Kundali Matching Calculator

Kundali matching hands back a number out of 36, and almost everyone reads it as a mark out of 36 — as though 30 were an A and 17 a fail. It is not that kind of number. I ran the eight-koota system over every pairing it can possibly produce, all 46,656 of them, and the conventional pass mark of 18 is cleared by 70.76%. Seven couples in ten pass. A test that almost everyone passes is not telling the family very much when it says yes.

What it is telling you sits underneath the total, in two all-or-nothing kootas and a separate Mars check that most calculators bury in a footnote. That is why the tool above prints those three first and the score second. Below is what each part is actually worth.

All 36 Points Come From One Number Each

Start with the thing nobody says out loud. Every one of the eight kootas reads the position of the Moon and nothing else. Varna, Vashya, Graha Maitri and Bhakoot come off the rashi; Tara, Yoni, Gana and Nadi come off the nakshatra; and both the rashi and the nakshatra are just two ways of slicing the same sidereal degree. Your Sun does not appear. Neither does Jupiter, your Ascendant, your birth city, or the other eight grahas that a full Vedic birth chart would put in front of an astrologer.

Two consequences follow, and both are checkable. First, any two people born when the Moon stood at the same degree get an identical 36-point verdict against the same partner, no matter what else differs. Second, the entire system is finite: at half-pada resolution there are 216 distinct Moon positions, so 46,656 possible matches and only 845 distinct eight-koota score vectors in existence. Guna Milan is a lookup table. A large one, an old one, a carefully built one — but a table.

Seven Couples in Ten Already Clear 18

Here is the distribution nobody publishes alongside the score. Across all 46,656 pairings the mean total is 20.90 and the median is 21. The commonest single score is 26, which turns up for one pairing in nineteen. And the cut-offs land like this:

• 18 or better: 70.76% of all pairings — the conventional pass
• 24 or better: 36.57% — the "very good" band
• 28 or better: 12.02%
• 32 or better: 1.63%

I checked that against reality rather than trusting the flat enumeration, because a grid of Moon positions assumes people are born evenly around the zodiac. Running 40,000 couples from real birth dates spread over 1930 to 2029 gives a mean of 20.88 and a pass rate of 70.14% — within a whisker of the theoretical figures. The distribution is genuine, not an artefact.

So a 22 is unremarkable. A 26 is genuinely above average. And an 18 or a 19, the numbers that produce the most anxious phone calls, sit in the bottom third of everything the system can produce while still technically passing.

Your 26 Out of 36 Can Still Be Refused

This is the part that catches families off guard, and it is not the astrologer being difficult. Two kootas are all-or-nothing. Nadi is worth 8 points and pays nothing at all if both partners share a nadi. Bhakoot is worth 7 and pays nothing if the two Moon signs sit at one of three condemned distances. Between them that is 15 of the 36 points decided by two switches — and traditional practice treats either switch being off as a reason to stop, regardless of what the remaining 21 points did.

Measured over every pairing, 56.01% of all matches that clear 18 carry a Nadi or a Bhakoot dosha.More than half of every "pass" the system issues comes with a blocker attached. That is why a couple can be handed 26 out of 36 by a website in the morning and told no by a pandit in the afternoon, with nobody having made an arithmetic error.

The relationship runs the other way too, and this is the cleanest evidence that the total and the doshas are not two independent opinions. Only 0.11% of all pairings score below 18 without carrying at least one of the two doshas. Put plainly: failing the score almost always means you failed a dosha. The number is largely a re-statement of the switches.

Split every pairing by which switches are off and the four groups look nothing alike:

Read the first row again. If both switches are on, you clear the pass mark 99.64% of the time. The score is barely doing any work.

Bhakoot Dosha Catches Exactly Half of Everyone

Bhakoot compares the two Moon signs by counting from one to the other in both directions. Seven of the twelve possible relationships are fine. Three — 2/12, 5/9 and 6/8 — void the koota outright. Because those three cover six of the twelve possible distances, Bhakoot dosha appears in exactly 50.00% of all pairings. Not approximately. The arithmetic cannot produce any other figure, and the real-birth-date sample lands on 50.59%.

Nadi is the same story with a different fraction. The 27 nakshatras split into nine Adi, nine Madhya and nine Antya, so two people share a nadi 33.33% of the time — measured at 33.34% on real dates. One couple in three carries the heaviest dosha in the system by pure counting.

None of that makes the doshas meaningless. It does mean they are common conditions rather than rare warnings, and the classical literature knows it: both carry well-established cancellations. Nadi dosha is traditionally waived when the pair share a nakshatra but sit in different padas, among other cases. Bhakoot dosha is waived when the two rashi lords are the same planet or mutual friends — which the Graha Maitri row above already tells you. A tool that reports a dosha without checking whether its own output cancels it is doing half a job.

Enter the Same Two People the Other Way Round

Try it in the calculator. For most couples the total changes.

Two kootas read the two slots differently. Varna awards its single point only when the groom's rashi rank is not below the bride's, so whenever the two ranks differ — 75% of pairings — the point exists in one direction and vanishes in the other. Vashya's published table is asymmetric too, though it bites less often, at 7.64%. Together they move the total for 74.44% of all pairings, by a mean of 1.07 points and a maximum of 2.5.

That is usually a curiosity. Occasionally it is not: for 4.08% of couples, about one in twenty-five, the swap carries the total across the 18-point line. Same two people, same two Moons, pass one way and fail the other. If a matrimonial site has ever given you a different score from the one your family was quoted, check the slots before you check the ephemeris.

Has Anybody Ever Actually Scored 36?

Short answer: with the tables this page uses, no. Not rarely — never. Brute-forcing all 46,656 cells returns a maximum of 35, reached by 92 of them, and in every single one either Gana or Yoni falls one point short.

The reason is that the kootas are not independent, and they cannot be, because they all read the same two numbers. Yoni pays full marks only when both partners share a yoni animal. Sharing an animal means being one of a fixed set of nakshatra pairs — and once you also require different nadis, so that the 8-point koota pays, the surviving pairs turn out to be hemmed in on gana or on Tara. Maxing one koota quietly constrains another.

Now the honest caveat, and it matters more than the finding. Sources genuinely differ on the Gana table. Under the symmetric version implemented here, 36 is unreachable. Under the two common variants, it becomes reachable by 32 and 64 cells respectively — that is 0.07% and 0.14% of all pairings, roughly one couple in 700 at the most generous reading. So the truthful statement is not "36 is impossible" but something sharper: whether a perfect score exists at all depends on which table your calculator happens to use. A number that changes its own ceiling depending on the source you picked is a convention, not a measurement. Treat it accordingly.

The Manglik Test Flags Half the Population

Mangal dosha sits outside the 36 points entirely and frequently outweighs them in a family discussion. The standard rule is that Mars occupying the 1st, 2nd, 4th, 7th, 8th or 12th house makes a person Manglik. That is six houses out of twelve — and running 30,000 real births through the ephemeris, counted in whole signs from the Ascendant, gives 49.67%. Half of everyone.

It gets stranger if you follow the common practice of checking from three reference points. Counted from the Moon the rate is 50.50%; from Venus, 53.47%. Flag anyone who fails any of the three and you have condemned 88.67% of the human race. Only 13.46% are Manglik on all three. Which reference point a practitioner uses is not a detail — it is most of the answer.

Two things save the rule from being useless. First, the classical cancellation: when both partners are Manglik, the dosha is traditionally held to neutralise. Drawing couples at random from the Lagna rule, 24.75% are both Manglik and 25.59% are neither, so half of all couples are already clear before anything else is considered. The flag genuinely bites only the 49.66% where exactly one partner carries it. Second, Mars in the 8th from the Ascendant and Mars in the 2nd are not remotely the same condition, and any reading that collapses them into one word is throwing away the information.

One Real Report, Read Line by Line

The calculator opens on a real pair of charts, chosen because it fires every part of the machine at once. A groom born 2 January 1990 at 09:15 in Delhi, a bride born 4 February 1992 at 16:30 in Mumbai. His Moon lands at 18°26′ Kumbha — Shatabhisha, pada 4. Hers lands at 28°23′ Makara — Dhanishta, pada 2. The total is 24.5 out of 36, a 65th-percentile result that comfortably clears the pass mark. The kind of figure a matrimonial site prints in green.

Now read the rows.

• Bhakoot: 0 of 7. Kumbha and Makara are adjacent signs, so they sit 2/12 apart — one of the three condemned distances. Seven points gone in a single line, and the report is now carrying a blocker.
• Nadi: 8 of 8. Adi against Madhya. The heaviest koota on the board is clean, which is the best thing in the report and the reason this match would survive a first look.
• Varna: 0 of 1.Kumbha ranks Shudra, Makara ranks Vaishya, so the groom's rank sits below the bride's and the point is withheld. Swap the two entries and it comes back — which is exactly why the swapped total reads 25.5.
• Vashya: 1.5 of 2.Her Moon sits at 28°23′ of Makara, past the 15-degree line where that sign changes vashya group, so she is scored as Jalachara rather than Chatushpada. This is the only koota in the system that can turn on the degree instead of the sign.
• Yoni: 1 of 4. Horse against Lion — not the same animal, not a classical enemy pair either. Low, but not a zero, and worth three-quarters of a point less than most people assume from the label.
• Gana: 6 of 6. Both Rakshasa. Worth pausing on, because the popular reading of Rakshasa gana is uniformly negative and yet two of them together take full marks. The koota scores the match, not the person.
• Manglik: bride only, on all three reference points. Her Mars sits in the 7th from her Ascendant, the 12th from her Moon and the 1st from her Venus. His is clear on all three. That unanimity is unusual — only 13.46% of people are Manglik by every reference — and it removes the usual escape route of arguing about method.

Then the part that should end the conversation about the number. Hold his birth data fixed and slide hers across her own birth day, and the total runs from 19 to 33. Her Moon crosses from Makara into Kumbha late that evening; the moment it does, Bhakoot flips from 2/12 to 1/1 and seven points reappear at once. Same two people, same date, a verdict that travels from "scraped through" to "exceptional" on the hour of birth alone.

So the honest summary of this 24.5 is: clean on the heaviest koota, void on the second heaviest, one-sided Manglik, a point of the total riding on which name went in which box, and fourteen points riding on a birth time. "24.5 out of 36, very good" does not survive contact with the detail.

Using a Number That Cannot Make the Decision

A practical order of operations, in the order the tool prints them. Check Nadi first, since it is the heaviest and the most commonly enforced. Check Bhakoot second, and if it is void, look immediately at Graha Maitri — same or friendly rashi lords is the classical cancellation and it is sitting right there in the same report. Check the Manglik status of both charts, not one. Only then read the total, and read it as a percentile rather than a percentage.

When not to lean on this at all: if either birth time is uncertain by more than a couple of hours, the Moon may have changed nakshatra during the window and half the kootas change with it — the tool's birth-day panel shows you exactly how far the total can slide. Nail the birth time down first. Guna Milan also says nothing whatsoever about the Sun, the seventh house, Venus, or Jupiter's condition, all of which a working astrologer weighs for marriage; and it is silent on everything a Western synastry reading of aspects and orbs would raise. Eight kootas from two Moon degrees is a screening tool. It was never designed to be the verdict.

Used that way it is genuinely good at what it does. The system has survived a very long time because reading two people's temperaments off the Moon — the graha the tradition assigns to the mind — is a defensible thing to want to do, and because the 27 nakshatras are a genuinely fine-grained instrument. Just do not ask a lookup table to carry a decision it was never built to make.