Tārā Kūṭa and Graha Maitrī: The Two Kūṭas Read From Counting
On this page, 8 sections
Of the eight kūṭas in the guṇa milan scheme, two are almost never explained properly. Tārā gets summarised as star compatibility and graha maitrī gets summarised as planetary friendship, and both summaries are so vague that they tell you nothing about what is being computed. Between them they carry eight of the thirty-six points, more than a fifth of the whole score, and both of them are simple enough to work out by hand once somebody shows you the counting.
They also share a feature that gets flattened in every report I have seen. Both are asymmetric. The score from one person to the other is not the same as the score in the opposite direction, and the classical method runs the count both ways deliberately. When you see a single number for either kūṭa, two calculations have been collapsed into one.
Where these two sit in the thirty-six
The aṣṭakūṭa scheme assigns points in a neat ascending run. Varṇa one, Vaśya two, Tārā three, Yoni four, Graha Maitrī five, Gaṇa six, Bhakūṭa seven, Nāḍī eight. That adds to thirty-six, and the arrangement is worth noticing because the point value is a statement of relative weight. Tārā at three sits at the lighter end. Graha maitrī at five sits in the middle, above Yoni and below Gaṇa.
The full scoring frame is in the piece on guṇa milan and the eight kūṭas. What follows here is the internals of these two, since they are the two you can actually verify yourself with nothing more than a nakṣatra list.
Tārā: nine steps, counted twice
Tārā kūṭa uses only the two janma nakṣatras. The method is a count and a remainder.
Count forward from the first person’s nakṣatra to the second person’s, inclusive of both ends. Divide that count by nine and keep the remainder, treating a remainder of zero as nine. Then do the same thing in the opposite direction. You now have two tārā numbers, one for each direction.
Work an example. Aśvinī is the first nakṣatra and Puṣya is the eighth. Counting from Aśvinī to Puṣya gives eight. Eight divided by nine leaves eight. Counting from Puṣya round to Aśvinī gives twenty-one, because you pass through the remaining nineteen nakṣatras before arriving. Twenty-one divided by nine leaves three.
So this pair produces an eight in one direction and a three in the other. Two different tārās from the same two nakṣatras. Anybody who quotes one number for this pair has thrown half the calculation away. The individual readings for these two are at Aśvinī and Puṣya.
The nine tārās and which three are set aside
The nine remainders have names, and the names are the whole content of the kūṭa.
- 1 Janma. The birth star itself. Treated as mixed.
- 2 Sampat. Wealth. Favourable.
- 3 Vipat. Danger. Set aside.
- 4 Kṣema. Wellbeing. Favourable.
- 5 Pratyari. Opposition. Set aside.
- 6 Sādhaka. Accomplishment. Favourable.
- 7 Vadha, also called Naidhana. Obstruction. Set aside.
- 8 Mitra. Friend. Favourable.
- 9 Parama Mitra. Best friend. Favourable.
Three, five and seven are the three the tradition avoids. The scoring is then straightforward. Both directions favourable earns the full three points, one direction favourable earns one and a half, neither earns nothing. Our Aśvinī and Puṣya pair lands on eight in one direction and three in the other, so it earns one and a half.
Change one input and the result flips. Revatī is the twenty-seventh. Counting from Revatī to Aśvinī gives two, which is Sampat. Counting from Aśvinī round to Revatī gives twenty-seven, and twenty-seven divided by nine leaves zero, which is read as nine, Parama Mitra. Both favourable, so the full three points. Two adjacent nakṣatras produced completely different answers against the same partner, which is a useful demonstration that this kūṭa is genuinely a count rather than a temperament match.
The same cycle you use to pick a day
Here is the thing about tārā that I almost never see written down. This nine-fold cycle is not specific to marriage matching. It is the same instrument used to judge a day.
Tārā bala in muhūrta works by counting from a person’s own janma nakṣatra to the nakṣatra the Moon is transiting today, taking the remainder over nine, and reading the same nine names. A day whose count lands on Sampat, Kṣema, Sādhaka, Mitra or Parama Mitra is a supportive day for that person. A day landing on Vipat, Pratyari or Vadha is one the tradition suggests you do not begin things on.
Same cycle, same names, same three exclusions. The only difference is the second input. In matching it is a partner’s birth star. In muhūrta it is today’s Moon. Realising this changed how I thought about the kūṭa, because it means tārā was never a marriage-specific concept. It is a general relational counter that the matching scheme borrowed, which explains why it carries only three points. It is answering a smaller question than the kūṭas around it.
Graha maitrī: five points from two lords
Graha maitrī is a different kind of test. It ignores the nakṣatras and looks at the two janma rāśis, the Moon signs, and specifically at the planets that rule them.
Find the lord of each person’s Moon sign. Then consult the naisargika maitrī table, the natural friendship scheme, and read the relationship in both directions. The usual scoring runs: mutual friends or the same lord for both people scores five, friend against neutral scores four, neutral against neutral scores three, friend against enemy scores one, neutral against enemy scores half a point, and mutual enemies score nothing. That version is the one reproduced most often, but its status is worth being honest about. The friendship table underneath it, the seven planets’ friends, neutrals and enemies set out below, is solid and consistent across every compilation checked for this piece, all tracing to Bṛhat Parāśara Horā Śāstra’s chapter on natural planetary friendship. The point-scoring conversion is a separate layer on top of that table. Digests that publish it attribute it to general Aṣṭakūṭa scoring convention rather than to one numbered BPHS verse, and they do not agree below the top and bottom rows: some run a flat five-tier scale with no half-point row at all, others split the lower end the way this piece does. Treat five points for mutual friends and zero for mutual enemies as settled, and the exact shape of the middle tiers as convention rather than a fixed classical number.
One precision point that trips people up. This kūṭa uses the Moon sign lord, which means you need the actual Moon sign, and several nakṣatras straddle a sign boundary. Kṛttikā’s first pada is in Meṣa and its remaining three are in Vṛṣabha, so two people both born in Kṛttikā can have different rāśi lords and therefore different graha maitrī scores against the same partner. A tool that derives the rāśi from the nakṣatra alone will get those cases wrong.
The friendship table, and the asymmetry inside it
The natural friendship scheme is where graha maitrī gets interesting, because it is not a symmetric relation. Two planets can disagree about each other.
- Sūrya counts Chandra, Maṅgala and Guru as friends, Budha as neutral, Śukra and Śani as enemies.
- Chandra counts Sūrya and Budha as friends and everyone else as neutral, with no natural enemies at all.
- Maṅgala counts Sūrya, Chandra and Guru as friends, Śukra and Śani as neutral, Budha as an enemy.
- Budha counts Sūrya and Śukra as friends, Maṅgala, Guru and Śani as neutral, and Chandra as an enemy.
- Guru counts Sūrya, Chandra and Maṅgala as friends, Śani as neutral, Budha and Śukra as enemies.
- Śukra counts Budha and Śani as friends, Maṅgala and Guru as neutral, Sūrya and Chandra as enemies.
- Śani counts Budha and Śukra as friends, Guru as neutral, and Sūrya, Chandra and Maṅgala as enemies.
Look at Chandra and Budha. Chandra counts Budha a friend. Budha counts Chandra an enemy. That is the sharpest asymmetry in the table, and it is not a copying slip: it is the naisargika maitrī scheme as printed in Bṛhat Parāśara Horā Śāstra’s chapter on natural planetary friendship, and the asymmetry is confirmed independently across sources that otherwise disagree on how to convert the table into points. It is also exactly why graha maitrī has a friend-against-enemy row in its scoring at all. Without asymmetric relations that row would never be reached.
Take our earlier pair. Aśvinī sits in Meṣa, ruled by Maṅgala. Puṣya sits in Karka, ruled by Chandra. Maṅgala counts Chandra a friend, Chandra counts Maṅgala neutral, so friend against neutral, four of five points. Against Revatī in Mīna, ruled by Guru, Maṅgala and Guru count each other friends in both directions, so the full five.
The misreadings worth correcting
The first is the collapse of the two directions into one number. Both of these kūṭas were built to be run twice and scored on the pair of results. A single reported figure has already done the averaging for you and hidden which direction was the weak one, which is the only genuinely actionable part of either calculation.
The second is the reading of Vadha as a prediction of harm. The name is severe and it gets translated in ways that frighten people. What the classical usage attaches to it is a recommendation about beginnings, that this is not a count on which to start something. It describes a quality of relation between two positions in a cycle, and it says nothing about outcomes for a person. In a matching context it costs one and a half points out of thirty-six. That is the actual weight the tradition gave it.
The third is treating graha maitrī as a personality match. It is a relation between two rulers, evaluated on a fixed table that has nothing to do with either person’s temperament. The temperament kūṭas are elsewhere in the scheme, in gaṇa and in yoni, and conflating them with this one produces readings that sound psychological while resting on a lookup table.
What these two are and are not telling you
Eight points out of thirty-six is a real share and it is also a minority share. Both of these kūṭas run on two inputs each, a pair of nakṣatras or a pair of Moon signs, and two inputs cannot describe two people. They are screens, computed quickly, designed to be one layer of a longer process that in classical practice also included the full charts of both people.
What I like about them is that they are honest about being arithmetic. There is a count, a remainder, a table and a score, and anybody can check the working. Everything a system states this plainly can be argued with, and that is a better position to be in than receiving a verdict with no method attached.
Why does tārā get counted in both directions?
Because the count is directional. Going from the first nakṣatra to the second is a different number of steps than going back, so the remainder over nine differs too. The classical method scores the pair of results, giving three points if both are favourable, one and a half if one is, and nothing if neither is.
Which tārā numbers are considered unfavourable?
Three, five and seven, which are Vipat, Pratyari and Vadha. One, Janma, is treated as mixed. The remaining five are favourable. A remainder of zero is read as nine, Parama Mitra, which is favourable.
Is tārā bala for choosing a day the same calculation?
It is the same nine-fold cycle with a different second input. In muhūrta you count from your own janma nakṣatra to the nakṣatra the Moon occupies today and read the same nine names. That shared machinery is the clearest sign that tārā is a general relational counter rather than a marriage-specific idea.
Does graha maitrī use the Sun sign or the Moon sign?
The Moon sign, always. The kūṭa scores the relationship between the lords of the two janma rāśis, and janma rāśi in this system means the sign the Moon occupied at birth. Any calculation that starts from a Sun sign is answering a different question.
How can two planets disagree about being friends?
The naisargika scheme is genuinely asymmetric. Chandra counts Budha a friend while Budha counts Chandra an enemy, and there are milder asymmetries elsewhere in the table. This is why the graha maitrī scoring has separate rows for mutual friendship, friend against neutral and friend against enemy.
Can a good score on these two make up for a poor total?
Between them they carry eight of thirty-six points, so they can move a score meaningfully without deciding it. More to the point, a total is a screening number and was never meant to stand in for reading two charts. Treating any composite score as the answer is the mistake the arithmetic makes easy.