Today has a number

The full working, both tables at once. Vowels are marked, because the inner number is built from those alone — and because which letters count as vowels is the third thing the two schools argue about.

The whole alphabet, laid out the way the Pythagorean table builds it — nine letters to a row, three rows, the last one two letters short. That makes the column number the Pythagorean value: A sits in column 1, so A is a 1.

Take every possible string of two letters, then of three, and so on, and count how often the two tables come out with the same answer. This is worked out in full, not sampled — and it is computed here, from the two tables above, rather than looked up in a table written earlier.

1. The Pythagorean table is the one almost every modern book uses. It runs the alphabet against 1–9 three times over: A to I take 1 to 9, J to R take 1 to 9 again, S to Z take 1 to 8. It is simple, it covers the whole alphabet with no gaps, and its connection to Pythagoras is a name rather than a lineage — the arrangement is a twentieth-century one.
2. The Chaldean table assigns 1–8 only, leaving 9 out on the grounds that it is not to be given to a letter. Its values are not alphabetical: A, I, J, Q and Y all take 1; F and P take 8. The groupings are usually explained as sound-based rather than order-based, which is why a letter’s position in the alphabet tells you nothing about its value.
3. They reduce differently too. Both add digits until one is left, but the Pythagorean school stops at 11, 22 and 33 — the master numbers — and reports them unreduced. The Chaldean school cares about the compound number before reduction and treats the single digit as only the outer half of the reading.
4. And they disagree about vowels. Y is the problem. Pythagorean practice usually counts Y as a vowel only when it is doing a vowel’s job; Chaldean practice generally does not. Since the inner number is built from vowels alone, one letter’s classification can move it by several points.

1. Two answers means at most one is right. The tables give different numbers for the same name. Any page that prints one number without saying which table it used is hiding a coin flip.
2. Spelling decides the result. A middle name, an initial, a married name, a transliteration, an accent dropped at a border — each changes the total. A quantity that moves when a form is filled in differently is describing the form.
3. It only works on one alphabet. Both tables assume the Latin alphabet of twenty-six letters. Applied to a transliterated name they read the romanisation, which was someone else’s choice, not the name.
4. It is not a prediction and not advice. No number here says anything about your health, your money or your future. It is a tradition, shown with its working, and that is all it is offered as.
5. The agreement figures count letters, not names. The chart above works through every possible string of a given length, each one weighted the same. Real names are nothing like that — E and A turn up constantly, Q and Z almost never — and this page holds no list of real names, so it cannot tell you which way that pushes the percentage. Take the figure as what the two tables do to the alphabet, not as a survey of people.