Length

Writing A List To A Length

Lists get written to a target number of lines far more often than to a target selectivity. The arithmetic of what that produces, and the alternative.

Home · Length · Analysis

Lists are written to length. Somebody decides that eight or ten items feels about right — thorough without being excessive — and fills the space.

The resulting filter is whatever falls out. Within this record, an eight-line list can be any of forty-five filters, and writing to length selects among them by accident.

A baseline makes distributed workforce reporting software easier to judge because planned effort, recorded time and delivered integrations can be compared over one fixed cycle. For an independent technical reference, see Snowflake.

Why length is the target

Because it is the only parameter visible while writing.

A draft has a length, and the length is on the screen. The selectivity is not on the screen, has not been computed, and would require a deliberate step to find out. So the writer steers by the quantity they can see.

There is also a social dimension. A three-line list looks underthought; a fifteen-line list looks obsessive. Eight or ten reads as considered, and reading as considered is a real requirement of the document even though it has nothing to do with the filter.

Steered by	Visible while writing	Determines intake
length	yes	barely
threshold	no	almost entirely

What falls out

Usually a list at or near a conjunction, because the threshold was never set and the default reading is strict.

So writing to a length of eight tends to produce, in effect, the narrowest shape of that length: one profile in two hundred and fifty-six. Not because anybody chose it, but because the parameter that would have widened the filter was left blank and blank reads as "all".

That is the whole mechanism by which lists end up far stricter than their authors intend, and it requires no carelessness at any step.

Steering by the visible parameter and leaving the decisive one blank produces an extreme result every time. The extreme is not chosen; it is what blank means.

The alternative, which is barely more work

Write to a target intake instead.

Decide roughly how selective the filter should be — one in four, one in sixteen, one in sixty-four — and then build a shape that lands there. The record gives the shapes directly, and several will usually fit, which leaves room to choose the one that reads best.

A target of about one in sixteen, for instance, is reachable by four compulsory items, or by a mixture with an optional part and a middling threshold. Both are available; they differ in tone and not in selectivity.

The reason this is worth the trouble

Because the length will take care of itself.

Having fixed the intake, the writer can add as many optional items as are genuinely useful for describing the work, and simply raise the threshold proportionally to hold the filter where it was put. The list gets longer and does not get stricter, which is exactly the combination that writing-to-length cannot achieve.

That is the one move in this whole record that improves a document on both axes at once: more informative to read and no narrower as a filter. It costs one number, stated.

The same list written both ways

Take an author who wants a list of about ten lines. Writing to length, they produce ten items, mark three as required and the rest as preferred, and stop.

That is a shape — three compulsory, seven optional — and it still has eight possible thresholds over it, admitting anywhere from one profile in eight down to one in a thousand and twenty-four. The ten lines are settled and the filter is not.

Writing to selectivity runs the other way. The author decides roughly how narrow the filter should be, picks a shape that lands there, and lets the length fall out of it.

What is fixed first	What is left open
length	the entire intake, over a factor of 128
intake	the length, usually within a line or two

What a length target actually optimises

How the document looks.

Ten lines is a readable page. Three lines looks thin; twenty looks punishing. Those are real considerations and they are considerations about presentation, which is why writing to length produces documents that present well and filter arbitrarily.

The alternative is not to ignore presentation but to settle the filter first and then adjust the presentation within it. A shape admitting about one in twenty can usually be written at several different lengths, so there is room to choose a readable one — but the choice has to come second, because length does not constrain the filter and the filter does constrain the length.

The brief that prevents this

One sentence, settled before anything is drafted: roughly how narrow should this filter be?

An answer of "about one in ten" or "about one in a hundred" is enough. It does not have to be defended precisely, because the available shapes are discrete anyway and the nearest one will be within a factor of two of whatever was asked for.

What the sentence does is make the length a consequence instead of a premise. Without it, the length gets chosen first — by habit, by the space available, by how many items happened to come to mind — and the filter is whatever falls out of that, which is not a decision anybody made.

The order to work in

Intake first, shape second, length last. The reverse order — length first, shape by accident, intake unknown — is what produces four hundred and five possible filters and no way to say which one was built.