Janmagyan study
How much does an error in birth time change a Kundli?
The finding
A 60-minute error in birth time changes the Lagna in 50.2% of charts, and a 15-minute error in 12.7%. The Moon is far steadier: its sign changes in 1.9% of charts at one hour. The Navamsa Lagna changes in 37.5% at just five minutes, and dasha dates move about 3.9 days per minute of error.
- 50.2%of charts change Lagna with a one-hour error
- 12.7%change Lagna with a 15-minute error
- 1.9%change Moon sign with a one-hour error
How we measured it
- Births. 20,000 random births, with every calendar day from 1960 to 2020 equally likely and every minute of the day equally likely.
- Places. Each birth was placed in one of the 7,246 Indian towns and cities in our place list, chosen in proportion to population, from the far south to the Himalayan foothills.
- What was varied. Each chart was computed at its time, then again with the clock time moved earlier and later by 2, 5, 10, 15, 30, 60 and 120 minutes (the date rolls over at midnight): 15 charts per birth, 300,000 in all.
- Settings. Janmagyan's chart engine with the Lahiri (Chitrapaksha) ayanamsa, whole-sign houses and the mean lunar node, the same settings as every chart on this site.
- What is reported. For each error size, the share of births in which the item differs, averaging the earlier and the later shift, with a 95% interval (Wilson) widened to allow for both shifts coming from the same birth.
- Repeatable. The random draws come from a fixed seed (20261117), so the run gives exactly these figures every time.
Results
Share of charts that change, in per cent, by size of the birth-time error. The 95% interval is under each figure.
| What changes | 2 min | 5 min | 10 min | 15 min | 30 min | 60 min | 120 min |
|---|---|---|---|---|---|---|---|
| Lagna sign | 1.71.6 to 1.8 | 4.24.0 to 4.4 | 8.58.2 to 8.8 | 12.712.4 to 13.0 | 24.924.5 to 25.3 | 50.249.7 to 50.7 | 94.394.1 to 94.6 |
| Lagna nakshatra | 3.83.6 to 4.0 | 9.29.0 to 9.5 | 18.818.4 to 19.1 | 28.227.7 to 28.6 | 56.756.2 to 57.2 | 99.699.6 to 99.7 | 100>99.99 |
| Any planet's house (whole-sign) | 1.81.7 to 1.9 | 4.44.2 to 4.6 | 8.98.6 to 9.2 | 13.212.9 to 13.6 | 25.825.4 to 26.3 | 51.451.0 to 51.9 | 94.694.4 to 94.8 |
| Navamsa Lagna | 15.114.7 to 15.4 | 37.537.1 to 38.0 | 75.274.7 to 75.6 | 99.799.6 to 99.7 | 100>99.99 | 100>99.99 | 94.193.9 to 94.4 |
| Moon sign | 0.060.04 to 0.08 | 0.140.11 to 0.18 | 0.320.26 to 0.37 | 0.480.42 to 0.55 | 0.920.84 to 1.02 | 1.91.7 to 2.0 | 3.73.5 to 3.9 |
| Moon nakshatra | 0.120.09 to 0.16 | 0.370.31 to 0.43 | 0.760.68 to 0.85 | 1.11.0 to 1.3 | 2.12.0 to 2.2 | 4.13.9 to 4.3 | 8.27.9 to 8.4 |
| Moon nakshatra pada | 0.550.48 to 0.62 | 1.41.3 to 1.5 | 2.92.8 to 3.1 | 4.44.2 to 4.6 | 8.58.2 to 8.7 | 16.616.2 to 16.9 | 33.032.6 to 33.5 |
| Starting Mahadasha lord | 0.120.09 to 0.16 | 0.370.31 to 0.43 | 0.760.68 to 0.85 | 1.11.0 to 1.3 | 2.12.0 to 2.2 | 4.13.9 to 4.3 | 8.27.9 to 8.4 |
| Manglik status | 0.440.38 to 0.51 | 1.11.0 to 1.2 | 2.22.1 to 2.4 | 3.33.1 to 3.5 | 6.46.1 to 6.6 | 13.012.7 to 13.3 | 24.624.1 to 25.0 |
| Day of week | 0.120.09 to 0.16 | 0.330.27 to 0.39 | 0.720.64 to 0.81 | 1.00.9 to 1.1 | 2.12.0 to 2.3 | 4.24.0 to 4.4 | 8.58.2 to 8.7 |
| Tithi | 0.120.09 to 0.16 | 0.340.29 to 0.41 | 0.760.68 to 0.85 | 1.11.0 to 1.2 | 2.22.1 to 2.4 | 4.34.2 to 4.6 | 8.58.2 to 8.8 |
| Day of week or tithi | 0.240.19 to 0.29 | 0.670.59 to 0.75 | 1.51.4 to 1.6 | 2.22.0 to 2.3 | 4.34.1 to 4.5 | 8.48.1 to 8.6 | 16.215.9 to 16.6 |
| First Mahadasha end date: median shift in days (90th percentile under it) | 8(10) | 19(25) | 39(50) | 59(75) | 116(154) | 231(306) | 477(659) |
Manglik status follows the site's rule: Mars in the 1st, 2nd, 4th, 7th, 8th or 12th house from the Lagna or from the Moon. The day of week follows the civil date; the tithi includes the paksha.
How fast the Lagna changes
What the figures mean
- A one-hour error changes the Lagna in about half of charts. 50.2% (49.7 to 50.7%). It is about one in eight at 15 minutes (12.7%) and almost every chart at two hours (94.3%). Because houses are whole-sign, every planet's house changes in exactly these charts.
- The Moon is far steadier than the Lagna. At one hour the Moon's sign changes in 1.9% of charts and its nakshatra in 4.1%; at two hours, 3.7% and 8.2%. The starting Mahadasha lord changes in the same charts as the nakshatra, because neighbouring nakshatras never share a lord.
- The Navamsa Lagna is the most fragile item. It changes in 37.5% of charts with a 5-minute error and 75.2% at 10 minutes. It dips at two hours (94.1%) because the Lagna has often moved a whole number of Navamsa cycles and lands on the same Navamsa sign again.
- Dasha dates move by months, not years, for realistic errors. The first Mahadasha end date moves by a median of 39 days at a 10-minute error, 231 days at one hour (90th percentile 306) and 477 days at two hours: about 3.9 days per minute. Every later Mahadasha date moves by the same amount.
- Manglik status and the calendar are mostly safe for small errors. Manglik status flips in 3.3% of charts at 15 minutes, 13.0% at one hour and 24.6% at two hours. The day of week or tithi changes in 2.2% at 15 minutes and 8.4% at one hour.
Check against theory
The Lagna passes through all twelve signs once a day, so a sign lasts two hours on average and a shift of s minutes should change it in about s/120 of charts. The expected figures come from scanning the Lagna minute by minute over a day for the largest Indian cities, and from the Moon's average daily motion. Each cell is computed / expected, in per cent.
| Item | 2 min | 5 min | 10 min | 15 min | 30 min | 60 min | 120 min |
|---|---|---|---|---|---|---|---|
| Lagna sign | 1.7 / 1.7 | 4.2 / 4.2 | 8.5 / 8.3 | 12.7 / 12.5 | 24.9 / 25.0 | 50.2 / 50.0 | 94.3 / 94.2 |
| Moon nakshatra | 0.1 / 0.1 | 0.4 / 0.3 | 0.8 / 0.7 | 1.1 / 1.0 | 2.1 / 2.1 | 4.1 / 4.1 | 8.2 / 8.2 |
| Moon sign | 0.1 / 0.1 | 0.1 / 0.2 | 0.3 / 0.3 | 0.5 / 0.5 | 0.9 / 0.9 | 1.9 / 1.8 | 3.7 / 3.7 |
| Moon pada | 0.5 / 0.5 | 1.4 / 1.4 | 2.9 / 2.7 | 4.4 / 4.1 | 8.5 / 8.2 | 16.6 / 16.5 | 33.0 / 32.9 |
The two agree within about half a percentage point everywhere, so the computation behaves as the astronomy says it should.
Limits
- Sensitivity, not truth. This measures how much the calculation changes when the time is wrong. It says nothing about whether any reading or prediction is true.
- Simulated births. Dates, times and places are random draws, not real birth records. Recorded times bunch at round figures such as 6:00 or 6:30, which the study does not model.
- India, 1960 to 2020. Places are weighted by today's population, not by where births happened in each year. Births far from India's latitudes may differ, mostly at the two-hour end.
- One engine, one set of settings. Lahiri ayanamsa, whole-sign houses and the mean node. Other ayanamsas, house systems such as Bhava Chalit, and the true node were not tested.
- Fixed shifts. Each error is exactly the stated size, earlier or later. A real uncertainty is usually a range, so read the table as a guide to a range.
- Intervals cover sampling only. The 95% intervals show the uncertainty from using 20,000 random births, not from the choices above. Dasha day figures are read from a fine histogram and are good to about two per cent.
How to cite
Cite the page and link to it where you can. The text, the figures and the data files are free to reuse with credit.
Janmagyan (2026). How much does an error in birth time change a Kundli? A study of 20,000 simulated Indian births. https://janmagyan.com/studies/birth-time
Download the data
The same figures as the table, read from the study's own results file.
Read next
Cast your free Kundli
Not sure of your time? Make the chart at the earliest and the latest time you think possible, and rely on what stays the same.