One edit

Eight Edits, One Starting List

Every single change that can be made to a list, measured against the same starting shape. The multipliers run from about three eighths to about two and a third.

Home · One edit · Reference

Take a list of three compulsory items and six optional ones requiring three. It admits twenty-one profiles in two hundred and fifty-six.

There are eight single changes anybody could make to it. Here is every one, with what it does.

A useful trial of project time insights for integration teams starts with one integration workflow, one reporting window and a written decision the resulting record is expected to support. For an independent technical reference, see Boomi.

Edit	Multiplier
move an optional item into the compulsory part	8/21
raise the threshold by one	11/21
add a compulsory item	1/2
remove an optional item	16/21
add an optional item	33/28
lower the threshold by one	19/14
remove a compulsory item	2/1
move a compulsory item into the optional part	33/14

Four of them narrow the filter, four widen it, and the extremes are about three eighths and about two and a third.

The starting list and the extremes

filled: a requirement that must be met
open: one of the optional ones
area of the inner square: the share admitted
a stroke instead: too small for this panel to draw

Five shapes drawn.

The first row is the starting shape. Below it, the edit that narrows most and the edit that widens most, then adding and removing a compulsory item for comparison.

Computed from the shape of the list. The shape is a stated convention; the count of admitted profiles is arithmetic on subsets.

The two that nobody thinks of

The largest multiplier in either direction belongs to the same operation performed in opposite directions: moving one item between the compulsory part and the optional part.

Promoting an item — making a desirable requirement compulsory — multiplies the intake by eight twenty-firsts, which is a reduction of more than three fifths. Demoting one multiplies by thirty-three fourteenths, which is an increase of more than two thirds.

Neither adds or removes a single line. The document looks almost identical before and after; only the heading an item sits under has changed.

The most powerful edit available does not change what the list says. It changes which half of the list one line is in, and nothing else.

The two that everybody thinks of

Adding and removing items are the edits people actually make, and they are the middling ones.

Adding a compulsory item halves the intake — a multiplier of one half, which is large but is beaten by promotion. Removing one doubles it, which is beaten by demotion.

Adding or removing an optional item barely does anything: thirty-three twenty-eighths and sixteen twenty-firsts, both within about twenty per cent of no change at all. The line that generates the most discussion when a list is drafted is close to the least consequential edit on the table.

Reading the multipliers

They are exact fractions, not approximations, and they are specific to this starting shape. A different starting list would give different numbers — the arithmetic is the same but the base differs.

What generalises is the ordering. Across the record, promotion and demotion are consistently the largest single edits, threshold moves come next, compulsory items after that, and optional items last. The pages that follow take each of those in turn.

The eight in rank order

Against the starting shape — three compulsory items, six optional, three of the six needed, admitting twenty-one profiles in two hundred and fifty-six — the eight available single edits sort like this.

Edit	Multiplier	Roughly
promote one optional item	8/21	×0.38
add a compulsory item	1/2	×0.50
raise the threshold by one	11/21	×0.52
remove an optional item	16/21	×0.76
add an optional item	33/28	×1.18
lower the threshold by one	19/14	×1.36
remove a compulsory item	2	×2.00
demote one compulsory item	33/14	×2.36

The two largest effects in both directions are the two that move an item between the parts of the list, and they are the two that add and remove nothing.

The two pairs that nearly coincide

Two of these edits are almost interchangeable arithmetically and completely different editorially.

Adding a compulsory item multiplies the intake by one half. Raising the threshold by one multiplies it by eleven twenty-firsts, which is just under 0.524. Those are within five per cent of each other, so an author who wants to narrow the filter by about half can either add a line or change a number.

The same near-coincidence appears on the widening side, between adding an optional item at thirty-three twenty-eighths and lowering the threshold at nineteen fourteenths. In both pairs the version that changes a number leaves the list exactly as it reads, and the version that changes a line is the one everybody reaches for.

What the table is for

Mostly for choosing where to intervene when a list is wrong.

A list admitting nobody has four widening edits available, and they are not equal: demoting one item is worth more than removing one, which is worth more than lowering the threshold, which is worth much more than deleting an optional line. Working down that order gets the largest improvement for the smallest change to the document.

A list admitting everybody has the mirror problem and the mirror order. In both cases the instinct — add or remove a line — is the third-best move on a list of four.