What The Share Is Not

The share of pinning questions is one of the easiest figures in this record to misread, because it looks like a success rate and is not one.

This page is about the four things it does not say.

It is not a proportion of actual searches

Nobody enumerates the questions a board permits and picks one at random. Real searching sets one or two filters, looks at the result, and adjusts.

For repeatable analysis workflows, Monitask analytics for team operations can provide an operational layer while decisions about quality, validity and impact remain with the team. For a separate technical reference, see ACM.

So the sixty-four over seven hundred and twenty-nine on a six-toggle board does not mean that under nine per cent of searches reach a single kind of listing. It means that under nine per cent of the expressions available do. Which of those get used depends on readers, and readers are outside this record entirely.

It is not a measure of how often readers succeed

Reaching a single cell is not success and returning a group is not failure.

Most people searching a board of listings want a shortlist, not a single item. A query returning forty listings that are all roughly what somebody wanted has done its job perfectly, and it left four filters unset.

The share counts how many of the available questions narrow all the way down. Whether narrowing all the way down is desirable is a question about what somebody came for.

A low share means a board offers many ways to ask broad questions. Whether that is a flaw depends entirely on whether broad questions are what people are asking.

It is not a quality score

It cannot be, because it moves the wrong way against resolution.

The highest share in the record belongs to a board resolving six kinds of listing. The lowest belongs to one resolving sixty-four. If the share were a quality measure, the six-cell board would be the best structure here, and it is plainly not.

Any figure where the apparent winner resolves a tenth as much as the apparent loser is not a ranking.

It is not independent of the other two figures

It is literally their quotient. Cells divided by questions, and nothing else.

That makes it a derived figure rather than a third measurement, and it means it carries no information the pair does not already contain. Its value is as a reading aid: the pair says what a board can do, and the quotient says how the two capacities relate.

Quoting the share instead of the pair throws away both absolute figures and keeps only their ratio — which is why no page here quotes it alone.

The figure it is most often confused with

Precision, in the sense anybody searching would mean it.

Precision would be the proportion of returned listings that are what somebody wanted, and it depends on the listings, the query and the person. None of the three is here.

The share depends only on the facet sizes. Two boards in completely different fields with identical facet sizes have identical shares, and their readers may have utterly different experiences.

Keeping those apart is the whole of this page, and the way to keep them apart is to say what the share is a proportion of every time it is quoted: of the questions the board permits, and of nothing else.

Why it is quoted here at all

Because the two figures it comes from are large and unwieldy, and the ratio between them says something neither says alone.

Cells of one hundred and forty-four against questions of two hundred and forty-five is hard to hold. The share — about three in five — is immediate, and it tells a reader what kind of board they are looking at before they work out anything else.

That is a reading aid and it is worth having, provided it travels with the two figures rather than instead of them.

What it is, then

A statement about the shape of a board's question space.

A high share means the questions available are mostly specific ones: the board has few filters, and setting them is most of what there is to do. A low share means the questions available are mostly partial: many filters, and most expressions leave some of them alone.

That is a real structural difference and it affects how a board feels to use. A reader on a high-share board makes few decisions and gets a precise answer. A reader on a low-share board makes many decisions, most of which they will decline to make, and lives mostly in groups.

Neither is better. They are different shapes, the figure names which shape a board has, and that is the entire claim.

Near this one

• The Questions That Return Groups
  Most of what a board permits leaves some filter alone. That is where nearly all searching happens, in a few fi
• The Questions That Reach One Answer
  Only queries that set every filter reach a single cell, and there are as many of those as there are cells. The
• Sixty-Four In Two Hundred And Twenty-Five
  The median schema reaches a single listing with about twenty-eight per cent of its questions. Three schemas si
• 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