Our ground rules
Jyotish is not one system but several, and they disagree on questions the texts themselves cannot settle. An astrologer who picks a different one after the event can explain any outcome โ which is exactly why such explanations are worth nothing. So we picked ours in advance, wrote them down, and this is the list.
So every such choice is fixed in advance and published here. The engine reads its defaults from the same file that generates this page, so the two cannot drift apart. Each decision is labelled with what kind of reason it rests on โ that matters more than the decision itself. Some are astronomically determined. Most are convention. A few are honestly unsettled, and say so.
Astronomically determined
| True or mean ayanamsa? The true value is what the sidereal transformation actually subtracts from a tropical longitude. Reporting the mean value instead leaves the printed ayanamsa disagreeing with the chart's own planetary positions by up to about 17 arcseconds. | True (nutation-corrected) not: Mean If changed: A constant inconsistency of up to 17 arcseconds. Small, but wrong. |
Verified against published almanacs
| Where does a festival fall when its tithi contains no sunrise? A kshaya tithi is shorter than the gap between two sunrises, so sunrise-matching alone drops the festival entirely. Ugadi 2026 is exactly this case. Placing it on the day the tithi begins gives 19 March, matching published almanacs. | On the day the tithi begins not: Skip the observance, Use the following day If changed: Affected festivals move by a day or vanish. |
| Are all festivals fixed by sunrise? Both are rites performed at a particular time of day: Shivaratri is a night vigil, Lakshmi Puja an evening one. Using sunrise for either puts it a day later than every published almanac. Verified against 2026: Shivaratri 15 February, Deepavali 8 November. | No โ Shivaratri by nishita kala, Deepavali by pradosha not: Sunrise for everything If changed: Those two festivals land a day late. |
| What counts as sunrise? This is the definition published panchangams use. Disc-centre sunrise differs by about a minute and would desynchronise every muhurta window. | Upper limb touching the horizon, with atmospheric refraction not: Disc centre, Civil twilight If changed: All sunrise-derived windows shift by roughly a minute. |
| How are Rahu Kaal, Yamaganda and Gulika computed? Matches published Telangana timings: Saturday Rahu Kaal computes to 09:03-10:35 at Nalgonda against a published 09:00-10:30. | Eighths of the sunrise-to-sunset day, by weekday not: Fixed clock times, Eighths of a 24-hour day If changed: Windows move, by more in winter and summer than at the equinoxes. |
Regional practice
| Sidereal or tropical zodiac? Jyotish is sidereal throughout. The Western pages on this site are tropical, and are labelled as such; the two are never mixed in one reading. | Sidereal not: Tropical If changed: Every position shifts by roughly 24 degrees, about 0.8 of a sign. |
| Which ayanamsa? The Indian government's Calendar Reform Committee adopted Lahiri in 1955 and it is what Telugu panchangams use. Note what this means: the value was settled by a committee, not by scripture โ the texts do not resolve it. | Lahiri (Chitrapaksha) not: Raman, Krishnamurti (KP), True Chitra, Yukteshwar If changed: Raman differs from Lahiri by about 1.2 degrees, enough to move a nakshatra and occasionally a sign. |
| Lunar or solar month reckoning? Telangana and Andhra Pradesh follow Chandramana: Ugadi is the new year and it moves across a 26-day range, which is the signature of a lunar reckoning. The solar month is displayed too, because the sankrantis live in it and some readers follow it. | Chandramana (lunar) primary, Souramana shown alongside not: Souramana (solar) primary, as in Tamil Nadu and Kerala If changed: Month names change; festival attribution to months changes. |
| Amanta or purnimanta months? Amanta is used across Telangana, Andhra Pradesh, Karnataka and Maharashtra. North Indian purnimanta reckoning names the Krishna-paksha half of each month differently. | Amanta (month ends at the new moon) not: Purnimanta (month ends at the full moon) If changed: The same lunar day carries a different month name for half of every month, which reassigns Krishna-paksha festivals. |
Dominant convention
| Mean or true lunar node for Rahu and Ketu? Mean node is what most Indian software and published panchangams use. The true node oscillates around it by up to about 1.7 degrees. | Mean node not: True node If changed: Rahu and Ketu move by up to 1.7 degrees, which can change nakshatra. |
| Which house division for Vedic charts? Whole-sign houses are standard in Parashari practice: the sign containing the ascendant is the first house entire. Placidus and others are offered but are not the default. | Whole sign (bhava = rashi) not: Placidus, Equal, Koch, Sripati If changed: Planets move between houses, which changes yoga and dosha detection โ Mangal dosha in particular. |
| Which dasha system? Vimshottari is the most widely used and the one Telugu practitioners expect. Be clear about what this is not: Ashtottari, Yogini and Kalachakra are equally scriptural and give different periods for the same chart. No evidence distinguishes them. We commit to Vimshottari so that our timing cannot be retrofitted after the fact. | Vimshottari not: Ashtottari, Yogini, Kalachakra, Chara (Jaimini) If changed: Period boundaries move by years. Any timing claim changes. |
| How long is a dasha year? Matches mainstream Jyotish software. A 365.25-day variant also circulates and the texts do not settle it. | 365.2425 days not: 365.25 days, 360 days (savana) If changed: Long-range period ends shift by several days over a lifetime. |
| When does the weekday begin? Classical practice runs the vara sunrise to sunrise, not midnight to midnight. | Sunrise not: Midnight If changed: Weekday-dependent windows such as Rahu Kaal shift by a day. |
| At what moment are the five limbs quoted? A day is named for the tithi running when it begins, even if that tithi ends minutes later. This is standard almanac practice. | Sunrise not: Midnight, Moment of query If changed: Roughly a third of days would carry a different tithi label. |
| Is Nadi dosha ever cancelled? A widely accepted exemption. Others circulate โ same rashi different nakshatra, or the same nakshatra lord โ and are not applied here. | Yes, when both share a nakshatra but differ in pada not: No exemption, Broader exemption sets If changed: Nadi is worth 8 of 36, so this flips many verdicts. |
Sources conflict โ one picked and disclosed
| Which houses carry Mangal dosha, and from what reference? North Indian practice uses 1, 4, 7, 8, 12; South Indian adds the 2nd. Rather than pick one silently, both are evaluated and each hit is labelled with which rule it satisfies. Worth knowing: Mangal dosha is not prominent in the earliest classical corpus. It is largely a later and regional doctrine, despite the weight it now carries in matchmaking. | Houses 1, 2, 4, 7, 8, 12 โ checked from Lagna, Moon and Venus, all three reported not: North rule only, South rule only, Lagna reference only If changed: A chart can be Manglik under one rule and not the other. |
| Which Vashya table? Published Vashya matrices agree on the maximum and on which pairs score zero, but differ in the intermediate 0.5/1/2 band. We use one and print the per-koota breakdown so it can be compared. | Class-based matrix, documented in the matching module not: Other published matrices If changed: The total can differ by up to 1.5 points. |
| Which Yoni table? The enemy pairs and the identical-yoni maximum are consistent across sources, but published tables conflict on the 1-3 middle band. Same treatment as Vashya: pick one, document it, show the breakdown so it can be compared. | Seven classical enemy pairs at zero, remainder by temperament grouping not: Other published 14x14 matrices If changed: The total can differ by up to 2 points. |
Does any of this stand up to modern measurement?
Some of it does, and we checked. The classical astronomy states numbers โ the length of a month, the length of a year โ and numbers can be tested against an ephemeris. Some pass to within a second. Others are out by minutes, and where a text is wrong this engine follows the measurement. See the full check.
The corpus we cite
Hand-curated. Nothing here is generated by a language model, and the interpretation prompts cannot invent citations โ a model asked for chapter and verse produces confident numbers that are often wrong. We name works, not verses: printed recensions differ, so verse-level precision would be false precision.
| Work | Author | Period | Note |
|---|---|---|---|
| Brihat Parashara Hora Shastra | attributed to Parashara | attributed to antiquity; surviving manuscripts much later | The most cited text in modern practice. Its manuscript tradition is late and contested, and printed editions differ in chapter count and content, so 'BPHS says' is less precise than it sounds. |
| Brihat Jataka | Varahamihira | 6th century CE | One of the securely datable foundations of the natal tradition. Varahamihira notes that his predecessors disagreed with each other. |
| Laghu Jataka | Varahamihira | 6th century CE | Varahamihira's condensed treatment of natal astrology. |
| Saravali | Kalyana Varma | c. 10th century CE | |
| Phaladeepika | Mantreswara | c. 13th-15th century CE | A systematic and widely used manual of interpretation. |
| Jataka Parijata | Vaidyanatha Dikshita | c. 15th-16th century CE | |
| Sarvartha Chintamani | Venkatesha Daivajna | c. 16th century CE | |
| Uttara Kalamrita | Kalidasa (not the poet) | medieval | |
| Muhurta Chintamani | Rama Daivajna | 16th century CE | A principal source for electional astrology and for the marriage compatibility tables. |
| Prashna Marga | anonymous, Kerala | 17th century CE | Horary astrology; also a source for several dosha doctrines. |
| Jaimini Upadesa Sutras | attributed to Jaimini | date uncertain | A distinct system, not a refinement of Parashari. Its dasha schemes give different timing for the same chart. |
| Surya Siddhanta | anonymous | c. 4th-5th century CE in surviving form | The astronomical half of the tradition, and the only part of it that measurement can adjudicate. Its lunar months are right to under a second; its sidereal year is about three and a quarter minutes too long and its precession rate seven per cent too fast. Every figure is checked live against the ephemeris on the ruleset page. |
Doctrines whose provenance does not hold up
Routinely presented as ancient. They are not. We still compute them, because readers ask โ and we label them.
Mangal (Kuja) dosha โ Routinely credited to the classical corpus, but it is not prominent in the earliest texts and is substantially a later and regional doctrine. Given how much weight it carries in matchmaking, that is worth stating plainly.
Kaal Sarp dosha โ Absent from the classical corpus. It appears in twentieth-century popular practice. We compute it because readers ask for it, and label it for what it is.
Pitru dosha โ Affliction of the 9th house and its lord. The name and the specific conditions are largely modern and vary widely by region.
What a citation does and does not do
"Phaladeepika describes this combination as auspicious" is a true, checkable statement about a text. It is not evidence that the combination is auspicious. Galen and Hippocrates are also authentic, ancient and carefully reasoned; nobody bleeds patients on their authority. Provenance establishes what a tradition holds, never whether it is right.