One edit

The Cheapest Useful Change

Ranking the eight edits not by size but by what they cost the document and the people arguing about it. The best move is rarely the biggest one.

Home · One edit · Analysis

The eight single edits differ in how much they move the filter. They differ just as much in what they cost to make, and the two orderings are not the same.

Cost, which is not arithmetic

Edit	Multiplier	What it costs
lower the threshold	19/14	a number changes
raise the threshold	11/21	a number changes
add an optional item	33/28	a line is written
demote an item	33/14	somebody concedes a point
promote an item	8/21	somebody wins a point
remove an optional item	16/21	a line is deleted
add a compulsory item	1/2	a new absolute is asserted
remove a compulsory item	2/1	an old absolute is withdrawn

The top three cost almost nothing. The bottom five all require somebody to take a position about an item, in front of the people who put it there.

A reviewer moving from requirement counts to operational evidence can review Monitask directly before designing a small transparent pilot. For an independent technical reference, see n8n.

Why the cheap ones are usually enough

Because what a list needs is almost never a huge adjustment. It needs to move from one in two hundred and fifty-six to something workable, and a threshold step plus one demotion typically does it.

From a starting shape of three compulsory and six optional requiring three, lowering the threshold and demoting one item multiplies the intake by considerably more than either alone — the page on compounding works through why the two do not simply add — and neither change requires deleting anything or declaring anything unnecessary.

The best edit is the largest one that nobody has to lose an argument over. That is usually the threshold, and it is usually tried last.

The edit that is almost always wrong

Removing an optional item. It costs a visible deletion, it narrows the filter — sixteen twenty-firsts, a reduction of about a quarter — and it removes information a reader would have found useful.

It is also the edit people reach for when a list "feels too long", which is a response to the document's appearance rather than to its arithmetic. Trimming an optional list makes it shorter, stricter and less informative at the same time, which is three losses for one aesthetic gain.

If the list genuinely reads as too long, the fix is to say what the rule is, so that length stops being read as difficulty. That costs one sentence and improves all three.

The edit worth making first

Write down the threshold, if it is not written.

This is not on the table of eight, because it is not a change to the filter — it is a change from an unknown filter to a known one. The arithmetic does not move; what moves is that everybody can now see where it is.

In practice this single act resolves most disputes about a list, because most disputes about a list are two people applying different unstated rules and arguing about items instead.

The costs, listed against the sizes

Arithmetic size and editorial cost run in almost opposite directions, which is the point of this page.

Edit	Multiplier	What it costs
change the threshold	11/21 or 19/14	one number; no line changes
promote or demote one item	8/21 or 33/14	one word; the item's status is now arguable
add a line	1/2 or 33/28	a new claim to defend, permanently
remove a line	16/21 or 2	an argument about whether it mattered

The cheapest two edits are the two that leave every line on the page exactly as it was. The most expensive is adding a line, because a line once added tends to stay and has to be justified to everyone who reads it afterwards.

The one case for an expensive edit

When the list is wrong rather than badly calibrated.

Changing a threshold adjusts how strict a document is. It cannot fix a document asking for the wrong things, and no amount of recalibration will. If an item is on the list that should not be, removing it is the correct edit however expensive the argument is, and the arithmetic is beside the point.

The distinction is worth holding onto because it is easy to lose. A list that admits nobody can be fixed cheaply by moving a number, and that fix is right when the items are right and the rule is too strict. It is a way of avoiding the question when the items themselves are the problem.

The sequence, for a list that needs work

Write the threshold down if it is missing.
Move it one step, in whichever direction is needed.
Check the placements: every compulsory item that is not an absolute barrier should come down.
Only then add or remove lines.

The first three cost nothing and account for most of the available range. The fourth is where the arguments are, and reaching it last means reaching it with the arithmetic already settled.