We computed all 729 nakshatra pairings through both of India's great matching systems. They disagree far more than anyone admits — and the pattern of disagreement is the real lesson.
India matches horoscopes in two great dialects. South of the Vindhyas: nakshatra porutham — ten (sometimes twelve) named compatibilities checked between the two birth stars, each answered mostly yes-or-no: Dina, Gana, Mahendra, Sthree Deergha, Yoni, Rasi, Rasi Athipathi, Vasya, Rajju, Vedha. North of them: guna milan — eight kootas scoring weighted points that sum to 36: Varna (1), Vashya (2), Tara (3), Yoni (4), Graha Maitri (5), Gana (6), Bhakoot (7), Nadi (8), with 18 the customary pass mark.
Marriages across that line happen every day, and every such family eventually holds two match reports computed from identical inputs. The natural questions — do the two systems mostly agree? when they disagree, which one is "right"? — have circulated for decades with answers made of anecdote. Yet the question is perfectly computable: both systems are deterministic functions of the couple's stars and Moon signs, and the whole input space is only 729 pairings. So we computed the entire space, through both engines, and counted.
For readers who know one system but not the other — the two checklists side by side, one line each:
| Porutham | What it checks | Weight in practice |
|---|---|---|
| Dina | Count from girl's star to boy's — favorable remainders promise health & prosperity | Critical tier |
| Gana | Temperament class (deva / manushya / rakshasa) compatibility | Minor tier |
| Mahendra | Boy's star at specific counts ahead — progeny & protection | Minor tier |
| Sthree Deergha | Boy's star sufficiently far ahead of girl's — the wife's welfare | Middle tier |
| Yoni | Animal-nature pairing — intimacy compatibility | Middle tier |
| Rasi | Moon-sign relationship between the two charts | Middle tier |
| Rasi Athipathi | Friendship of the two Moon-sign lords | Middle tier |
| Vasya | Mutual attraction classes of the signs | Minor tier |
| Rajju | The stars must not share a body-zone "rope" — longevity guard | Critical gatekeeper |
| Vedha | Specific star pairs that "pierce" each other — excluded outright | Critical tier |
| Koota | Points | What it checks |
|---|---|---|
| Varna | 1 | Spiritual class order between the charts |
| Vashya | 2 | Sign-to-sign influence/amenability |
| Tara | 3 | Star-count relationship (the Dina cousin) |
| Yoni | 4 | Animal-nature pairing (the Yoni cousin) |
| Graha Maitri | 5 | Friendship of Moon-sign lords (the Athipathi cousin) |
| Gana | 6 | Temperament class (the Gana cousin) |
| Bhakoot | 7 | Moon-sign distance — 6/8, 2/12, 5/9 are the dreaded doshas |
| Nadi | 8 | Constitutional class — same class is Nadi dosha, the heaviest single item |
Tables 0a/0b — The checklists. Note the inverse weighting: the South's heaviest tests (Rajju, Vedha, Dina) are yes/no gates missing from the North; the North's heaviest points (Nadi 8, Bhakoot 7) are middle- or lower-tier in southern practice.
| Measure | Ten-porutham (Vākya school) | 36-guna milan |
|---|---|---|
| Pairings that pass (≥6/12 poruthams · ≥18/36 gunas) | 656 / 729 = 90.0% | 548 / 729 = 75.2% |
| Mean score | 7.54 / 12 (62.8%) | 21.73 / 36 (60.4%) |
| Score range observed | 3 – 11 | 4 – 36 |
| Score correlation between the systems: Pearson r = 0.292 | ||
Table 1 — Both systems are generous on average — but they are generous to different couples.
That correlation figure deserves a pause. If the two systems were the same wisdom in different notation, r would approach 1. If they were unrelated lotteries, r would approach 0. Measured: 0.29 — the systems share a thin common signal (mostly the Nadi/Gana/Yoni family, as we will see) atop largely independent judgments. Knowing a couple's porutham score tells you surprisingly little about their guna score, and vice versa.
| Guna milan: pass (≥18) | Guna milan: fail (<18) | |
|---|---|---|
| Porutham: pass (≥6) | 513 pairings — 70.4% both systems bless | 143 pairings — 19.6% South blesses, North objects |
| Porutham: fail (<6) | 35 pairings — 4.8% North blesses, South objects | 38 pairings — 5.2% both systems object |
Table 2 — The verdict quadrants over all 729 star pairings, at each system's customary pass mark.
Three couples in four get a consistent answer. The other 178 pairings — 24.4% — get a blessing from one tradition and an objection from the other. Notice the asymmetry: the South-blesses/North-objects quadrant (19.6%) is four times the size of its mirror (4.8%). The porutham system, at its customary six-of -twelve bar, simply passes more of the space than guna milan does at eighteen — 90% versus 75%. A family hearing "the match fails" should therefore first ask which system failed it, because the base rates of failure differ by a factor of 2.5.
The disagreement is not noise, and it is certainly not an "error" in either system. The two systems examine overlapping but distinct axes, weight them differently, and gate on different failures.
Exactly one koota is mathematically identical on both sides of the line: Nadi. The North's 8-point Nadi and the South's nadi test both partition the 27 stars into the same three classes and object to same-class pairs. Our data confirms the identity: the 243 pairings failing northern Nadi are precisely the 243 failing southern nadi. Gana and Yoni are shared in name and spirit — both systems classify stars into temperament groups and animal natures — but the pass/fail matrices differ in detail, and the North grades on a scale where the South often rules yes/no. Tara/Dina are cousins: both count star-to-star distances, with different favored remainders. Rasi/Bhakoot and Athipathi/Graha-Maitri are similarly cousins — sign-relationship and lord-relationship tests with different exception lists.
The South's most feared gatekeeper — Rajju, the rule that the two stars must not share a "rope" (body-zone class), classically read as a longevity protection — has no counterpart anywhere in the 36 gunas. A Rajju failure can sink an otherwise-perfect southern match; the North never performs the check. Conversely the North's Varna (spiritual class) and Vashya (mutual influence by sign) have no porutham equivalents; and Bhakoot's specific dread of the 6/8, 2/12 and 5/9 sign pairs is a northern institution the South handles differently (and more forgivingly) under Rasi porutham. Two inspection regimes, overlapping on perhaps half their checklists — a 24% verdict disagreement is not a scandal; it is the arithmetic consequence of asking different questions.
Fifty-five pairings — 7.5% of the space — are strong southern matches (eight or more poruthams) that fail the northern bar. Example: girl Ashwini, boy Hasta — a comfortable southern pass at 8/12, but 10/36 in guna milan: Nadi 0 (same class), Bhakoot 0 (6/8 sign relation), Graha Maitri near 0 (Mars– Mercury enmity) stack three heavy zeros the porutham system either doesn't ask about or forgives. For a Tamil family this pair is simply a good match; a Delhi astrologer would open with two doshas.
The mirror quadrant is smaller but contains the study's most striking family: same-nakshatra pairs. Ashwini with Ashwini, Rohini with Rohini — the North scores these very highly: identical Moon signs give Bhakoot its full 7, identical lords give Graha Maitri its full 5, Tara and Vashya cooperate, and the total lands at 28/36, an excellent match (with even the same-class Nadi dosha classically cancelled for same-sign couples — a cancellation our engine applies and the texts endorse ⚖️). The South counts the same facts and returns 5/12: fail — Dina fails (distance zero), Mahendra fails, Sthree Deergha fails (both want the boy's star comfortably ahead of the girl's), and Rajju fails outright (same star, same rope). One system reads sameness as harmony; the other reads it as insufficient distance. Neither is confused — they encode different marital philosophies: guna milan prizes similarity; porutham prizes complementary offset, a wife's star placed to "lengthen the thread" of the husband's.
Exhaustive computation has one pleasure sampling never gives: you meet the extremes personally.
There is exactly one perfect 36/36 in the whole universe of star pairings: girl Rohini, boy Mrigashira. Every koota cooperates — adjacent stars, compatible ganas and yonis, Moon signs (Rishaba, pada-1 Mithuna–adjacent) in benefic relation, different nadis. And the punchline writes itself: the porutham system hands this uniquely perfect northern couple a workmanlike 8/12 — good, unremarkable, nowhere near its own maximum. Meanwhile the porutham system's own summit — 11/12, achieved by nine pairings (Ashwini–Shravana, Rohini–P.Bhadrapada, Ardra–Revati among them; no pairing achieves 12/12 in this pada-1 space) — scores between 20 and 30.5 in guna milan: solid, never spectacular. Each system's ideal couple is merely ordinary in the other's eyes. No sentence in this study summarizes the two philosophies better.
At the other end, the floor of the northern scale is girl Jyeshtha, boy Ardra at 4/36 — yet the southern engine passes the reversed Jyeshtha-girl pairing at 6/12 and gives Ardra-girl/Jyeshtha-boy a failing 5. And the single most schizophrenic pairing we found: girl Dhanishta, boy Purva Phalguni — 9/12 in porutham, 5.5/36 in guna milan. A match one tradition rates near-excellent and the other rates near-worst-possible, from identical inputs, with both engines functioning exactly as designed. Any cross-regional family that has felt whiplashed by contradictory reports may take comfort: the whiplash is structural, ancient, and shared with every couple in their quadrant.
First: do not average the scores. A 8/12 and a 10/36 do not blend into a "medium" match; they are answers to different questions, and averaging them destroys both. Second: read the shared core first. If Nadi passes (both systems), Gana and Yoni pass in both dialects, and neither Rajju (South) nor Bhakoot (North) objects, the couple stands on ground both grandmothers' traditions respect — that describes just over half of all pairings. Third: where one system objects, check that system's own forgiveness rules before anything else — the North cancels Nadi and Bhakoot doshas under specific lord-friendship conditions (56% of Nadi-dosha pairings cancel — see our companion study), and the South's own authorities treat several poruthams as minor when Rajju and Dina hold. Fourth: remember that both systems examine only the star layer. The full charts — planet strengths, dasa alignment, the Manglik layer — sit above either scorecard in every serious tradition, northern or southern.
On juno.date's match page the two verdicts are computed side by side from the same inputs — porutham in the southern engines (both Vākya and Thirukanitha schools), guna milan in the northern engine, with each system's own cancellations applied and each labeled by tradition. The disagreement rate you have just read about is visible there on live couples every day; we regard showing it as basic honesty, and hiding it behind a single merged number as the one genuinely wrong answer.
A final widening of the lens: the porutham-versus-guna divide is itself a simplification. Telugu matching practice keeps its own weighting of the southern checks; Kerala tradition adds its distinctive emphasis (and its own papa-samya sin-point comparison, which neither scorecard above contains); Kannada practice differs again. juno.date computes the Telugu, Kerala and Kannada variants alongside the Tamil and North-Indian ones on every match, because 'the traditional verdict' has always been plural — the plural is the tradition. The two-system comparison in this study is the largest single seam, but every seam in that patchwork produces its own version of Table 2, and a couple standing on any seam deserves to see both sides of it.
"The" porutham system is itself a family. Tamil practice divides between the Vākya-almanac school and the Thirukanitha school, whose porutham matrices differ enough to flip about one verdict in five between them — a disagreement we measure in a separate article. The porutham figures above use our Vākya-school engine; rerunning the whole study with the Thirukanitha engine moves the pass rate from 90.0% to 77.0% and the North–South quadrant sizes by a few points each, but leaves every qualitative conclusion — the ~25% disagreement, the asymmetry, the same-star paradox, the r≈0.3 correlation — intact. The North–South gap is robust to the South's internal argument.
The question has no empirical meaning as posed — there is no agreed outcome variable against which 729 pairings have been scored. What the data can say: the systems measure different constructs (r=0.29), so "accuracy" claims for either are claims about tradition and philosophy, not measurement. Choose the system your family's tradition speaks; consult both when two traditions marry.
It is a match the northern philosophy endorses and the southern philosophy questions — 4.8% of all pairings live in that quadrant. Families in it should look at which poruthams failed: if the failures are Rajju or Dina, the southern objection is at its most serious tier; if they are Mahendra or Vasya, southern authorities themselves rank the objection minor, and the strong northern score describes the same couple through its own valid lens.
Beyond the school differences measured here, apps differ in pada handling, in which cancellations they implement, and occasionally in the ayanamsa that assigns the stars themselves. Ask any tool two questions: which school, and which cancellation rules. A tool that cannot answer has answered.
Both are real variants. Many southern practitioners append Nadi and Varna (borrowed from the northern list) to reach twelve checks; some Kerala and Karnataka lineages count differently again, and several modern Tamil sites blend a scored version where each porutham earns fractions rather than a plain yes or no. The ten measured here are the classical Tamil core on which all the variants agree, and adding the two borrowed checks moves no verdict in our data that the shared Nadi analysis has not already flagged. Fractional-scoring blends are presentation choices layered over the same matrices — pleasant to read, identical underneath.
For six of the 27 stars, the four padas straddle two Moon signs, so a couple's rasi-dependent kootas (Bhakoot, Graha Maitri, Vashya, Rasi, Athipathi) can shift with the pada even when the stars are fixed. Our sweep pinned each star to its first pada — the convention used when a family knows stars but not full charts. Sliding those six stars to their alternate signs re-scores roughly one pairing in five by a few points but moves the quadrant boundaries only marginally: the 24% disagreement figure is robust because its causes — Rajju versus Bhakoot, similarity versus offset — do not live in the padas. With full birth details, juno.date computes from the actual Moon longitude, so the pada question resolves itself; the sweep's convention matters only for star-only matching.