Why Birth Time Accuracy Matters in Saju: What to Do If You Are Not Sure
Separate recorded birth times from estimates, compare changing calendar outputs, and keep uncertainty visible without guessing an Hour Pillar.
Lucky By Birthday · Content revised 2026-09-13
A birth time is an input with a source, not a detail to fill in for a more complete-looking chart. First write down where it came from: a contemporaneous record, a later transcription, a family recollection, or no source. A precise-looking minute does not establish that the underlying record is accurate.
Use a private three-column note: recorded value, source, unresolved questions. For example, a hypothetical note might say ‘around 07:00’, ‘relative recalling breakfast’, and ‘could mean before or after 07:00; exact date and clock convention need confirmation’. Keep that wording rather than turning it into a fictional 07:00 timestamp. Do not upload birth certificates just to use this calculator.
Next distinguish the calendar from the clock. Check whether the recorded date is Gregorian or lunar, and retain any information about location and local time rules. This site takes Gregorian dates and whole entered hours only. It does not apply historical timezone, daylight-saving, longitude, or true-solar-time corrections, so a precise record may still be insufficient for a precise historical chart here.
If the hour is unknown, choose Unknown. We omit the Hour Pillar and use noon provisionally for the remaining calculation. This changes the visible-element denominator from eight to six; the resulting percentages are not directly comparable as a measure of a person's strength. On a solar-term transition date, even Year or Month may vary with the unknown time.
Use the birth-time explorer for a sensitivity check. For the public 2026-02-04 example, this engine produces different Year and Month pairs between the early and later sampled hours. It retains the same Day pair in those samples. That is a statement about sampled calendar outputs, not evidence that a shared personality theme is trustworthy. Twenty-four whole-hour samples also do not locate the exact transition minute.
When records conflict, preserve both versions and their sources. Compare the outputs under each input, then label any differing field unresolved. A reading that feels more accurate does not identify the correct timestamp. If an exact historical conversion is necessary, seek a source or specialist who states the relevant clock and calendar conventions; if it is only for casual exploration, leaving the hour blank is a valid endpoint.
A record worksheet, not an accuracy score
Separate a recorded value from an estimate. A document with a time gives you something to trace, but it does not independently validate the calculator's clock convention. Family memory can be useful context while remaining an estimate. Keep the original wording, who supplied it, and any disagreement; do not turn these categories into numerical confidence or accuracy scores.
If a remembered time is close to a boundary, calculate both neighboring hours and compare only the parts that change. The Year, Month, and Day Pillars normally remain the same while the Hour Pillar may switch. That comparison shows how much of a reading actually depends on the uncertain input. If the narrative changes dramatically even though only one uncertain field changed, treat the narrative with less confidence rather than forcing a single answer.
Location can matter because a clock reading is tied to civil-time rules. Historical timezone changes, daylight-saving rules, and true-solar-time methods are outside this site’s input model. For most casual users, the honest choice is to enter the recorded local clock hour and note the limitation. For research, genealogy, or a professional consultation, retain the original record, birthplace, timezone, and the method used by the calculator.
Four fields to retain privately
- Recorded: copy the value and name of its source without altering it.
- Estimated: retain words such as about, before, or after.
- Conflicting: keep both values and the unresolved difference.
- Unknown: omit the hour; check whether the remaining pillars vary across samples.
Use the unknown-hour/noon comparison to see what is omitted. Noon is provisional: Year and Month Pillars can remain uncertain on a solar-term boundary date. The tool does not correct historical timezones.