Two BaZi Calculators Disagree? Check the Inputs and Rules.

From the YUVYN team. Written with AI assistance.

You use one birth record in two calculators and get different BaZi charts. Before choosing the reading that feels more convincing, put the entered dates, places, times and disclosed rules side by side. The two worksheets below help you locate differences and record what remains unknown.

BaZi belongs to traditional Chinese astrology and is used culturally and interpretively. This worksheet documents calculation assumptions and chart components. It does not establish scientific predictions or recover an actual birth time.

New to chart labels? Start Here: read your BaZi chart before the story. With a known birth time, open the free English BaZi calculator and use the worksheet below to record its inputs and chart. A personal AI interpretation is separate, optional and paid.

Build an input ledger

Label the two results A and B. Copy the entries as entered, including the written date and clock value. Record the calculator name, any displayed version, and when you checked it.

Confirm the calendar before entering the date. If a form requests a Gregorian date, use the corresponding Gregorian date rather than copying lunar month/day numbers into that field. For example, the Hong Kong Observatory's 2025 Gregorian-Lunar conversion table places the first day of the first lunar month on 29 January 2025. Keep the original record and note any conversion. Date conversion is separate from a calculator's rules for changing year and month pillars.

Copy each calculator's stated Year-, Month- and Day-pillar boundary rules. If a rule is not disclosed, record "not disclosed"; do not assume it matches the calendar used for date entry.

Field	Chart A	Chart B
Date as entered	[date]	[date]
Calendar mode as stated	[shown calendar]	[shown calendar]
Year-pillar boundary rule as stated	[stated convention]	[stated convention]
Month-pillar boundary rule as stated	[stated convention]	[stated convention]
Day-pillar boundary rule as stated	[stated convention]	[stated convention]
Clock value as entered	[time]	[time]
Birthplace as entered; coordinates if shown	[place; shown coordinates]	[place; shown coordinates]
Timezone and displayed offset	[shown value]	[shown value]
Civil-time or solar-time convention	[stated convention]	[stated convention]
Solar-time correction setting	[shown setting]	[shown setting]
Recorded-time uncertainty interval	[exact, range, or unknown]	[exact, range, or unknown]
Record source	[source label]	[source label]
Calculator, version, check date	[shown information]	[shown information]

If the birthplace name could refer to more than one place, check the town and region against the birth record before comparing results; see Same Town Name, Different Birthplace: Check Before You Calculate.

When a calculator does not disclose a setting, write "not disclosed" in that cell. Do not silently fill it with the setting used by another tool. A blank leaves the comparison ambiguous; an explicit missing value shows what still needs clarification.

Keep the original record separate from the value entered into the calculator. If the record is approximate but the form requires a precise time, label that entry as an assumption. A precise-looking input does not make the underlying record precise.

Capture outputs before interpretations

Copy the raw stem-and-branch pair for each labelled pillar. Chart layouts can differ, so locate the Year, Month, Day and Hour labels rather than assuming a column position. Record the Day Master separately: it is the Heavenly Stem of the Day Pillar.

Use this output ledger beneath the inputs:

Component	A	B	Observation
Year Pillar	[pair]	[pair]	[same, different, or not readable]
Month Pillar	[pair]	[pair]	[same, different, or not readable]
Day Pillar	[pair]	[pair]	[same, different, or not readable]
Hour Pillar	[pair]	[pair]	[same, different, or not readable]
Day Master	[Day stem]	[Day stem]	[same, different, or not readable]

This preserves unchanged components as well as differences. A changed Hour Pillar does not, by itself, imply a changed Day Master. If a value is unreadable, keep that limitation visible rather than guessing.

Change one item at a time

First identify entries or disclosed conventions that differ. For a controlled comparison, keep the other recorded settings fixed and change one item. Save a new labelled row with the exact change and resulting outputs. Changing several assumptions together makes it harder to tell which change affected a component.

If a setting remains undisclosed, the comparison remains incomplete. The ledger alone does not decide which calculator is correct.

Keep candidates when the record is uncertain

If the recorded-time interval crosses a relevant boundary under the chosen convention, retain both labelled candidates and their assumed times. Do not round the record into certainty or choose a time because its personality narrative sounds more flattering. The comparison shows dependence on an assumption; it cannot establish the actual time.

Our existing fictional example makes this visible: on 15 June 2000, using recorded local civil time in Asia/Shanghai, UTC+08:00, without true solar-time correction, 10:59 gives Ji Si (己巳) and 11:00 gives Geng Wu (庚午). The other three pillars and the Jia (甲) Day Master stay unchanged. These local Tyme 1.5.0 results illustrate that fixed convention, not a universal 11:00 civil-clock boundary.

Read the free worked example, then reuse the ledgers above for your own comparison:
https://yuvyn-bazi.blogspot.com/2026/10/1059-vs-1100-why-one-minute-changed.html

With a known birth time, use YUVYN's free English BaZi calculator to get your Four Pillars and Day Master. Record its inputs and disclosed conventions in the ledgers above. It does not infer unknown birth time. A personal English AI reading is a separate, optional paid Natal product.

Choose the output you actually need: compare YUVYN's free BaZi chart, free Natal preview and optional paid AI reading.

Explore YUVYN's free fictional English reading sample to see how a reading is organized. No birth details are needed.