Keep a shared input record

Constructed example: Gregorian 12 May 1990, local 09:30, Guangzhou, UTC+08:00. This demonstrates recordkeeping, not the resulting chart or a prediction. Substitute evidence-based information and record which input path you used. Enter the same information in the other tool.

Check the calendar first

Establish whether the original record is Gregorian or lunar and whether a leap month is relevant. Review the final converted date in both tools. Do not enter a date already converted to Gregorian into a lunar input again. Resolve different displayed dates before comparing the Four Pillars.

Review time and place together

A local clock reading and a converted time can have the same digits while referring to different instants. Record birthplace and time zone or UTC offset. Historical timekeeping needs separate verification. Check each tool’s actual documentation for time adjustments, day boundaries and seasonal conventions rather than assuming a setting exists.

Separate calculation from interpretation

Compare the final date and time, Four Pillars, luck-cycle periods and person selected. If those match but the text differs, review the sources, school and period discussed. Guan Shi Zhi Li offers two Bazi conversation preferences, while Bazi analysis reports use Ziping. Different scopes are not the same comparison.

Make an unresolved difference reproducible

Record tool names and versions, shared inputs, displayed dates and the exact difference. Redact names and unnecessary details from screenshots before contacting support. Keep approximate birth data marked as uncertain. Agreement between calculators tests consistency, not predictive accuracy.