Our house rules

Why your number is different somewhere else

Numerology has no standards body. It has traditions, and the traditions disagree with each other on four specific points: which letters carry which digits, what to do with Y and W, whether certain totals are held whole instead of reduced, and when a personal year turns over. Any two sites that make different choices on any one of those will hand you different numbers for the same birth date, and both will present theirs without qualification.

That is the actual reason your number changes between calculators. Not that one of them is broken — that nobody says what they did. So this page says what we did, in full, including the places where we depart from a tradition somebody else follows.

One consequence worth stating up front: because we publish this, you can catch us. If a number on this site ever contradicts a rule on this page, that is a defect and it is visible to you, which is a deliberately uncomfortable position to put ourselves in and the whole reason for doing it.

The alphabet: Pythagorean, not Chaldean

We use the Pythagorean alphabet. Letters take values one through nine by their ordinary position — A through I take one through nine, J starts again at one, and S starts again after that. It is the western convention and the one that essentially all English-language numerology assumes.

The Chaldean system assigns different values, treats the number nine as special and leaves it out of the letter set, and consequently produces different name numbers from the same name. We do not use it and we do not mix the two. Mixing is the thing that produces a number nobody, including the person who published it, can reproduce.

Y and W

This is the rule that causes the most disagreement, because a name number splits into vowels and consonants, and Y sits on the line. Some calculators treat it as always a vowel, some as never one, and one well-known site treats both Y and W as vowels unconditionally — which makes WILLIAM a four-vowel name, and its own copy concedes that this contradicts most numerology books.

Our rule is printed below, exactly as the module that applies it states it. Beneath that, five names split letter by letter by that same module. We did not decide these splits by hand for this page; they are output.

A name with more than one word in it

Every name above is a single word, which leaves the question most people actually have unanswered: with a first, a middle and a last name, are all the letters summed once, or is each name reduced on its own and the results then added? Both methods are in print. They do not agree.

We sum every letter of the whole name and reduce that total once. Reducing each part first is the method that quietly loses things: a whole-name total of twenty-nine is an eleven, and the same name added part by part can land on a two with the eleven never having existed at all. A karmic debt disappears the same way. If a number you worked out by hand does not match ours, this is the first place to look — and because the aggregation is stated here rather than left to be inferred, you can tell in one step which of us did which.

Master numbers

Eleven, twenty-two and thirty-three are held whole at a named set of positions rather than reduced, and folded everywhere else. The exact set of positions is in the table below, so there is no ambiguity about where a master survives and where it does not.

Two things follow from that, and both are worth saying out loud. First, we admit thirty-three — some systems recognise only two masters, which means a reader whose total lands there is a question those systems structurally cannot answer. Second, we write a master compound-over-fadic, so an eleven appears as 11/2. Living an eleven as a two is not a failure; it is the ordinary way most people carry one, and printing both digits says so without a paragraph of reassurance.

Karmic debt

Four two-digit totals are read as carrying a karmic debt — thirteen, fourteen, sixteen and nineteen. What matters far more than the list is where a debt is looked for, because that is where systems silently differ: a rule that only inspects the last two-digit value before the final collapse will miss a nineteen that reduced through a ten on the way down, and the reader will simply never see a section that should have been there.

Our detection rule is printed below. A debt is written compound-over-fadic in the same way a master is, so it appears as, for example, 19/1 — the total that carried it and the digit it became, both on the page.

The personal year boundary

This is the departure that affects the most people, so it is stated plainly: we turn the personal year on your birthday, not on 1 January. Your year opens on the birthday you have just had and closes the day before your next one. The twelve turnings inside it open on your own day of the month rather than on the 1st.

Almost every free calculator online uses the January boundary. The practical difference is narrow and predictable: if you have not yet had your birthday this year, a January-boundary site gives you a number one higher than ours, and after your birthday the two agree. Because that gap is where every “your calculator is broken” message comes from, the Personal Year calculator prints the January answer too, once, beneath ours.

We prefer the birthday because it is the only boundary that is actually about the person. A year that turns on a shared calendar date is a year everybody starts together; a year that turns on your birth date is one you can locate yourself in, and it is the boundary that makes the twelve turnings mean anything at all.

The Life Path, and the leap day

A Life Path can be produced two ways: by reducing the whole birth date in one go, or by reducing the month, the day and the year separately and then summing those three. The two methods usually agree and occasionally do not — the separate method is the one that can produce a thirty-three, which the all-at-once method quietly cannot. Ours is named in the table below rather than left for you to infer.

The leap day gets its own rule too. A birth on 29 February is a real date with an unreal anniversary, and any system that counts years from a birthday has to decide what to do in the years that do not contain one. Ours decides, and says so.

The rules, verbatim

Below is the policy block itself, read out of the module that applies it — not a description of it written separately and left to drift.

What we will not do

We do not give financial, medical, legal or fertility direction from a number, in any product, at any price. We do not grade a year or a number as good or bad. We do not publish lore we would immediately have to hedge, and we do not assert a total without printing how it was reached.

Those are not modesty. They are the same commitment as the rest of this page: say what we do, do only what we said, and leave the working visible so somebody can check.

Every rule on this page, applied to your own birth date.