One edit

The Same Item Either Way: A Factor Of Six

One item, moved up or moved down, spans a factor of ninety-nine sixteenths — a little over six. No other single decision in this record carries as much.

Home · One edit · Reference

Promoting one item multiplies the intake by eight twenty-firsts. Demoting one multiplies it by thirty-three fourteenths.

The ratio between those two is ninety-nine sixteenths — a little over six.

For a workflow with several system handoffs, it is useful to check this implementation resource and test one normal project rather than a scripted demonstration. For an independent technical reference, see Tray.ai.

So a single line, placed in one half of a list or the other, spans a factor of six in what the whole document admits. It is the widest span any one decision commands in this record.

One item, three positions

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

Three shapes drawn.

The middle row is the starting list. Above it, one item promoted to compulsory; below it, one item demoted to optional. The same line, in two different halves of the same document.

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

The three positions, side by side

Shape	Admits
4 compulsory, 5 optional, 3 required	1/32
3 compulsory, 6 optional, 3 required	21/256
2 compulsory, 7 optional, 3 required	99/512

Nine items in every row. The same nine items, in fact — only one has moved between the two parts, twice.

The top row admits one profile in thirty-two. The bottom admits ninety-nine in five hundred and twelve, which is a little under one in five. Between them, a factor of a little over six.

Why this is the decision that matters

Because it is made constantly, informally, and without being recognised as a decision at all.

Every item on every list sits in one half or the other, and for most items the placement was never deliberated. It was inherited from a previous version, or it followed the item above it, or somebody wrote the list in one sitting and grouped things by feel.

Each of those placements is worth a factor of six. A list of nine items contains nine such decisions, and a handful of them being wrong in the same direction accounts for most of the distance between a filter that works and one that admits nobody.

Nine items, nine placements, each worth a factor of six. The compounding is why two lists naming the same things can differ by orders of magnitude.

The question that settles a placement

Only one, and it has to be answered about the work rather than about the ideal candidate:

Could somebody who lacks this do the job?

A flat no puts the item in the compulsory part. Anything else — "not as well", "it would take longer", "we would have to train them" — puts it in the optional part, where the threshold decides how many such gaps are tolerable.

That is the whole method. It takes a few seconds per item and it is the highest-value few seconds available anywhere in this record.

Why the answer drifts upward

Because the question is usually asked in the abstract, with no candidate in front of anybody, and in the abstract everything sounds necessary.

Asked concretely — about a specific person who is strong in eight areas and missing this one — the answer changes for most items on most lists. The abstract framing is what loads the compulsory part, and the compulsory part is where the factors of six accumulate.

Six, and why it is not two

The instinct is that moving an item between required and preferred is worth a factor of two, because one compulsory item halves the intake.

That instinct accounts for only half the move. Going from compulsory to optional does remove a halving, but it also adds an item to the optional part, and the optional part is sitting under a threshold that the new item helps to satisfy. Both effects push the same way.

So the move is worth more than a halving in each direction, and the two directions compound. Ninety-nine sixteenths — a little over six — is what separates the item's two placements, and nothing on the page distinguishes them except one word.

Placement of the one item	Admits (out of 512)
compulsory	16
optional	99

The one line that prevents the drift

A sentence next to the list recording what the threshold is and when it was last set.

The drift described above happens because placements change one at a time, each time for a locally good reason, with no record of what the filter admitted before. Nobody is tracking the cumulative effect because there is nothing to track it against.

A single line — "three of the following six; set at the last review" — gives the next person a baseline. They can recompute, see that the shape has moved, and decide whether that was intended. Without it, every edit starts from an unknown position, and a factor of six accumulates in steps that each look like a word change.

What to do with the figure

Count the compulsory items on a list and ask the question of each. Every item that moves down is worth about two thirds more intake; every item that should not have moved up in the first place was costing three fifths.

For a list that is admitting nobody, that audit usually finds two or three items, and two or three factors of six is the difference between a document nobody passes and one that works.