30Z | 辛巳, 甲午; 壬子, 庚子 |
| B: 2001-06-17 16:30, UTC | 2001-06-17 16:30Z | 辛巳, 甲午; 辛亥, 丙申 |
| C: 2001-06-18 00:30, UTC | 2001-06-18 00:30Z | 辛巳, 甲午; 壬子, 庚子 |

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md)

## Try it: which records name the same instant?

Cover the last two columns and compare the fictional inputs.

1. Subtract eight hours from A and go back to June 17. Its UTC timestamp matches B.

2. Compare A and B. UTC matches, but the day and hour pillars differ. Check which local clock the original chart uses.

3. Compare A and C. The entered digits and the pillars match, but their instants are eight hours apart.

A and B describe the same instant. Use A to repeat the Shanghai chart.

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Current complete en canonical article](https://wenbu.app/knowledge/en/why-bazi-charts-differ.md)

## FAQ

### What if all I have is a UTC timestamp?

If you know the birthplace zone, convert UTC to the local date and time in that zone on that date. If the zone is unknown, record the uncertainty rather than substituting the zone where you live now.

[Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Current complete en canonical article](https://wenbu.app/knowledge/en/why-bazi-charts-differ.md)

### Why can different instants share year and month pillars?

Those pillars follow solar-term boundaries. Two instants can fall on the same side of the relevant boundaries. Compare timestamps to check the instant; pillar names cannot settle it.

[Wenbu · calculation conventions](https://wenbu.app/methodology/) · [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Current complete en canonical article](https://wenbu.app/knowledge/en/why-bazi-charts-differ.md)

## Glossary

- **Absolute instant**: A moment that does not change when expressed in another zone. A and B have the same UTC timestamp. [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Wenbu · calculation conventions](https://wenbu.app/methodology/)

- **Local civil clock**: The clock date and time in a particular zone. Keep all three fields when repeating a calculation based on that clock. [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) · [Current complete en canonical article](https://wenbu.app/knowledge/en/why-bazi-charts-differ.md)

## Evidence notes

- Wenbu calculates year and month by comparing the birth instant with solar-term instants. — “年、月柱按出生绝对时刻与节气时刻比较” (fact); [Wenbu · calculation conventions](https://wenbu.app/methodology/)

- Without solar-time adjustment, day and hour use the selected local clock. — “Day and hour pillars use the selected local clock.” (fact); [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md)

- Clock digits alone do not identify an instant. The zone in effect on that date matters. — “A clock time means something only with the time zone in effect on that date” (fact); [Current complete en canonical article](https://wenbu.app/knowledge/en/why-bazi-charts-differ.md)

## Sources

- [Current complete zh canonical article](https://wenbu.app/knowledge/zh/why-bazi-charts-differ.md) — Current public original and earlier additions; add substantial coverage without repeating or replacing them.; fetched 2026-11-12T01:08:00.619Z; complete response

- [Current complete en canonical article](https://wenbu.app/knowledge/en/why-bazi-charts-differ.md) — Current public original and earlier additions; add substantial coverage without repeating or replacing them.; fetched 2026-11-12T01:08:00.667Z; complete response

- [Wenbu · birth time and time zones](https://wenbu.app/knowledge/en/birth-time-timezone.md) — Existing editorial and boundary examples; fetched 2026-11-12T01:08:00.759Z; complete response

- [Wenbu · calculation conventions](https://wenbu.app/methodology/) — First-party implementation; not evidence of predictive accuracy; fetched 2026-11-12T01:08:00.725Z; complete response

- [Wenbu · BaZi guide](https://wenbu.app/knowledge/en/bazi-basics.md) — Existing editorial and reproducible chart conventions; fetched 2026-11-12T01:08:00.773Z; complete response

[Compare A and B in the BaZi calculator](https://wenbu.app/en/bazi/)