Prahar

Ashta Kuta

The eight kutas, one at a time

Ashta Kuta is eight separate tests worth eight different amounts, and the weights are the argument: Nadi alone is worth more than the first four put together. Knowing which of the eight a pair lost is the whole of what a total conceals.

What the eight read

Two facts from each chart, and no others: which nakshatra the Moon is in, and which rashi. Everything below is a table lookup on those two numbers. That is what makes the system cheap to compute and it is also its limit — a matcher can be exactly right about all thirty-six points and still be reading two facts out of a chart that contains several hundred. Four of the eight are one-directional, so which chart takes which role changes the answer; those four are marked.

The eight, and what each is worth

Varna1 point · one-directional
Whether the two temperaments can work at all, read from the element of each Moon rashi: water is Brahmin, fire Kshatriya, earth Vaishya, air Shudra, ranked four down to one. This is the only kuta scored by a rank rather than by a cell in a grid — the point is given when the vara's varna is not below the vadhu's, and withheld otherwise. It is the lightest of the eight and it settles nothing on its own.
Vashya2 points · one-directional
Which of the two naturally holds sway over the other. Each Moon rashi belongs to one of five classes — quadruped, human, watery, wild, insect. Sagittarius and Capricorn split at their halfway degree, so this kuta needs the exact longitude and not just the sign, which is why a matcher that asks only for a rashi cannot compute it correctly. The vara's class is the row and the vadhu's the column: 2 is the same class, 1 amenable, 0.5 partial, 0 not at all.
Tara3 points · one-directional
How each one's birth star falls when counted from the other's — the classical measure of whether the pairing is fortunate for each of them separately. Count from one nakshatra to the other inclusive and take the remainder on division by nine. The nine taras are Janma, Sampat, Vipat, Kshema, Pratyari, Sadhaka, Vadha, Mitra and Ati-Mitra; the third, fifth and seventh — Vipat, Pratyari, Vadha — are the inauspicious ones. A remainder of nought is the ninth, Ati-Mitra, which is the best of them and not a failure. Counted both ways because the count is not symmetric, one and a half points each way.
Yoni4 points · symmetric
Physical compatibility, read through the animal each nakshatra carries. Each of the twenty-seven is assigned one of fourteen animals and a gender. The same animal scores the full four; the seven sworn pairs — Horse and Buffalo, Elephant and Lion, Sheep and Monkey, Serpent and Mongoose, Dog and Deer, Cat and Rat, Cow and Tiger — score nought; everything between comes off a fourteen-by-fourteen grid. Reproductions of that grid disagree in a handful of its enemy-versus- neutral cells; the diagonal and the seven zeroes are not in dispute anywhere.
Graha Maitri5 points · symmetric
Whether the two minds get on, read from the natural friendship between the lords of the two Moon rashis — not the planets as they sit in either chart. Only the naisargika (permanent) relation counts; the tatkalika half of the five-fold relation is chart-specific and is not part of this kuta. Rahu and Ketu lord no sign, so only the seven classical grahas appear. Mutual friends, or one planet lording both rashis, take all five; mutual enemies take nothing.
Gana6 points · one-directional
Temperament in the ordinary daily sense — how each one behaves when not being watched. Each nakshatra is Deva, Manushya or Rakshasa. The table is sharply asymmetric and that is the point of it: a Manushya vara with a Deva vadhu takes the full six, while a Rakshasa vara with a Deva or Manushya vadhu takes nothing. It is the second heaviest kuta and the one most likely to zero a score that otherwise looked fine.
Bhakuta7 points · symmetric
How the two Moon rashis stand to each other — the classical read on the material fortunes of the household. Count between the two rashis both ways, each count inclusive, so the pair always sums to fourteen. Full marks unless the pair is 2/12, 5/9 or 6/8, each of which scores nought outright. Every other relation — 1/1, 3/11, 4/10, 7/7 — takes all seven. There is no partial credit anywhere in it, which is why a page that tells you a bhakuta dosha was found without naming which of the three it hit has withheld the only part worth knowing.
Nadi8 points · symmetric
Constitution, in the ayurvedic sense — vata, pitta, kapha — read from the nakshatra as Adi, Madhya or Antya. Different nadis take the full eight; the same nadi takes nothing. It is the single heaviest kuta, which is exactly why its exceptions are the most argued part of the whole system. Three pariharas are commonly cited and each is denied by some tradition; this page applies none of them and prints what that strictness cost. Why no parihara is applied →

Mangal dosha is not one of the eight

It is worth saying plainly, because a great many matchers fold it in and never mention that they have. Classical Ashta Kuta does not contain Mangal (Kuja, Manglik) dosha at all — the eight tables sum to thirty-six without it, and it is a separate test entirely: whether Mars sits in one of a set of houses counted from the ascendant, the Moon or Venus. Some practitioners treat an unmatched dosha as a veto over any score whatever; others as one factor among many. Folding it into the thirty-six would make it invisible, so this page computes it, reports it beside the total, and leaves the total alone. Both carrying it, or neither, is the classical cancellation; the one-sided case is what the whole doctrine is about.

Where the tables come from

The system is Ashta Kuta, or Ashtakoot Milan, as set out in the Muhurta Chintamani of Daivajna Ram in the sixteenth century, reproduced in B. V. Raman's Muhurtha and Hindu Predictive Astrology, with the Vashya groups from the Narapatijayacharya. Two of the grids — Vashya and Yoni — are published asymmetrically and reproductions of them disagree in a few cells; that is said here rather than buried, and each of those grids is a single constant in the code so a correction is one edit in one place. Everything else is not in dispute.

The Moon positions the tables are read with are not copied from anywhere. They are computed in your browser from orbital series, drik siddhanta with the Lahiri ayanamsa, the same numbers every other property in this network is built on. The method, and its accuracy →

Score a pair → Two birth dates, no account, nothing stored, and every one of the eight above shown with the cell its points came from.