The Founder Archives are Dr. Armaan Sharma's own papers, written for Astro River from 2008 and restored here in his words, with modern tools used for structure and editing.
Abstract
This paper sets out what the Founder Archives mean by statistical work in Vedic astrology. The term names a record kept beside the practice — a dated claim, an interpretive basis, a time window, and a later note of outcome — and a set of codes by which such records may be counted. It does not name laboratory science, and it makes no claim of proof. The paper first describes the order in which a chart is read before any prediction is made, so that a recorded claim can be traced to the step it came from. It then defines the unit of record, the fields of an entry, separate codes for occurrence and for timing, and the only tallies the method treats as meaningful. A section on Ashtakavarga explains the numeric layer on which ambiguous questions are settled. The paper closes with the limits of any consultative record — selection, incomplete follow-up, remedy, and birth-time quality — and with the paper's own relation to Parashari and Krishnamurti method. Further sections treat negative claims, the other chart genres, repeat consultations, remedies, ayanamsa and window width, and birth-time quality. Two worked examples — a hit and a miss — a note on sources, and a glossary follow.
What I mean by "statistical"
I do not mean laboratory science. I do not mean trials, control groups, or a machine that turns a horoscope into a probability.
I mean something older and plainer: keeping a book. Writing down what was said, when it was said, what window was given, and what later happened — then counting from that book by rules fixed before it was opened.
This is a clerk's discipline placed beside an interpreter's art. The interpreter reads the chart by the rules of Parashara — houses, lords, dignity, aspects, yogas, dasha — and, where I use them, by the sub-lords of Krishnamurti and the bindus of Ashtakavarga. The clerk writes down what the interpreter said and checks it later. The two must never be confused. An interpretation that is never written down cannot be checked. A count that pretends to be more than a count becomes a lie with a decimal point.
Why it matters
Every astrologer remembers the hits. Memory is generous that way. A sitting that produced one clear success, one failure and one unclear result becomes, years later, "a case I got right" or "a case I got wrong" depending on how the story is told. The book removes that choice. The book does not remember; it records.
There is a second reason. Writing a claim so that it can be checked forces the claim to be clear. "A difficult period ahead" cannot be checked. "A change of employer during the Mercury antardasha, between these two dates" can. The discipline of the book improves the speech of the astrologer, and clear speech is most of what a client needs.
How a chart is read before any claim is made
A prediction is the last thing that happens in a reading, not the first. Before a single claim is written, the chart has to be understood as a whole. This is the order in which I do it — the Astro River reading order, as the Institute now calls it — and a claim in the book should be able to point to the step it came from.
The feel of the chart. Open it and take it in. After enough charts, a horoscope tells you at a glance whether it is a fortunate one or a hard one. This is not a rule; it is the sediment of experience, and it sets the tone for everything after.
The ascendant. What is this person about. The lagna, its lord, and the lord's placement and condition.
The struggle and the purpose. Where the Moon sits, where Saturn sits, and where Ketu sits. Between them they show what this life will be asked to carry and what it is here to learn. The client has not asked about this. It is what tells me how to guide them correctly when I answer what they did ask.
Strength, and whether strength is welcome. Where Capricorn falls. Wherever the sign of Capricorn sits in a chart, that area of life — and the people who belong to it — carries an element of struggle, lessons and challenge, whatever else the chart says about it. It is the first place I look for where the life will have to work hardest. Then: which planets are strong — and whether a strong planet is a friend to this chart or a strong ruler of the wrong houses. Then the weak planets, and what their weakness costs. Shadbala, degrees, and the aspects between planets: who touches whom, and how hard.
The houses, in order of what a life needs. The fourth first — sukha sthana, the seat of comfort and peace. Then the seventh, the partner. Then the fifth, love and children. Then the rest. Even when a client has asked only about career, this is still the order, because career is lived by a person and the person has to be understood first. It is why the order holds however short the question is: a person has to be seen before any part of their life can be read.
Where the real astrology begins. The nakshatras in which the planets sit, and their lords. The sub-lords, in Krishnamurti work. This is the level at which two charts that look alike on the surface come apart.
The divisional charts, read beside the birth chart, never on their own.
The dasha in force, and within it the antardasha in which the client is asking. Then the transits — Saturn and the Moon above all, the others in their place.
Ashtakavarga. The back kitchen. The questions the chart leaves ambiguous are settled here. It has its own section below.
Only now is a claim made, and only now can it be written in a form the book will accept.
Ashtakavarga
Ashtakavarga is a whole way of doing Jyotish, not an add-on. Ashta is eight; varga is a division or group. Instead of relying only on interpretation, the system assigns benefic points — bindus, dots — to the twelve signs, based on the positions of eight reference points: the seven classical planets and the lagna. Rahu and Ketu are excluded. It is numeric.
Each of the seven planets has its own table, the Bhinna Ashtakavarga, in which it contributes 0 to 8 points to each sign according to where the other planets and the lagna stand relative to it. Adding the seven tables together gives the Sarvashtakavarga, the master grid, whose points across the whole chart always total 337.
The total in a sign — and so in the house it occupies — shows that house's general capacity to give good results. Above 28, and especially past 30, a house is strong: planets transiting through it tend to bring success and ease. From 25 to 28 is average, and results are mixed. Below 25, and especially under 22, the house is sensitive: the areas of life it rules are where struggle, delay and challenge tend to gather, and where the practitioner looks for adjustment or remedy.
Two uses follow. Ashtakavarga In transit work, a planet crossing a sign with high bindus gives its good results; crossing a sign with low bindus, it creates friction, whatever its natural character. In assessing a life, the grid shows at once which areas — livelihood, partnership, wealth, peace of mind — are inherently robust and which will always ask for extra effort. Where the birth chart is ambiguous on a question, the grid usually is not. That is why it comes last in the reading, and why it settles more than it is given credit for.
The unit is the claim, not the person
A consultation is not a unit. A person is not a unit. The unit is a single stated claim: one native, one chart, one statement, one time window.
One sitting may produce five claims — marriage, a job, a parent's health, property, a court matter. Each goes in the book on its own line. They are never rolled together into "a successful case", because rolling them together hides which rule of interpretation was being tested.
What goes on the line
Each entry has these parts. Without them it may be a teaching note; it is not a countable entry.
- An identifier for the claim, not for the person. Names stay in a separate place.
- Chart type. Natal, prashna, varshaphala, muhurta, or transit over natal. Each has its own rules; the book keeps them apart.
- Birth data as used, and its quality — stated by the native, from a record, or rectified — and whether the time was treated as exact, approximate or unknown.
- Date the claim was made. A reading of an event already known is labelled post hoc and never enters the forward count.
- The claim, in the words used. Not a later tidied version that fits better.
- The interpretive basis — the classical factors the claim rests on, named. The house and its lord; the karaka; the dasha and antardasha lords and their connection to the house; the sub-lord of the cusp where KP is used; the bindus of the house and of the transiting planet; the transit of Saturn and Jupiter to the house and the Moon's position for the trigger; any yoga that bears on the matter and whether it is live in this period. A claim that cannot name its basis is merely a guess in Sanskrit clothing. It may be logged, marked uncited, but it never enters a count that claims to test method.
- Outcome domain, from a fixed short list: marriage or partnership; livelihood; residence or travel; health; litigation; birth or death in the family; property; education; other, named.
- The window. Start and end dates, or a named dasha segment with its calendar bounds. "Someday" is not a window.
- Follow-up. When the outcome was checked and by what source — the native, a document, a third party with standing.
- Codes (below).
- A note on anything unclear.
Prashna, varshaphala and muhurta
The fields above are written for the natal chart. Three other genres appear in the book and each needs a word.
Prashna. A chart cast for the moment of the question answers that question and nothing else. Its windows are short — weeks or months, set by the prashna itself — and the claim text should carry the question as asked, because the outcome domain is whatever the querent meant. A prashna entry is never pooled with natal entries in a tally; the rules of judgment are different and so is the selection: people ask a prashna when something is already in motion.
Varshaphala. The annual chart gives claims windowed to the year, or to a Mudda dasha segment inside it. Such entries are counted on their own, cut by the year of the chart. A varshaphala claim that agrees with a natal dasha claim is one claim with two bases, not two claims.
Muhurta. An election is not a prediction; it is a choice. The astrologer does not say what will happen, he says when to begin so that what happens is good. The book therefore records a muhurta entry differently: the matter begun, the time chosen and the chart of that time, the factors the choice rested on, and a later note on how the matter fared over a stated period. The codes are used, but O means "the undertaking begun at this time reached the outcome for which the time was chosen" — the marriage held, the business survived its first years, the house was lived in — and the window is the period over which that is judged. Muhurta entries are never mixed with natal entries. They are also the entries most exposed to the remedy problem below, since choosing a time is itself an intervention.
The codes
Two questions, scored separately. Did it happen? Did it happen when it was said?
Outcome
- O — occurred. The event happened in a form a second reader would recognise from the claim text. A civil marriage where a ritual one was implied still counts. A transfer where a resignation was said still counts, if the domain and the action match.
- N — did not occur. The window closed, follow-up was adequate, the thing did not happen.
- A — ambiguous. Only a stretch makes it a hit or a miss. A long live-in partnership where marriage was claimed. Sources disagree. A is a real code, not a parking place for failures.
- I — incomplete. Not yet checked, or the window still open. I is a status, not a result. Silence from the native is I, never N.
Timing — scored only after O, and sometimes A
- H — hit. Inside the window, including the ends.
- M — miss. Clearly outside it. A marriage said for the Venus antardasha that came two antardashas later is O on outcome and M on timing. Both facts stay visible.
- P — partial. Began inside, completed outside; or the claim named a range of segments and the event fell on a boundary it did not resolve. Used sparingly. Never a courtesy upgrade of M.
A near miss is still M. Nearness may be noted as a signed day-count at the side; it does not change the code. If a second reader, given only the claim and the follow-up, would not agree on O versus N, the code is A, and the disagreement is written down.
Negative claims
Half of honest counsel is telling a person what will not happen. "No marriage in this antardasha." "No transfer this year; the change comes in the next period." "This illness does not become surgical." These are claims like any other and they go in the book.
They are coded by the same two questions, read the right way round. A negative claim is O when the non-event held — the window closed and the thing did not happen. It is N when the thing did happen inside the window. Timing is scored as for any O: a negative claim that held for the whole window is H; one that held for most of it and then failed near the end is N, not P, because the claim was about the whole window.
Two cautions. A negative claim must be as specific as a positive one — "nothing much this year" is not a claim. And a negative claim is only worth recording when the positive outcome was live: telling a widower of eighty that he will not remarry this year is not a prediction, and the book does not dignify it with a code.
What may be counted
Only these:
- entries opened in a period;
- entries that reached O, N or A;
- entries still I;
- O/N/A against H/M/P, for entries that have both;
- the same table cut by one domain, one chart type, or one kind of dasha claim.
No single "accuracy percentage" is required by this method, and none is published. If a ratio is ever formed, its denominator must be said in words: "O among terminal outcomes in the marriage domain", "H among O entries that had a window". Different denominators answer different questions. Mixing them gives a number that means nothing.
Small numbers are the normal state of an honest book. A worker who logs only clear, forward-dated claims and waits for windows to close will have a modest count for a long time. That is why tables of counts are better than percentages: one changed code moves a percentage a long way and a count by only one.
A worked example
The entry below is illustrative. It is not a real client; the chart data are invented for teaching and should not be read as anyone's horoscope. Its purpose is to show what a complete line in the book looks like.
Entry A-0417 Chart type Natal, with transit overlay Birth data Time from hospital record, treated as exact; place and date as stated Date of claim 14 February Claim, as spoken "Change of employer, not a promotion within the same firm, during the Mercury antardasha of the Saturn mahadasha — that is, between 3 June this year and 11 March next year. Most likely in the second half of that window." Interpretive basis Tenth lord Mercury, placed in the sixth, running as antardasha lord under Saturn, which occupies the tenth. Mercury in Jyeshta, whose lord Mercury itself sits in the sixth — service, but service elsewhere. KP: tenth-cusp sub-lord Saturn, signifying the sixth and the tenth, not the eleventh — a change of job rather than gain in the same job. Ashtakavarga: sixth house 31 bindus, tenth house 24; Saturn's transit enters the sixth in August, crossing a strong sign. Moon transit over the natal sixth lord in the same month for the trigger. Domain Livelihood Window 3 June – 11 March Follow-up Native's dated note, 2 December; offer letter seen Outcome code O — native joined a different firm on 9 November Timing code H — inside the window, in the second half as stated Note The native was also offered a promotion in the same firm in July and refused it. The claim said "not a promotion within the same firm"; the refusal is consistent with the claim but is not itself an event coded here.
Three things to notice. The claim names its basis before the outcome is known. The window is a pair of dates, not a mood. And the note records a nearby fact without letting it become a second hit.
If the native had instead been promoted in July and stayed, the entry would read O on the domain (a livelihood change did occur) but A on the claim as worded, with a note that "change of employer" was not met. The temptation to call that a hit is exactly what the codes exist to resist.
A second example — a miss
The book must hold misses as plainly as hits, so here is one, equally invented.
Entry A-0422 Chart type Natal Birth data Time stated by the native as "around seven in the morning"; treated as approximate, ±30 minutes Date of claim 2 March Claim, as spoken "Marriage during the Venus antardasha of the Rahu mahadasha — between 20 August this year and 20 August three years on. Most likely in the first eighteen months." Interpretive basis Seventh lord Venus, running as antardasha lord. Venus in Rohini, whose lord the Moon occupies the seventh. Rahu, the mahadasha lord, in the eleventh, aspected by Jupiter. KP: seventh-cusp sub-lord Moon, signifying the second, seventh and eleventh. Ashtakavarga: seventh house 29 bindus. Jupiter transits the seventh in the second year of the window. Domain Marriage Window 20 August – 20 August, three years Follow-up Native's dated note, seven weeks after the window closed; family confirmed Outcome code N — no marriage in the window Timing code not scored Note An engagement was announced in the final month of the window and the marriage took place four months after it closed. The claim said marriage, not engagement. Under the rules this is N with a note, not A: a second reader given "marriage" and the dates would not call it fulfilled. The birth time was approximate; with a thirty-minute error the seventh cusp moves enough to change the sub-lord, which is recorded as a possible cause but not used to excuse the code.
This entry is worth more to a student than the first. It shows the code refusing a near miss, the note preserving the near miss for teaching, and the data-quality field doing its job — pointing at a likely cause without being allowed to rewrite the result. Had the book allowed "engagement counts", or "four months is close enough", or "the time was probably wrong so it doesn't count", the file would drift toward flattery one entry at a time.
Remedies
Astro River prescribes remedies — mantra, puja, gemstone, the standard Vedic upaya of the Parashari tradition first and as the main line. Lal Kitab's small daily corrections are used only now and then, sprinkled in for quick relief, never as the primary method, because they carry a real risk of backfiring. Either way, a remedy is an intervention, and so the book must reckon with them. A remedy is an intervention. Once one has been given, the later fact can no longer be laid at the door of the prediction alone.
Every entry therefore carries a flag: remedy present or none. The remedy, when given, is named in the note, with its date. Entries with a remedy present are tabulated separately from those without, in every count. This does not decide whether the remedy worked; the book cannot say that and does not try. It only prevents the file from quietly crediting the astrologer's foresight with what may have been the client's discipline, or crediting the remedy with what the chart had promised anyway.
One more rule. A remedy given after an outcome is known — to steady a person after a loss — is pastoral work, not an intervention on the claim, and is not flagged.
Ayanamsa and the width of the window
Two things that are easy to leave unsaid, and that a careful reader will ask about.
Ayanamsa. Every degree in this paper, and in every chart behind the Archives, is sidereal by the Lahiri ayanamsa. A chart cast by another ayanamsa can move a planet across a sign boundary, a nakshatra boundary, or a sub-lord boundary, and a claim made from one cannot be checked against the other. The book records the ayanamsa on every entry. This paper does not argue for Lahiri over its rivals; it only refuses to count across the difference.
Width. A claim windowed to a whole mahadasha may be twelve years wide; one windowed to an antardasha, a few months; one windowed to a pratyantardasha, a few weeks. A hit inside a wide window is a weaker thing than a hit inside a narrow one, and an honest file should be able to show the difference. Every entry therefore records the width of its window in days, and a tally may be cut by width. No rule here converts width into a score. The point is only that "hit" must not hide whether the astrologer said "this decade" or "this fortnight".
What the count can and cannot say
It can say: of the claims made by these rules, in this file, these many occurred and these many occurred on time.
It cannot say: planets cause events. It cannot say: Vedic astrology has been proved or disproved. Everyone in the book came to the astrologer — selected by worry, by belief, by money, by geography, by reputation. Those who never consult are not in the book. Those who stopped replying are I, and they may be the unhappy ones. Remedies are often suggested, and after a remedy the book can record a sequence but never a cause. Birth times are of mixed quality, and a bad clock is not a bad doctrine. The book carries all of this in its notes. It removes none of it.
So the book warrants one modest sentence, and no more: here is how a practitioner may keep a record of claims and outcomes without deceiving himself about what it contains.
Repeat consultations
A native who returns is not a new native. By the second sitting the astrologer knows the first outcome, and by the fifth he knows the life. This is good for counsel and bad for counting.
The book handles it by lineage. Every entry for a returning native carries the identifiers of the earlier entries for that native, so that a reader can see the chain. Claims made after an outcome is known about a closely related matter — a second marriage claim after the first was N, a career claim after the native has explained why the last job ended — are marked informed. They remain countable, but a tally may be cut to exclude them, and any tally that claims to test method should be.
The reverse case matters too. A native who returns to report an outcome is the best follow-up source there is, and the note should record that the report was volunteered rather than asked for. Volunteered reports skew toward happy news; the note lets a later reader weigh that.
Birth-time quality
Vedic judgment is sensitive to time. The lagna moves a degree every four minutes; a cusp sub-lord in KP can change with a shift of a few minutes; the balance of the birth nakshatra that sets the whole Vimshottari sequence depends on the Moon's degree, which moves roughly one degree every two hours. A wrong clock is not a wrong doctrine, and a count that mixes the two learns nothing.
The book therefore records the quality of every birth time and keeps three strata apart:
Exact. A documented time — hospital record, a written note made at the birth — used as given.
Approximate. A stated time with a known or estimated spread: "around seven", "before sunrise", "in the evening". The spread is written down. Where the spread crosses a lagna boundary, a nakshatra boundary for the Moon, or a sub-lord boundary for the cusp in question, the note says so before the claim is made.
Rectified. A time arrived at by working back from known life events. Rectification is legitimate practice, but a rectified chart has already been fitted to the life, and a claim made from it is partly a claim about the rectification. Such entries are marked and counted on their own.
Any tally may be cut by stratum, and should be. If exact-time entries score well and approximate ones poorly in the same domain, the book has said something useful about the clock. If they score alike, it has said something useful about the rule. Either is worth knowing; a pooled figure would hide both.
One further rule: a birth time is never revised after an outcome is known in order to make an entry fit. If a time is corrected — by a document coming to light, by a later rectification — the old entry keeps its old code and a new entry is opened for the new chart.
How this sits with Parashara and KP
Nothing in the book changes a single rule of interpretation. The houses still mean what they mean. Dasha still sequences the life. The book asks a narrower question after the interpretation is done: what later attached to it?
A high count in some domain does not rewrite a sutra. A low count shows which of one's own applications were too vague to score, and which rule one leans on that deserves a harder look. Disagreements between authorities — on dasha systems, on divisional charts, on the weight of the sub-lord against the sign lord — are not settled by one man's file. The file shows how these rules behaved in this book. That is all, and it is enough to teach from.
Two practical consequences. An entry with no interpretive basis is uncited and stays out of any tally that claims to test method. An untimed claim — "this chart indicates a second marriage", with no window — goes in a separate tally of fact-claims and is never mixed into the timing table.
Closing
Vedic statistical astrology, as I use the phrase, is a record kept beside a practice. The practice reads the chart in the order set out above — the whole before the part, the person before the question, the nakshatra and the sub-lord where the real work begins, and the bindus where the ambiguities are settled. The record writes down the claim, the window and the later fact, and codes them so that a second reader could do the same.
Confuse the two and you get either a mystified science or an art no one can question. Keep them apart and you get a smaller, plainer, more useful book than any legend of cases. That is the standard I set for these Archives, and the standard by which I ask to be read.
Pages that hang from this paper
This paper is a hub. The pages below are its spokes, each a short reference article that links back here. Where a spoke is not yet written, the link waits.
- How a chart is read: the reading order, step by step, as a page of its own.
- Ashtakavarga: the tables of bindus, the Sarvashtakavarga, and its use in transit.
- Prashna, muhurta and varshaphala — one page each on the genre and its rules.
- Birth-time quality and rectification.
- Remedies: what Astro River prescribes and why.
- Negative claims, and the discipline of saying what will not happen.
- The sub-lord in KP.
- Papers B, C and D of this series.
Sources
The interpretive side of this paper stands on the received tradition and claims no novelty. The house and lordship rules, dasha sequence, and the yogas are those of the Brihat Parashara Hora Shastra as taught in the north Indian line. Ashtakavarga is set out in the same text and in Varahamihira's Brihat Jataka; the tables of bindus used here are the standard ones, and the 337 total and the thresholds around 25 and 28 are conventional. The sub-lord method is that of K. S. Krishnamurti, as given in the Krishnamurti Padhdhati readers, especially the readers on horary and on marriage and profession. The place of prashna and muhurta as separate genres with their own rules is taken from the Prashna Marga and the muhurta literature respectively.
The recording side stands on no shastra. It descends from ordinary scholarly habit — the dated note, the frozen statement, the later inspection, the refusal to count what cannot be inspected. The same habit appears in case registers, field diaries and clinical logs, without any astrological content. This paper treats it as transferable clerical technique, not as evidence that astrology has entered the experimental sciences.
No modern journal, no institutional dataset and no accuracy figure is cited, because the paper advances none.
Glossary
Antardasha — sub-period within a mahadasha. Ashtakavarga — the system of benefic points assigned to signs from eight reference points; see the section above. Bhinna Ashtakavarga (BAV) — a single planet's table of points. Bindu — a benefic point in Ashtakavarga. Dasha — a planetary period; the Vimshottari sequence unless stated otherwise. Karaka — the natural significator of a matter (Venus for marriage, Jupiter for children, and so on). Lagna — the ascendant. Mahadasha — a major planetary period. Mudda dasha — the dasha sequence of the annual chart. Muhurta — electional astrology: choosing a time. Nakshatra — one of the twenty-seven lunar mansions. Pratyantardasha — the sub-period within an antardasha. Prashna — horary astrology: a chart cast for the moment of the question. Rectification — arriving at a birth time by working back from known life events. Sarvashtakavarga (SAV) — the sum of the seven planetary tables. Shadbala — the six-fold measure of planetary strength. Sub-lord — in Krishnamurti Paddhati, the lord of the sub-division of a nakshatra in which a cusp or planet falls. Sukha sthana — the fourth house, the seat of comfort. Upaya — remedy. Varshaphala — the annual chart.
Later papers in this series apply these principles to particular questions — dasha systems (Paper B, Dasha Systems), the nakshatras (Paper C, Nakshatras), and Jaimini practice (Paper D, Jaimini). They inherit the limits stated here.
