Sixty-Four In Two Hundred And Twenty-Five

Sort the four hundred and sixty-one schemas by the share of their questions that reach a single cell. The middle value is sixty-four over two hundred and twenty-five — a little over twenty-eight per cent.

So on a typical structure in this record, roughly seven questions in ten leave at least one filter unset and return more than one kind of listing.

The strongest case for online timesheet reporting is a narrow question, a short pilot and a review that includes outcomes, records and participant feedback. For a separate technical reference, see Google Cloud Looker.

The three schemas at the median share

• one bar is one facet; the divisions are its values
• cells: how many listings the schema can tell apart
• queries: how many different questions it lets anybody ask
• sizes: how many different result sizes are reachable

Three schemas drawn.

All three reach a single cell with sixty-four questions in every two hundred and twenty-five. The first has sixty-four cells and two hundred and twenty-five questions outright; the others reach the same share by different routes.

Computed from the schema alone. How many facets and how many values in each is a stated convention; the three figures are products of whole numbers.

The three schemas that sit there

The first is the one to remember. Its cells are sixty-four and its questions are two hundred and twenty-five, so the median share is simply those two figures written as a fraction. Nothing has been reduced or rearranged.

The other two reach the same share from different sizes entirely — one with five facets and six hundred and forty cells, one with six facets and six thousand four hundred. The share is a ratio, and ratios do not care about scale.

What the figure is a median of

The four hundred and sixty-one schemas, not any collection of real boards.

This caveat attaches to every median in the record and it matters most here. The enumeration leans heavily towards long schemas: two hundred and ten of the four hundred and sixty-one carry six facets, one hundred and twenty-six carry five.

Long schemas have low shares, because every facet multiplies the share by a fraction under one. So the median is pulled down by the shape of the enumeration, and a collection of actual boards — which would carry fewer facets on average — would sit higher.

The median describes what the record contains. It is not a forecast about any board, and the record says so wherever it quotes one.

Reading twenty-eight per cent correctly

It is a property of a schema, not of anybody's searching.

Nobody works through the available questions in order. Real readers set one or two filters, look, and adjust. The share says nothing about which questions get asked and everything about which are available.

What it does say is structural and useful: on a median schema, if somebody wanted to name one particular kind of listing, under three in ten of the expressions available to them do that. The rest return groups.

Whether returning groups is good or bad depends entirely on what the reader wanted, and a record that counts structures cannot know.

Where the median sits against the ends

The range runs from sixty-four over seven hundred and twenty-nine at the bottom — under nine per cent — to six sevenths at the top, which is nearly eighty-six.

Twenty-eight per cent is nearer the bottom than the middle of that range. That is the enumeration's bias again, and it has a plain cause: there are far more ways to build a six-facet schema than a one-facet one, and every extra facet pushes the share down.

Why three schemas and not one

A share is a ratio, so any schema whose fractions multiply to the same value lands on it.

Two, two, four and four gives two thirds times two thirds times four fifths times four fifths. Two, four, four, four and five adds a five sixths and replaces a two thirds with a four fifths, and the product is unchanged.

That is how a ratio behaves and it is why the share, alone, is the weakest of the four figures for identifying a board. It discards both absolute quantities and keeps only their relation.

The fraction worth keeping

Sixty-four over two hundred and twenty-five is not a memorable fraction and does not need to be.

What is worth keeping is the shape: on a schema of middling size, most of what a reader can ask returns a group rather than a single kind. Reaching one particular thing requires setting every filter, and on a long sidebar that is a demanding requirement which most readers will not meet and should not have to.

The next page takes the two ends of the range, where the same structural fact is much starker in both directions.

Near this one

• Six Sevenths, And Under A Tenth
  On a board with one filter of six choices, six questions in seven reach a single kind of listing. On a board w
• One Fraction Per Filter
  The share of pinning questions is the product of size over size-plus-one. Each filter contributes a fraction,
• What Each Filter Costs The Reader
  Adding a filter raises resolution and lowers the share of questions that reach one listing. The two effects pu
• The Board With One Filter
  Five schemas carry a single facet. They hold the highest share of pinning questions, the fewest result sizes,