What are Torah codes?

Write out the letters of the text one after another, with no spaces between the words. Choose a starting letter and a number — the skip — and read every letter that many places apart. If the letters you read spell a word, that is an equidistant letter sequence, an ELS — popularly called a Torah code.

For example: from the first ת of Genesis — the last letter of בראשית — every 50th letter reads ת, ו, ר, ה. The skip is 50; the sequence is תורה. A skip of 1 is simply the text as written, so the laboratory starts at 2 unless you ask otherwise.

A sequence can run forward, in the text's own order, or backward. A backward sequence is shown with a negative skip: its first letter comes later in the text and each next letter stands that many places earlier.

The search runs in your browser over the whole letter stream, so it is exact: every sequence in the scope and skip range you choose is found, not a sample. Every letter of every sequence is linked to its word and verse, and the matrix — the text laid out in rows of a chosen width — shows where the letters stand around it.

For every skip, a term of L letters can start at (letters in the scope) − (L − 1) × skip places; multiply that by the product of each of its letters' frequencies in the scope, add it up over every skip and direction searched, and you have the number of sequences chance alone would be expected to produce.

Beside the formula, control runs shuffle the scope's letters (keeping exactly the same letters) and search again, so you can see how many sequences a text with no order at all would give.

We search the Westminster Leningrad Codex — the oldest complete manuscript of the Hebrew Bible, copied about 1008 — as the Open Scriptures Hebrew Bible publishes it: the consonants as written (the ketiv), 304,850 letters in the Torah by our count and 1,197,042 in the whole Tanakh. The figure most commonly cited for the Torah is 304,805; manuscripts and printed editions differ by a few letters in full and defective spellings, and we call neither count the correct one.

Because a sequence depends on every letter between its first and last, a single letter's difference between editions can move or remove it. Each result here is a statement about this text.

A result is a start and a skip: a measurement. Short words made of common letters appear at some skip almost everywhere, and long words almost nowhere, which is why the count expected by chance stands beside every count found.

This laboratory measures and claims nothing: it finds the sequences, shows where every letter stands, and prints what chance alone would produce beside every count.