Experimental working-time estimate with explicit uncertainty
Rectification is an estimate, not a measurement. Astrobot compares your birth data, questionnaire answers and dated events against several heuristic methods, then shows a working time, an interval, data quality and warnings. This is not a certified birth minute: on a blind 24-hour window the median error is ~283 min and the Ascendant sign is right ~12% of the time; with a ±1-hour prior it is ~28 min and ~72%. Add events and narrow the window to make the estimate more stable.
1
Birth data
Date, city and a time window. The narrower the window, the more stable the estimate — a ±1-hour prior is best. A full day gives a rough interval, not a precise minute.
2
Questionnaire: 6 questions about you
Appearance, character, relationship with your mother — this is a weak prior about the Ascendant's element. Skipping is allowed; the questionnaire alone does not prove the time.
3
Life events — the main signal
Dated events: marriages, moves, career shifts. At least 3 are needed, preferably 8–15; a single marriage is a weak anchor. Without events or answers the result is marked as insufficient data.
4
Working estimate and interval
You see the working time, interval, methods that actually contributed, data quality and warnings. Five control questions run a consistency check on the same engine — not independent confirmation.
More in the Knowledge Library — breakdowns of signs, planets and houses
Free, no sign-up required·This tool is available to all visitors. Sign-in is required only if you want to send the finished result to AI Chat for a detailed interpretation and saved history.
~28 min
Median error
only with a ±1 h prior window
72%
Ascendant sign
with ±1 h prior; 12% blind 24 h
≤6
Methods listed
only methods actually used
50
Pilot charts
not an independent validation study
1
2
3
4
Birth Data
The exact birth date is critical — even a 1-day shift changes the entire calculation.
The narrower the window, the more reliable the estimate. A ±1-hour prior is best. Pilot benchmark (50 known charts): ±1 h prior → median ~28 min, Asc sign 72%; blind 24 h → ~283 min, Asc sign 12%. This is not a general accuracy guarantee.
Gender selects the marriage ruler: men — Venus/Moon, women — Sun/Mars.
City coordinates determine Ascendant and MC. Time zone is filled in automatically.
How this estimate is produced
What the numbers do and do not mean
Marriage ruler heuristic
Uses the directed MC at marriage events as one weak, disputed signal. It works only when marriage dates are provided and is never treated as proof.
Symbolic directions
Uses a mean solar arc of 0.9856°/year. This is a simplified symbolic arc, not true primary directions in right ascension.
Conception Moon prior
A fixed ~273-day prenatal Moon check against the Asc/Desc axis. The method is disputed and is used only as a low-weight prior, never as the sole signal.
Annual profections
Whole-sign annual profections. In this engine the score is close to non-discriminative, so it is not presented as independent evidence.
Heuristic scoring
Signals are weighted and combined. Agreement between methods is not proof and is not a calibrated probability.
Interval and warnings
You see the working interval, the methods that actually contributed, the data quality and all warnings — not just a single minute.
Experimental estimate
This is an open heuristic, not a validated method. Pilot benchmark: blind 24 h → median ~283 min, Asc sign 12%; ±1 h prior → ~28 min, Asc sign 72%.
Rectification is an inference, not a measurement. Treat the output as a working chart and interval. The pilot benchmark is not a validation study, and no exact birth minute is claimed.
AI Rectification (working estimate) | Astrobot · Astrobot
