The heaviest verdict in guna milan strikes exactly a third of humanity — and the same classical texts that raise it cancel it more than half the time. We computed every case.
Guna milan awards points across eight tests, and its scale is deliberately top-heavy: the last test, Nadi, is worth 8 of the 36 points — more than Varna, Vashya and Tara combined. But its real weight exceeds even that number. In classical practice an uncancelled nadi dosha is not an eight-point deduction; it is a gate. Reports mark it in red; families stop reading at the word. The juno.date engine honors that tradition explicitly — a live nadi dosha caps the verdict at "needs review" no matter how handsome the total — because pretending the gate is a subtraction would misrepresent how the system is actually used.
A rule with that much power over real weddings deserves total transparency about two questions: how often does it fire? and how often does tradition itself disarm it? Both questions turn out to have exact answers, because nadi depends only on the two birth stars — a space of 729 pairings that a computer can examine exhaustively. We did, and both answers are below. Neither has, to our knowledge, been published with its working before.
The word nāḍī means channel or pulse — in the Ayurvedic frame, the constitutional current of the body. The 27 nakshatras are dealt into three classes in a fixed repeating pattern (1-2-3-3-2-1, folding back and forth through the star list):
| Nadi | Constitution | Nakshatras (9 each) |
|---|---|---|
| Ādi (first) | Vāta — wind | Ashwini, Ardra, Punarvasu, Uttara Phalguni, Hasta, Jyeshtha, Mula, Shatabhisha, Purva Bhadrapada |
| Madhya (middle) | Pitta — bile | Bharani, Mrigashira, Pushya, Purva Phalguni, Chitra, Anuradha, Purva Ashadha, Dhanishta, Uttara Bhadrapada |
| Antya (last) | Kapha — phlegm | Krittika, Rohini, Ashlesha, Magha, Swati, Vishakha, Uttara Ashadha, Shravana, Revati |
Table 1 — The three nadi classes. The classical concern: a couple sharing a class shares a constitutional excess, read as a risk to health and progeny.
The dosha rule could not be simpler: same class, dosha; different class, full 8 points. There are no intermediate scores. And the frequency follows immediately: with nine stars per class, the 729 ordered pairings contain 3 × 9 × 9 = 243 same-class cases — exactly one third. Our engine sweep confirms the count to the pairing (and finds zero half-point cases, unlike the graded kootas). One couple in three, for eternity, by construction — the dosha's frequency contains no information about any particular couple. It cannot be rare, because it was never designed to be rare.
The rule's severity makes sense only through its medical cosmology. In the Ayurvedic frame every constitution balances the three doṣas — vāta, pitta, kapha — and each nadi class marks a nakshatra's inclination toward one of them. The classical worry about a same-nadi union is compounding: two vāta-dominant constitutions, the reasoning goes, reinforce rather than balance each other, and the texts attach the concern specifically to the couple's health and to the constitution of children — which is why Nadi carries the maximum points in a system whose stated purpose is progeny and long companionship, and why the gate outranks the total. One may hold any opinion about the medicine; the design intent is coherent, and it explains features the point-count alone cannot: why the dosha is all-or-nothing (a constitutional match either compounds or it doesn't), and why the cancellations all describe ways the rest of the charts re-establish balance — different signs, shared signs, allied lords. The system is not superstition bolted onto arithmetic; it is a consistent theory whose empirical premises belong to its era ⚖️. We compute it faithfully and present it with its own logic visible, so families can weigh it within their own tradition's frame.
Folklore holds that Ādi-nadi (vāta) doshas are the most dangerous, and some practitioners refuse those matches while tolerating Madhya cases ⚖️. The sweep can at least test the structural version of the claim: are Ādi doshas less likely to cancel? No. The 243 dosha cases split perfectly evenly — 81 per class, another consequence of the nine-nine-nine deal — and the survivors split 34 (Ādi), 32 (Madhya), 40 (Antya): cancellation rates of 58%, 60% and 51%, differences well within what the sign-lord geometry explains. No class is structurally doomed; whatever weight a family gives the severity folklore, the cancellation machinery treats all three channels alike.
The classical literature never left the dosha absolute. The standard treatments attach parihāra — cancellation — conditions under which the same texts that define the dosha declare it void ⚖️ (the exact lists vary somewhat across sources and regional practice; we implement the three most widely attested). The three rules, and what each contributes across all 243 dosha cases:
| Cancellation rule | Classical logic | Cases cancelled (of 243) |
|---|---|---|
| Same nakshatra, different rasi | Shared star but different Moon signs — the padas separate the couple's constitutions | 0 in this star-level sweep (requires pada detail; see §5) |
| Different nakshatra, same rasi | Shared Moon sign with distinct stars — the sign's unity overrides the class clash | 6 |
| Moon-sign lords same or friends | The two charts' ruling planets are allies — their friendship carries the couple past the constitutional clash | 131 |
| Total cancelled | 137 / 243 = 56.4% | |
| Surviving (live) doshas | 106 / 243 — 14.5% of all pairings | |
Table 2 — The classical cancellations applied to every dosha case, as implemented in the juno.date engine.
The headline: more than half of all nadi doshas cancel — overwhelmingly through the lord-friendship rule, the least-known of the three among laypeople and the one most reports never check. A report that prints "Nadi: 0/8" without testing the lords has done a third of its job: out of every nine couples frightened by that line, five are, by the tradition's own authorities, not in dosha at all.
What of the cases no rule cancels? Here the sweep produced this study's most striking regularity: every one of the 106 surviving nadi doshas occurs in a pairing whose total score is already below 24/36 — not a single "excellent match ruined only by nadi" exists in the whole space. The reason is structural, and rather beautiful: the nadi classes are interleaved with the zodiac in a way that correlates same-class pairs with hostile sign geometries, so pairings that fail nadi and lack the friendly lords that would cancel it have usually already lost points at Bhakoot and Graha Maitri for the same underlying reason. The dreaded scenario — a perfect couple destroyed by one hidden channel — turns out to be mathematically impossible in the star-level system. Where nadi survives, the match was already visibly strained; where the match is strong, nadi either passed or cancels. The dosha never ambushes; it corroborates.
This finding also explains a piece of practitioner folklore: experienced astrologers often say they "rarely see a good match blocked by nadi alone." The sweep shows why — it is not luck or leniency; the configuration does not exist.
It is worth dwelling on what kind of fact this is. It was not designed in by any single author — the koota matrices and the nadi deal evolved separately, in different centuries — yet they interlock so that the system's harshest gate can never contradict its own totals. Whether one reads that as the wisdom of long selection (rules that embarrassed practitioners by blocking obviously fine couples would not have survived transmission) or as a happy accident of modular arithmetic, the practical consequence is the same and is checkable by anyone with the tables: a star-level report showing 24+ points alongside a live nadi dosha contains an error (full-chart reports, where padas shift the sign-based kootas, can in principle strain the boundary — but must still test the cancellations before printing the dosha). We know of no other publicly stated invariant of the guna-milan system of this kind, and we offer it to practitioners as a free consistency test for the software they use.
The strangest corner of nadi arithmetic is the couple born under the same nakshatra. Same star means same class — automatic dosha — yet these are also the cases tradition treats most gently. The first cancellation rule exists precisely for them: same star with different Moon signs (possible for the six stars whose padas straddle a sign boundary) is classically void. Same star, same sign has a subtler classical treatment: several authorities void the dosha when the couple's padas differ, and some star lists exempt specific nakshatras (Rohini, Revati, Pushya among the commonly named) from same-star dosha entirely ⚖️. Our star-level sweep — which pins each star to its first pada — necessarily shows these couples as dosha cases cancelled by the lords rule (same sign, same lord, hence "friends"); with full birth details the engine resolves the actual padas and signs, and the cancellation is decided by the chart, not the convention. The practical takeaway for same-star couples: your case is the single most nuanced in the whole system, the alarmist reading is the least supported one, and a tool (or astrologer) that does not examine your padas has not yet examined your case.
Cross the Vindhyas and nadi's throne is occupied by a different tyrant. The Tamil ten-porutham system does check nadi-style class compatibility, and our sweep confirms its 243 failing pairs are exactly the northern dosha set — the two traditions share this one test perfectly (the only test they share exactly, as our porutham-vs-guna study maps). But the southern system neither weights it heaviest nor gates on it: in Tamil practice the fearsome gatekeeper is Rajju, and nadi sits mid-list. Kerala practice adds its own distinctive check (papa-samya, the balancing of malefic burdens) above either. A couple flagged for nadi in a Delhi report may find a Chennai report barely mentions it — not because either is careless, but because each tradition chose a different guardian for the same gate of longevity and health. Families spanning that line should read both reports' gatekeepers, not just both totals.
Step 1 — confirm the classes. Table 1 is complete; check both stars yourself. (One report in our review series showed a nadi verdict from a mis-entered star — thirty seconds of checking would have caught it.)
Step 2 — test the lords. Find both Moon signs and their lords, and consult any friendship table (ours is in the match report): same lord or friends → the dosha cancels by the most productive classical rule, the one that clears 131 of 243 cases.
Step 3 — check the sign/star escapes. Different stars sharing a Moon sign: cancelled. Same star with different signs: cancelled. Same star, same sign: examine the padas (Step 4 territory) and the exempt-star lists.
Step 4 — look at the whole match. If your total is 24+ and nadi shows live, our sweep says one of three things happened: a cancellation applies that the tool ignored, a star was mis-entered, or the tool computes nonstandard matrices. In the entire 729-pairing space there is no legitimate high-scoring match with a live nadi dosha — so a report showing one is describing its own limitations, not your marriage.
On juno.date's match page, every nadi verdict is computed with the cancellations applied and displayed — dosha present, cancellation rule (if any), and what survives, in that order. The same transparency covers Bhakoot, whose dosha strikes an even larger fraction (50.2% of pairings — the 6/8, 2/12 and 5/9 geometries between them cover half of all sign pairs) and cancels by lord-friendship in 72.7% of its cases. The heaviest doshas in the system are also its most frequently cancelled — a proportionality the tradition built in, and reports owe their readers.
Agreement between reports usually means only that both checked the class table — the thirty-second step. The load-bearing questions are the ones after it: did either report test the lord friendship (the rule that clears 54% of all cancellations' worth of cases)? Did either examine your padas if you share a star? Does your total sit above or below the 24-point line that, in the entire space of possibilities, no surviving dosha crosses? Two reports that skipped the same steps are one opinion, printed twice.
Now nothing — and this case deserves saying plainly, because it produces real anguish. A third of all couples, including happily married ones across every generation, carry the same-class configuration; most of those carry a cancellation they never knew they had; and a marriage's own years of evidence outrank a star table in every tradition's practical hierarchy. The dosha framework exists to counsel prospective matches, not to indict standing ones; no classical authority instructs a married couple to retro-fear their stars ⚖️.
Remedial practice is a genuine part of living tradition, and between a family and its purohit it is not ours to police. What we can offer is sequence: the cancellations are free and textual — check them first (Steps 1–3 above); more than half of frightened families need nothing further. A remedy purchased before the parihāra check is a solution bought before reading the diagnosis. juno.date sells no remedies, no gemstones and no rituals, so our order of operations has no thumb on the scale.
Because honesty requires it. Everything here is computed on the 729 star pairings with each star at its first pada — the resolution at which most families first meet the dosha ("her star is Hasta, his is Mula"). Full charts refine two things: the pada-dependent cancellations (§5) and the exact Moon signs for the six boundary-straddling stars. The dosha set itself cannot change — it is fixed by the stars alone — but the cancellations get re-decided on the real signs and padas: for most couples the full chart opens escapes the coarse level cannot see (the pada rules exist only there), while for a minority it can withdraw a sign-based cancellation the pada-1 convention had granted. Either way the full chart's answer supersedes the star-level one, which is why juno.date always recomputes from complete birth details when a family has them.
Nadi dosha is real classical doctrine, and we compute it faithfully, gate and all. But the complete classical picture — the one the sweep quantifies — is this: a condition that strikes one couple in three by arithmetic; cancels in more than half its cases by the tradition's own rules; never once, in the entire space of possibilities, spoils an otherwise-excellent match; and reserves its gravest readings for couples whose reports already counsel reflection on other grounds. That is not a curse; it is a well-engineered caution system with proportionate brakes. If a single red line on a report has been weighing on your family, run the four steps above — the odds are genuinely, computably good that the weight lifts.