What is met

The Fraction One Reader Gets

How much of a set a reader reached is a simple fraction and the only quantity here that behaves as intuition expects. What is built on it does not.

at most 30240 arrangements

Home · What is met · Reference

Divide how much a reader met by how much was published and the result is the share of the set they reached. One of ten is a tenth. Three of ten is three tenths. Five of ten is a half.

This is the only quantity in the record that behaves proportionally, and it is worth noting precisely because everything derived from it does not.

One tenth, three tenths, five tenths

dashed: a statement a reader can check from outside
solid: published, but nothing to check it against
short and pale: never reached by this reader

Three shapes drawn; the left-to-right order is the drawing's own and carries no meaning.

The same ten published statements met one, three and five at a time. The fraction of the set reached triples and then rises by two thirds again; the arrangements figure multiplies by more than three thousand.

Computed from the three numbers alone. How many statements exist, how many are met and how many can be checked is a stated convention; the counts are arrangements and combinations of whole numbers.

A linear number with non-linear consequences

Going from one met to five met, on ten published, moves the share from a tenth to a half: a factor of five, exactly as arithmetic suggests. For a practical example beyond the counting model, teams can visit Monitask and map the available records to a specific planning decision.

Over the same step the contents figure goes from ten to two hundred and fifty-two, a factor of more than twenty-five. The arrangements figure goes from ten to thirty thousand two hundred and forty, a factor of three thousand and twenty-four. The topic can also be checked against public guidance and research published by GitHub.

Nothing about this is paradoxical and all of it is easy to lose track of. The share is the input. The two figures are counts of possibilities, and counts of possibilities grow combinatorially in the input. Five times the share, twenty-five times the states of information, three thousand times the experiences.

Why the share is nonetheless the useful handle

Despite being the least informative of the record's quantities, the share is the one a set of pages can reason about, for two reasons.

It is the only one that can be estimated from outside without combinatorics. Somebody who knows roughly how far down a page gets read knows roughly what the share is.

And it determines both of the other figures completely, given the published count. There is no additional information in the contents or arrangements figures beyond what the share and the published count already fix. They are derived, and the derivation is the content of this record.

The share is small, simple, and estimable. The two figures built on it are large, awkward, and exact. Confusing which is which is how a set of pages ends up reasoning about a tenth as if it were a half.

The third number does not enter

Five of the ten published statements are checkable in every row of the drawing, and none of the figures discussed so far notice.

That is not an oversight in the drawing. Contents and arrangements depend on how much was published and how much was met, and on nothing else. A page could publish ten statements of which every single one is externally verifiable, or ten of which none is, and both figures would be identical.

What changes between those two cases is the chance that a reader meets something they can check, and that is a separate calculation with a separate drawing. The two halves of the record do not interact in either direction: the share tells you nothing about verifiability, and the checkable count tells you nothing about how many impressions are possible.

The fraction that is not the share

One figure in this record looks like the share and is not: the proportion of arrangements that open with a checkable statement.

That proportion is the checkable count divided by the published count — five tenths in the drawing's rows — and it has nothing to do with how much the reader met. It is the same on all three rows, although the share runs a tenth, three tenths and a half down the column.

The coincidence of form is a trap worth naming. Both quantities are a count divided by the published count. One is about how much got read and the other is about what was available to read. They look alike written down and measure unrelated things.

Why rearranging does not move it

The share is how much of the set a reader reached, and nothing about the order of the set enters it.

That matters because rearrangement is the cheapest thing a publisher can do and it is the one intervention this record says has no effect on either main figure. Moving a statement from the end to the beginning changes which arrangement a front-door reader gets. It changes neither how many arrangements exist nor how many contents, and for a reader who arrived in the middle it changes nothing whatever.

What moves the figures is the published count and the share. The first is a decision and the second is not, which leaves rearrangement as an act with real effects on individual readers and no effect on any quantity counted here.

What a small share does not excuse

A set of pages whose readers reach a tenth of it might conclude that the arithmetic here is not its problem: with ten contents possible, there is nothing much to be uncertain about.

The ten contents are each a single statement. The reader's entire account of the set is one sentence, chosen by circumstance from ten, and nine tenths of what was published played no part in it. The contents figure is small because the encounter is thin, not because it is predictable.

Thin and predictable are different properties, and only one of them is desirable. A low share buys a small contents figure at the price of nearly all the content, which is the least favourable trade available in this record.