Answer. Vedic calculators disagree mainly because they use different conventions, not different astronomy. Measured on 200,000 random charts, switching the ayanamsa from Lahiri to Raman changes the Moon's nakshatra for 11% of people and, for everyone else, moves every Mahadasha date by 238 to 792 days (median 634); counting a dasha year as 360 instead of 365.25 days moves them by 210 days by age 40; and dropping the 2nd house from the Manglik rule changes the Mars-placement rate from 50% to 42%.
By AstroSutra · Published 15 November 2026 · Last updated 15 November 2026 · Swiss Ephemeris for all comparisons · Data: CSV · JSON · CC BY 4.0
| Convention | Typical choices | What it changes | Size of the effect (measured) |
|---|---|---|---|
| Ayanamsa | Lahiri (the Indian national calendar, Rashtriya Panchang, and most Indian software), KP, Raman, True Chitra, Fagan-Bradley | Every sidereal position | Raman vs Lahiri: nakshatra differs for 11.0% of charts; for the rest, Mahadasha dates move 238–792 days |
| Dasha year length | 365.25 days; 360-day (savana) year; 365.2564-day sidereal year | Every dasha date after birth | 5.25 days per dasha-year between 365.25 and 360: 210 days by age 40 |
| Birth time | As recorded vs rounded | Lagna, houses, dasha dates | ±15 min: Lagna sign in doubt for 24.9% of charts; a 15-minute error moves Mahadasha dates a median 58 days when the first lord is unchanged (birth-time study) |
| Manglik houses | Six houses (with the 2nd) or five | Manglik dosha | 50.1% vs 41.7% of charts with Mars in a Manglik house from the Lagna (Manglik study) |
| Manglik reference points | Lagna only, or Lagna, Moon and Venus | Manglik dosha | 50.1% from the Lagna; 88.1% from any of the three |
| Manglik cancellations | None, or a set of traditional exceptions | Manglik dosha | 50.1% have the placement from the Lagna; under AstroSutra's set, 11.8% are reported as Manglik and 11.9% as mild (under 28) |
| Kuta cancellations | None, or Bhakoot/Nadi exceptions | Guna Milan total | Average score 20.9 without, 23.3 with AstroSutra's set (Guna Milan study) |
| Bhakoot 5/9 axis | Counted as dosha or not | Guna Milan total | Bhakoot dosha in 50.0% of pairings with 5/9, 33.3% without |
| Lunar node | True node or mean node | Rahu/Ketu positions and transit dates | Rahu's sign differs for 3.0% of charts; sign-change dates differ by up to a month |
| Sade Sati rule | Whole signs from the Moon sign, or 45° either side of the Moon's degree | Sade Sati start and end | Up to 15° of Saturn motion apart: more than a year (Saturn averages about 12° a year) |
| Time zone of the date | Local, IST or UTC | The calendar date of a transit | One day when a change falls near midnight |
Vedic astrology uses the sidereal zodiac, which is offset from the tropical zodiac by the ayanamsa. Schools disagree on the exact offset.
Ayanamsa on 1 January of each year (Swiss Ephemeris):
| Ayanamsa | 1950 | 2000 | 2026 | 2050 | Difference from Lahiri |
|---|---|---|---|---|---|
| Lahiri (Chitra Paksha) | 23°09′31″ | 23°51′25″ | 24°13′13″ | 24°33′20″ | — |
| True Chitra | 23°08′32″ | 23°50′24″ | 24°12′10″ | 24°32′16″ | −1.0′ |
| Krishnamurti (KP) | 23°03′43″ | 23°45′37″ | 24°07′24″ | 24°27′32″ | −5.8′ |
| Raman | 21°42′45″ | 22°24′39″ | 22°46′26″ | 23°06′34″ | −1°26.8′ |
| Fagan-Bradley (Western sidereal) | 24°02′31″ | 24°44′25″ | 25°06′13″ | 25°26′20″ | +53.0′ |
How often each ayanamsa gives a different answer from Lahiri, in 200,000 random charts:
| Ayanamsa | Moon sign | Nakshatra | Pada | Lagna sign | Sun sign | First Mahadasha lord | Mahadasha dates move, same first lord (median; range) |
|---|---|---|---|---|---|---|---|
| True Chitra | 0.06% | 0.11% | 0.47% | 0.05% | 0.04% | 0.11% | 5 days; 2–12 |
| KP | 0.33% | 0.70% | 2.91% | 0.31% | 0.32% | 0.70% | 42 days; 16–53 |
| Fagan-Bradley | 2.91% | 6.65% | 26.4% | 2.92% | 2.93% | 6.65% | 387 days; 145–484 |
| Raman | 4.87% | 10.95% | 43.4% | 4.84% | 4.76% | 10.95% | 634 days; 238–792 |
For KP, Raman and Fagan-Bradley the offset from Lahiri is constant, so the shift depends only on the first Mahadasha lord's period: the median is the value for a Jupiter (16-year) start. For the charts whose nakshatra changes (10.95% with Raman, 6.65% with Fagan-Bradley), the first Mahadasha lord changes too, and the whole sequence of periods is different, not just shifted.
A rule of thumb: an ayanamsa difference of D arc-minutes changes the nakshatra for about D ÷ 8 percent of charts (a nakshatra is 800′ wide), and moves Mahadasha dates by about 0.46 × D × Y days, where Y is the first Mahadasha lord's period in years. For Raman (D = 87′) and a Venus first Mahadasha (Y = 20), that is about 790 days. Small ayanamsa differences matter much more for dasha dates than for signs.
Vimshottari periods are given in years, and calculators convert years to days differently. AstroSutra uses 365.25 days. Some software offers a 360-day (savana) year, which follows a reading of the classical texts; others use the 365.2564-day sidereal year.
| Dasha-years since birth | 365.25 vs 360 days |
|---|---|
| 10 | 52.5 days |
| 20 | 105 days |
| 40 | 210 days |
| 60 | 315 days |
| 120 (full cycle) | 630 days (about 21 months) |
The 365.25 vs 365.2564 difference is 0.0064 days per year: under a day over a lifetime.
Sade Sati starts when Saturn enters the sign before the natal Moon sign, so its dates are Saturn's sign-change dates. Those move with the ayanamsa.
Saturn's sidereal sign changes, 2025–2030 (IST dates with Lahiri; difference in days for the others):
| Saturn enters | Lahiri (IST) | KP | Raman | Fagan-Bradley |
|---|---|---|---|---|
| Pisces | 29 Mar 2025 | −0.8 | −11.8 | +7.4 |
| Aries | 3 Jun 2027 | −1.0 | −13.8 | +9.5 |
| Pisces (retrograde) | 20 Oct 2027 | +1.2 | +19.0 | −11.2 |
| Aries | 23 Feb 2028 | −1.0 | −16.6 | +8.7 |
| Taurus | 8 Aug 2029 | −1.9 | −21.8 | does not happen |
| Aries (retrograde) | 5 Oct 2029 | +1.9 | +22.7 | does not happen |
| Taurus | 17 Apr 2030 | −0.8 | −12.5 | +7.2 |
With the Fagan-Bradley ayanamsa, Saturn's 2029 retrograde loop stays in Aries, so the brief entry into Taurus in August–October 2029 does not occur at all and Saturn first enters Taurus on 24 April 2030: 259 days after the Lahiri date. For a Gemini Moon that is the start of Sade Sati; for a Pisces Moon, its end. A retrograde "pause" in a Sade Sati can exist under one ayanamsa and not another.
Two further conventions:
If two apps disagree on a Mahadasha date, the size of the gap points to the cause:
| Gap | Most likely cause |
|---|---|
| A day or two | Time zone, rounding the birth time, or adding calendar years instead of 365.25-day years |
| Weeks | KP vs Lahiri ayanamsa, or a birth time differing by about 5–15 minutes |
| Months, growing with age | 360-day vs 365.25-day dasha year |
| A year or more, or a different sequence | Raman vs Lahiri ayanamsa, or a Moon in a different nakshatra |
From the Manglik study (200,000 charts, 1950–2009):
| Convention | Share of charts |
|---|---|
| Mars in 1, 2, 4, 7, 8 or 12 from the Lagna | 50.1% |
| Mars in 1, 4, 7, 8 or 12 from the Lagna (no 2nd house) | 41.7% |
| Mars in a Manglik house from the Lagna, the Moon or Venus | 88.1% |
| Reported as Manglik under AstroSutra's cancellation rules (uncancelled 1st, 4th, 7th, 8th or 12th from the Lagna), under 28 | 11.8% |
| The same, at age 28 or over (severity halved) | 4.7% |
A calculator that says "Manglik" for half of all charts and one that says it for about 1 in 10 can both be applying a documented rule correctly. The number only means something next to the convention.
Rahu and Ketu are the lunar nodes. The mean node moves smoothly; the true node wobbles around it by up to about 2°. AstroSutra uses the true node.
| Rahu enters | True node (IST) | Mean node (IST) |
|---|---|---|
| Aries | 17 Mar 2022 | 12 Apr 2022 |
| Pisces | 29 Nov 2023 | 30 Oct 2023 |
| Aquarius | 30 May 2025 | 18 May 2025 |
| Capricorn | 25 Nov 2026 | 5 Dec 2026 |
| Sagittarius | 24 May 2028 | 24 Jun 2028 |
The Rahu transit dates commonly reported in India (12 April 2022, 30 October 2023, 18 May 2025) are mean-node dates. Because the true node wobbles, it can cross a sign boundary more than once within a few days; the table gives its first crossing.
AstroSutra's choices are listed on one page: How AstroSutra calculates.
| Item | Value |
|---|---|
| Sample | 200,000 birth moments, uniform to the minute, UTC 1950-01-01 to 2009-12-31, seed 20261116, seven cities (Delhi, Mumbai, Chennai, London, New York, Toronto, Singapore) |
| Ephemeris | Swiss Ephemeris; every ayanamsa applied to the same tropical positions; Lahiri is the reference |
| Ayanamsa modes | Swiss Ephemeris built-ins: SIDM_LAHIRI, SIDM_TRUE_CITRA, SIDM_KRISHNAMURTI, SIDM_RAMAN, SIDM_FAGAN_BRADLEY |
| Dasha | Vimshottari; the arithmetic matches AstroSutra's live dasha service to 0.0 days on 1,000 checked dates |
| Transits | Saturn sign changes found by daily scan, node sign changes by 2-hour scan, each bisected to one minute; dates in IST |
| Manglik and Guna Milan figures | From the linked studies, same engine |
| Reproduce | backend/scripts/geo_research_studies.py conventions --n 200000 |
Why do different websites give different Sade Sati start dates? Mostly the ayanamsa: Saturn's sign changes fall 11–26 days away from the Lahiri dates with the Raman ayanamsa and 7–22 days away with Fagan-Bradley, and months apart when a retrograde loop crosses a boundary under one ayanamsa but not the other (259 days in 2029–30). A site that uses 45° around the Moon's degree instead of whole signs can differ by more than a year, and a change near midnight can land on different dates in different time zones.
Why does my Mahadasha date differ between two astrology apps? In order of size: a different ayanamsa (Raman vs Lahiri moves dates by 238–792 days, or changes the whole sequence for 11% of people), a 360-day instead of 365.25-day dasha year (210 days after 40 dasha-years), a different birth time (about 1.5 to 5 days per minute), or a different time zone.
Which ayanamsa is most accurate, Lahiri, Raman or KP? They are conventions rather than measurements, so accuracy is not the right test. Lahiri is the most widely used in India; KP differs from it by 5.8′ and Raman by 1°26.8′. What matters is that a calculator says which one it uses.
What is the Lahiri ayanamsa value in 2026? 24°13′13″ on 1 January 2026 (Swiss Ephemeris, Lahiri/Chitra Paksha).
Why do Manglik calculators disagree? They count Mars from different houses (five or six), from different points (the Lagna, or also the Moon and Venus) and apply different cancellation rules. These choices move the share of Manglik charts from 88% down to about 5–12%.
How to cite: AstroSutra (2026). "Why do Vedic astrology calculators disagree?" https://astrosutra.app/research/why-vedic-calculators-disagree/
Compute yours: AstroSutra's free calculators show the calculation method with each result: Mahadasha, Sade Sati, Manglik, Nakshatra. No signup.