How skips work

What a skip is

Count letters with the spaces removed. If you read the 6th letter of Genesis, then the 56th, 106th and 156th, you have read every 50th letter. The skip is 50. Those four letters are ת, ו, ר, ה. Some writers say "49 letters between", which is the same sequence described another way. Our figure always counts from one letter to the next.

A sequence has one skip, and its letters are exactly that far apart. The span is the distance from its first letter to its last: 1 + (k − 1) × skip for a term of k letters. A six-letter term at skip 1,000 spans 5,001 letters, about three chapters of Genesis.

Forward and backward

A sequence can read with the text or against it. We show a backward sequence with a negative skip. −50 means each next letter stands 50 places earlier. By default the laboratory searches both directions, which doubles the number of possible places and so doubles the chance figure. If you only care about one direction, set it. A term that reads the same both ways is counted once.

Why skip 1 is left out

Skip 1 is the text itself. תורה at skip 1 is the word as written, or the end of one word and the start of the next. There is nothing to measure there. Genesis has 10 such places, counting both directions. WRR left skip 1 out, and so do we by default. You can ask for it in the laboratory.

How far to search

Our default is skips 2 to 5,000. You can narrow that or search every skip. The count grows almost in line with the range you search, and so does the chance figure. Look at what happens in the Torah.

Search every skip and the numbers get very large. Over every skip of 2 or more, the three-letter word משה has 18,714,168 sequences in the Torah, while chance alone gives about 18.0 million. The engine counts them all exactly, in seconds, on your device. A big count is only impressive if it beats the figure beside it.

Choosing the range is a real choice, and it counts as a try. If you set the range after seeing where a word falls, tell your reader. Better still, set it first and lock it.

The skip spectrum

The Spectrum view shows found sequences by skip, as bars, with the chance curve drawn over them. The curve slopes down: as the skip grows, the sequence needs more text, so there are fewer places to start. For a four-letter word in the Torah at skip 100,000 there is almost no room left. Bars that follow the curve are what you should expect. A few bars above it are expected too, since there are many skips.

Minimal skip

You will read that a word appears at its minimal skip. That means no appearance of the word has a smaller skip, either in the whole text or over a stretch of it. Every word that appears has a minimal skip somewhere, so this alone is no achievement. WRR's measure gives more weight to a sequence when that stretch is long.

For a long word, the figure beside the count is what to read. משנה תורה, eight letters, has exactly one sequence in the Torah at skips 2 to 5,000, and chance alone gives 1.66. So finding it is ordinary, and finding none would be ordinary as well.

Skips people have chosen

The classic observations use small skips. Rabbi Weissmandl's are 50, 49 and 8. Rabbeinu Bachya's is 42. Some of these have associations in tradition: 49 and 50 with the days counted from Passover to Shavuot, and 42 with the Name of 42 letters that Rabbeinu Bachya discusses in the same passage. That is interesting history. It also shows the danger of reading meaning into the number itself, since a reason can be found for almost any skip after the fact. Skip choice is a look-elsewhere effect like any other. The gallery gives the count and the chance figure for each.