Astro Clock AI
ExplainersExplainers · 88 of 137

Is 18 gun milan good or bad?

Where the 18-of-36 line comes from, which eight kootas make up a Guna Milan score, why two matches at 18 can share almost nothing, and how to read your own koota breakdown instead of the total.

Article text is currently English only. The calculation below and the rest of the interface follow your chosen language.

In short

Eighteen of 36 is neither good nor bad on its own. It is a family convention, exactly half the maximum, and no book in this app's library sets it. In this engine every score of exactly 18 has a zero on Nadi or Bhakoot, and the same 18 can come from 18 different koota splits, so the rows tell you more than the total.

The short answer: 18 out of 36 is a middle score, and whether it is "good" depends on which rows produced it. Many families treat 18 as the line a match must clear. That line is half of 36 and nothing more; no text in this app's library sets it, and this app's kundli matching calculator does not use it. What the number hides is more useful than the number: every pairing that scores exactly 18 in this engine has lost all of its Nadi points, all of its Bhakoot points, or both.

This page does not repeat how each koota is scored; that is set out in nakshatra matching and in Guna Milan. It looks only at the 18 line and at how to read a score near it.

Where the 18-of-36 line comes from

Guna Milan, also called Ashtakoot Milan, adds up eight scores (each one a koota) to a maximum of 36. A search of every astrology text in this app's library (BPHS, Saravali, the Bhrigu Sutras, Jyotisha Siddhanta Sara, Advanced Predictive Astrology Vol. II and Daivajna Vallabha) finds no 36-point scheme, no list of eight kootas and no 18-point line. The only "Koota" in them is Saravali's Koota Yoga, one of the Nabhasa yogas, which has nothing to do with matching.

The engine's own header says the same thing in plain words: no source book in the library contains the Ashtakoot rules, so the owner adopted published references and took a rule only where independent references agreed. That decision is recorded on the sources page and the methodology page. So when someone says 18 is the minimum, the honest reply is that it is a convention, half of the maximum, and this page cannot point you to a sloka behind it.

This app draws no pass line. The calculator puts a total into one of four bands, cut at 75%, 55% and 35% of 36, and refuses no match:

Band in the calculatorShare of 36TotalsPairings in the sweep
Strong match75% and above27 to 361,980 (17.0%)
Workable match55% to under 75%20 to 26.55,285 (45.3%)
Mixed - proceed with awareness35% to under 55%13 to 19.53,357 (28.8%)
Challenging matchunder 35%4 to 12.51,042 (8.9%)
the compatibility calculation bands; totals move in half points. The sweep is every ordered pair of the 108 nakshatra quarters (11,664 pairings), run 2026-11-18.

In these bands 18 sits in "Mixed", two points below the start of "Workable". The sweep also shows why a line at 18 is a weak filter: 8,472 of the 11,664 pairings, 72.6%, score 18 or more. A rule that passes nearly three pairings in four tells you little about the one in front of you. The median total is 21.5, so 18 is a little below the middle of what this engine produces.

The eight kootas behind the score

The eight rows are not equal. They are worth 1, 2, 3, 4, 5, 6, 7 and 8 points, so the two heaviest, Bhakoot (7) and Nadi (8), carry 15 of the 36 points between them. Each row reads one of two inputs: the Moon sign or the Moon nakshatra of each person.

KootaMaximumReadsZero in the sweep
Varna1Moon sign37.5%
Vashya2Moon sign0.0%
Tara3Moon nakshatra0.0%
Yoni4Moon nakshatra7.1%
Graha Maitri5Moon sign (its lord)5.6%
Gana6Moon nakshatra22.2%
Bhakoot7Moon sign50.0%
Nadi8Moon nakshatra33.3%
ashtakoot() maxima and a table in the app inputs; zero rates over 11,664 pairings. Vashya and Tara never fell to zero in the sweep (their lowest scores there were 0.5 and 1.5).

Two rows decide most of the story. Bhakoot is all or nothing: 7 points, or 0 when the two Moon signs are 2nd and 12th, 5th and 9th, or 6th and 8th from each other, which happens in half of all pairings. Nadi is all or nothing too: 8 points, or 0 when both Moons fall in nakshatras of the same nadi, which happens in a third of them. Only 11.7% of pairings come through with no zero in any row.

Why every 18 hides a zero

Put the two heavy rows side by side and the 18 line looks different. In the sweep, 3,644 pairings keep both Nadi and Bhakoot in full, and the lowest total any of them reaches is 19.5. So a pairing with both of those rows intact can never score 18 or less in this engine. The reverse also holds: 1,700 pairings lose both, and the highest total any of them reaches is 21.

At exactly 18 there are 349 pairings. In 209 of them Nadi is 0; in 276 Bhakoot is 0; in 136 both are 0; and in none are both kept. Below 18, all 3,192 pairings lose at least one of the two. A score of 18 is therefore not "half good": it is a match where one or both of the two heaviest rows scored nothing, and the six smaller rows made up the difference.

Three matches at 18 that share almost nothing

The 349 pairings at exactly 18 are spread over 18 different koota splits. Three of them, taken from the sweep, show how far apart they can be. Each person is given by the Moon's sign, nakshatra and quarter (pada), so any two births whose Moons fall in these quarters give the same rows.

Koota (max)A: Rohini 1 + Vishakha 4B: Ashwini 1 + Vishakha 4C: Bharani 1 + Purva Phalguni 1
Moon signsTaurus + ScorpioAries + ScorpioAries + Leo
Varna (1)111
Vashya (2)0.50.51
Tara (3)1.51.53
Yoni (4)222
Graha Maitri (5)355
Gana (6)306
Bhakoot (7)700
Nadi (8)080
Total181818
ashtakoot() run 2026-11-18 at the quarter midpoints, first person first.
  • Match A keeps Bhakoot (Taurus and Scorpio are 7th from each other) and loses Nadi: both Moons are in the same nadi. Its smaller rows are middling. The Taurus and Scorpio pair is covered in Taurus and Scorpio Moon sign compatibility.
  • Match B keeps Nadi and loses Bhakoot (Aries and Scorpio are 6th and 8th) and also Gana, so three rows worth 13 points between them give nothing. Graha Maitri is full, because Mars rules both Moon signs. See Aries and Scorpio Moon sign compatibility.
  • Match C loses both Nadi and Bhakoot (Aries and Leo are 5th and 9th) and reaches 18 only because Tara, Graha Maitri and Gana are all full. See Aries and Leo Moon sign compatibility.

Anyone asking "is 18 good?" about these three is asking three different questions. A family that weighs Nadi most heavily would read A and C one way and B another; a family that weighs Bhakoot most heavily would group them differently again. The total cannot make that choice for you.

Why 18 can turn into 17

Varna is the only row in this engine that depends on who is entered first: it gives its one point when the first person's class number is equal to or higher than the second's. Matches A and B above score 18 one way and 17 the other; match C scores 18 both ways. Of the 349 pairings at exactly 18, 107 drop to 17 when the order is reversed. If a family's decision turns on whether a match is 17 or 18, it is turning on this one row and on the order the names were typed.

The Moon itself can also move a score. A nakshatra quarter is only 3 degrees 20 minutes wide, and four of the eight rows read the nakshatra while the other four read the sign. When a birth time is missing, the free match tool checks whether either Moon changed sign or nakshatra during the birth day and names the rows, and the points, that ride on it. Why two apps can give the same birth two different nakshatras is covered in why apps disagree on your nakshatra.

How to read your own breakdown instead of the number

  1. Confirm both Moons first. Run the Moon sign calculator and the nakshatra calculator for each person with an exact birth time, and note the pada. If either Moon is near the edge of a quarter, the rows that read it may change.
  2. Run the kundli matching calculator and read all eight rows, not the headline. Write down which rows are 0 and which are full.
  3. Look at Nadi and Bhakoot first. If both are full, the total is at least 19.5 in this engine. If one or both are 0, you now know where the missing points went.
  4. Then read the six smaller rows. Two matches with the same zero in a heavy row can still differ by several points in Graha Maitri, Gana and Tara.
  5. Run the match in both orders. If the total changes by one point, that is Varna, and nothing else about the match has changed.
  6. Ask what the family is actually weighing. Guna Milan reads only the two Moons; it does not look at the rest of either chart. Whether matching is needed at all, and what it leaves out, is the subject of is kundli matching necessary.

What the score cannot tell you

These are the limits no calculator can remove, this one included.

  • It cannot tell you where the 18 line came from. The library holds no source for it, and this page will not invent one.
  • It cannot settle the rules the references disagree on. For Gana and two cells of Graha Maitri the engine keeps its earlier scores because the published references give different tables; another app using another table can give the same couple a different total.
  • It reads two points in two charts: each Moon. Everything else in both birth charts, and everything about the two people themselves, is outside the score.
  • It is not a forecast. This app does not predict whether a marriage will last or how it will go, from 18 or from any other number.
Is 18 gun milan good or bad?
Neither on its own. It is half of 36 and sits in the calculator's "Mixed" band. In this engine every score of exactly 18 has a zero on Nadi, Bhakoot or both, so the useful question is which rows scored nothing.
Who decided that 18 is the minimum?
No book in this app's library sets it. None of the astrology texts here describes the 36-point scheme at all; the engine's rules come from published references adopted where they agree. The 18 line is a family convention.
Is 17 much worse than 18?
Not necessarily. Of the 349 pairings at exactly 18, 107 become 17 just by entering the two people in the other order, because Varna, a one-point row, depends on the order. Read the rows rather than the single point.
Can a match above 18 still have Nadi or Bhakoot at zero?
Yes. Of the 8,472 pairings at 18 or more, 1,427 have Nadi at 0 and 3,607 have Bhakoot at 0. A pairing can lose both and still reach 21.
Does this app reject matches under 18?
No. It has four bands, cut at 75%, 55% and 35% of 36, and refuses no match. It names the weakest row in the summary so you can look at it.
Why does my score differ from another app's?
Two common reasons: the other app may use a different Gana or Graha Maitri table, or it may place one Moon in a different nakshatra. Check both Moons in the nakshatra calculator first, then compare the eight rows one by one.
Now on yourself

Kundli Matching (Guna Milan)

The classical Ashtakoot score out of 36, computed from two birth dates.

This one has its own page

This one runs on its own page, built for it, where the full result is laid out properly. No account is needed there either.

Run it on its own page

Sources for this page