The 27 nakshatras, and what matching actually reads from them

Your nakshatra is a coordinate, not a character. The sidereal zodiac is divided into twenty-seven equal arcs of thirteen degrees and twenty minutes each, and your birth star is whichever arc the Moon occupied at the minute you were born. That is the entire definition, and everything a matching report says about your star is read off a table indexed by that one number.

Five columns of that table do the work. Knowing which five, and knowing how each column is laid out, is enough to follow any report you are handed and to spot the places where two reports could honestly differ.

What exactly is a nakshatra measuring?

The Moon’s longitude, expressed in twenty-sevenths of a circle instead of in degrees. Nothing else about the chart is involved: not the Sun, not the ascendant, not the date.

The Moon travels roughly thirteen degrees a day against the background stars, which is very close to the width of one nakshatra. So a nakshatra is, in practical terms, about a day wide. Two babies born twenty hours apart usually share a birth star. Two born forty hours apart usually do not, and if they are close to a boundary the answer can turn on the recorded minute.

The word roughly is carrying weight there. The Moon’s orbit is an ellipse, so it does not cross the background at a constant rate: near perigee it moves noticeably faster than near apogee, and the time it takes to travel one thirteen-degree arc varies by a few hours either side of a day depending on where in the month the birth fell. That is why “a nakshatra is a day” is a rule of thumb for conversation and never a thing a calculation does. The Moon’s longitude is read at the instant and divided; nothing counts days forward from an anchor, because counting days would drift by a whole star inside a single lunar month.

Where the arcs begin is the one genuinely contested part. The twenty-seven divisions are fixed relative to each other, but the zero point of the sidereal zodiac is a convention, and several are in use. Shift the zero point and all twenty-seven boundaries shift together. That is why a Moon sitting a few arcminutes from an edge can be given two different birth stars by two competent calculations, and it is the whole subject of the post on the ayanamsa.

Everything downstream of the zero point is astronomy rather than convention. The Moon’s position at a given instant is computed against the published JPL planetary and lunar ephemerides, the same tables spacecraft navigation uses, and can be checked independently in JPL Horizons by anyone who wants to. The disagreement between traditions is never about where the Moon was.

A hundred and eight positions, not twenty-seven

Each nakshatra divides again into four padas of three degrees and twenty minutes. Twenty-seven times four is a hundred and eight, and that is the full vocabulary of star-based matching: nothing finer than a pada is visible to any of the eight kootas, and nothing coarser than a star is enough for four of them.

That number is why the base rates on this site are stated over 11,664 pairings. A hundred and eight positions for one person, a hundred and eight for the other, ordered because several of the rules are not symmetric. Multiply them and you have the whole population, not a sample drawn from it. Every rate quoted here was computed by scoring every one of those pairings under the rule in question, so there is no sampling error to report and no seed that would make a second run come out differently.

This matters more than it sounds. A base rate computed from a survey of couples inherits whatever biases put those couples in the survey. A base rate computed exhaustively over the position space avoids sampling error within that space, but still depends on the chosen rules and equal weighting of positions. It answers a narrower question — how often does this rule fire across all possible birth stars, not how often does it fire in Maharashtra — but it answers it without an error bar.

Which five attributes does matching actually read?

The ruling lord, gana, yoni, nadi and rajju. That is the whole of it. The classical description of a nakshatra is far richer: a presiding deity, a symbol, a shape, a direction, a temperament, a set of activities held suitable under it. Marriage matching consults almost none of that.

• The ruling lord. Nine planets in a fixed repeating order — Ketu, Venus, Sun, Moon, Mars, Rahu, Jupiter, Saturn, Mercury — which runs three complete times across the twenty-seven stars. Ashwini takes Ketu, and so do Magha and Mula, nine and eighteen places later. Matching uses this indirectly: the lord drives the dasha timing cycle, while it is the Moon-sign lords, not the star lords, whose friendship graha maitri scores.
• Gana. One of three temperament groups, Deva, Manushya and Rakshasa, nine stars in each. Worth 6 points, and a mismatch between particular groups is what gana dosha names.
• Yoni. An animal and a sex. Worth 4 points, and scored from a matrix of animal against animal, not from a rule you could state in a sentence.
• Nadi. One of three channels, Adi, Madhya and Antya, nine stars in each, laid down in a repeating six-star wave instead of three blocks, so how far apart two birth stars sit tells you nothing about whether they share one. The largest single koota at 8 points, the one that most often decides a total on its own, and the subject of the post on same-nadi refusals, which walks the wave, the membership of the three groups and the cancellation rules the schools disagree about.
• Rajju. Five body-part groups running from foot to head. Worth nothing at all in Ashtakoota, because Ashtakoota does not contain it. It belongs to the South Indian systems, where it is not a score but a veto.

All five are printed side by side, for every star, on the nakshatra grid. The rest of this piece is about the shape of those columns, because the shape explains several results that look arbitrary on a report.

The rajju column runs foot to head and back

Foot, waist, navel, neck, head. Then back down again.

That is rajju: a nine-star wave of Pada, Kati, Nabhi, Kantha, Shiro, then Kantha, Nabhi, Kati and Pada, running three complete cycles across the twenty-seven. The head sits at the turn, so each cycle passes through it once while passing through the other four groups twice. That is why the counts are uneven: six stars each in Pada, Kati, Nabhi and Kantha, and only three in Shiro — Mrigashira, Chitra and Dhanishta, the fifth star of each cycle.

The ascending and descending halves are named, aroha and avaroha, and some South Indian formulations care about which half a star sits in as well as which group. Two stars in the same group on the same limb of the wave and two stars in the same group on opposite limbs are not always treated alike. That distinction never appears on a North Indian report, because rajju never appears on a North Indian report at all — a family that has had a match refused on rajju in Tamil Nadu and cleared in Delhi is not looking at an error by either astrologer.

Why are there fourteen yoni animals for twenty-seven stars?

Thirteen of the animals appear twice, once as a male and once as a female, and the fourteenth appears only once. Thirteen pairs is twenty-six stars; the mongoose makes twenty-seven.

Ashwini is the male horse and Shatabhisha is the female horse. Rohini is the male serpent and Mrigashira the female. The pairing is not by adjacency and there is no pattern to which member of a pair comes first, so the column has to be read rather than derived. Uttara Ashadha is the male mongoose, and inside the twenty-seven there is no female mongoose at all.

That gap is where Abhijit comes in, and it is the only job Abhijit does in our tables. It supplies the female mongoose, and it is used for yoni and for nothing else — never for gana, never for nadi, never for rajju, never for the tara count. A source that treats Abhijit as a twenty-eighth division of the zodiac has done something unusual, and it ought to say so, because every count of nine and every count of twenty-seven downstream would change if it were.

The animal is a lookup key into a fourteen-by-fourteen matrix of scores, not a description of anybody. Full marks on yoni go to just 7.3% of all pairings, which makes a partial yoni score the normal case and not a warning. The piece on that figure works through why the matrix is so stingy.

Which nine stars are split across two rashis?

Krittika, Mrigashira, Punarvasu, Uttara Phalguni, Chitra, Vishakha, Uttara Ashadha, Dhanishta and Purva Bhadrapada — and for those nine, the pada alone decides the Moon sign. The other eighteen sit entirely inside one rashi whichever pada you were born in.

The arithmetic is unavoidable. A rashi is thirty degrees and a nakshatra is thirteen and a third, so two and a quarter stars fill a sign and the boundaries cannot line up except every fourth sign. Three of the twelve rashi boundaries fall exactly on a nakshatra edge; the other nine fall inside a star and cut it.

The three that line up are worth knowing by name, because they are the only places the two grids agree. Ashlesha ends exactly where Cancer ends, Jyeshtha exactly where Scorpio ends, Revati exactly where Pisces ends. The spacing is not an accident. Two and a quarter stars fill a sign, so the leftover quarter closes only after four signs have gone by, and nine whole nakshatras cover four whole rashis. That puts an alignment at the end of every fourth sign and nowhere else, and it is another thing a reader can confirm on the grid in under a minute.

The cut is in a repeating rhythm of one, two and three padas. Krittika gives its first pada to Aries and the rest to Taurus. Mrigashira splits two and two between Taurus and Gemini. Punarvasu gives three padas to Gemini and the last one to Cancer. Then the pattern starts again with Uttara Phalguni, and again with Uttara Ashadha, so once you know the rhythm you can work out any of the nine without a table.

This is where a sloppy birth time does its damage. A wrong pada inside one of these nine stars produces a wrong Moon sign, and the Moon sign is what varna, vashya, graha maitri and bhakoot are all computed from — four kootas and 15 of the thirty-six points, moving together on one wrong quarter of a star. The birth-time sensitivity tool shows how far a given chart would have to move before any of that changes.

For nine of the twenty-seven stars a wrong quarter of a star is a wrong Moon sign, and the Moon sign is what four of the eight kootas are computed from.

The columns are not equally consequential

They decide things very unevenly, and the ranking is not the one most reports imply. The difference is measurable across the whole population, not estimated from some corner of it.

Nadi produces a zero on 33.3% of pairings, gana on 22.2%, and bhakoot on 50% — that last one is a coin flip, by construction, and it is the second-largest koota. Between them these three account for most of the zeroes anybody ever sees on a report.

Now the other direction. Varna awards full marks on 62.5% of pairings, graha maitri on 38.9%, and yoni on 7.3%. These frequencies show how often the tables return each result in the enumeration. They do not establish predictive value or explain why a practitioner refused a match. A single report does not supply this frequency context, which is why we publish it separately.

The ruling lord is the odd one out on this list, because it does not feed a koota directly at all. It sets the dasha sequence, which is a timing scheme and not a compatibility test, and it is the reason two people with the same birth star are told they are running the same planetary period at different ages. When a consultation moves from “these charts score such-and-such” to “wait two years”, it has usually moved from the koota tables to the dasha cycle without saying so, and those are two different systems with two different sets of assumptions behind them.

How would you catch an error in a table like this?

A twenty-seven-row lookup table looks too simple to get wrong, which is exactly why nobody checks one, and a single wrong cell moves every score that touches that star. Two habits catch most of it.

The first is transcription against a second independent statement of the same table, cell by cell, so that a slip has to be made twice in the same direction before it survives. The second is a set of assertions the tables run against themselves whenever they load. Gana must come out nine, nine and nine. So must nadi. Rajju must give six, six, six, six and three, with the three sitting at the turn of each wave. Abhijit must never appear in a gana, nadi or rajju group, because it is a yoni entry and nothing else. Every nakshatra must have exactly one of each attribute, and no yoni animal may appear three times.

None of that proves the tables match any particular manuscript. It proves something narrower and still useful: that they are internally consistent, that a copy-paste error in the middle of the rajju column will stop a build instead of quietly changing somebody’s verdict, and that the structural facts described above — the nine-star rajju wave, the thirteen animal pairs, the nine stars a rashi boundary cuts — are properties the code enforces rather than claims this page makes.

A reader has a cheaper version of the same test available. Count the Adi stars on the grid. If there are not nine, something is wrong, and you have found it without knowing anything about astrology. Most claims in this field cannot be checked that easily; the parts that can be, should be.

“Deva, Manushya, Rakshasa” is a three-way sort

The gana names and the yoni animals do far more work in ordinary conversation than they do in the calculation. Deva, Manushya and Rakshasa carry a great deal of freight in everyday Indian usage; in the koota they are a sort with nine stars in each bucket, and the rule tests only whether two people fall in buckets a text calls compatible.

The same holds for the animals. The matrix scores mongoose against serpent and cat against rat, and those scores are what the koota reads. Nobody in the classical source is claiming the person is a mongoose. Reading yoni as a character assessment is a modern habit and a fairly recent one.

We report what these tables conclude and how common each conclusion is. The step from “this pairing scores zero on gana under the Charak reading” to “these two people are incompatible” is a step we do not take, and it is the step at which most of the harm in this subject is done.

What should you check on your own report?

Four things, in order, and all four are checkable in a few minutes. None of them requires you to accept or to reject the system.

Start with the birth star and pada for each of you, and confirm them. Then find the Moon sign and confirm it follows from that pada and not from a Sun sign; a surprising number of confusions start there, especially for anyone who also reads a Western horoscope. Then read the five attributes off the grid for both stars. Then check that the koota scores in the report follow from those attributes.

If they do, the report is at least internally consistent, and any disagreement with a second report is about rules rather than about arithmetic. That is a far easier conversation to have with a family than an argument about whether the first astrologer was careless, and it is usually the true one. The nakshatra finder gives the star and pada from a birth time, and prints the boundary distance so you can see how close a call it was.

Where can you check each of the 27 nakshatras?

Use the reference pages below to inspect the exact five fields matching reads for any birth star. Each page is generated from the same table as the calculators, so the reference and the result cannot drift apart.

• Ashwini
• Bharani
• Krittika
• Rohini
• Mrigashira
• Ardra
• Punarvasu
• Pushya
• Ashlesha
• Magha
• Purva Phalguni
• Uttara Phalguni
• Hasta
• Chitra
• Swati
• Vishakha
• Anuradha
• Jyeshtha
• Mula
• Purva Ashadha
• Uttara Ashadha
• Shravana
• Dhanishta
• Shatabhisha
• Purva Bhadrapada
• Uttara Bhadrapada
• Revati

One last thing to check, which almost nobody does. Confirm that the time zone applied to the birth was the right one. India has run on a single national offset for a long time, but births recorded before it settled, or abroad, or in a family that kept local time, can land in the wrong hour entirely — and the historical time zone data is more complicated than a single offset suggests. An hour is most of a pada.