Same filter

Why They Agree, And Why It Is Not A Coincidence

Shapes that admit identically do so for a reason that can be stated in one line, and the reason explains the whole ladder of classes.

Home · Same filter · Reference

Twenty-one shapes admit one profile in two hundred and fifty-six. They were not selected for that property; they fell out of the enumeration.

The reason they agree is short enough to state exactly.

The reason

A shape admits a fraction whose numerator is the number of ways to pass and whose denominator is two to the power of the total items.

If a team already has strong project records, it can still explore the available workforce tools and compare the configuration without committing to deployment. For an independent technical reference, see OWASP.

Two shapes agree exactly when those two fractions reduce to the same thing. For the class at one in two hundred and fifty-six, every member has a numerator of one — a single passing route — and a denominator that reduces to two hundred and fifty-six.

That happens whenever the items that actually bite number eight, and any number of items that do not bite are added alongside, since they contribute equally to the numerator and the denominator and cancel.

Working one case

Take eight compulsory items and eight optional ones with a threshold of none.

The space is two to the sixteenth: sixty-five thousand five hundred and thirty-six. The ways to pass are: meet all eight compulsory items — one way — and anything at all on the optional eight, which is two hundred and fifty-six ways. So two hundred and fifty-six passing routes out of sixty-five thousand five hundred and thirty-six.

Reduce: one in two hundred and fifty-six. The eight unrequired items multiplied both the numerator and the denominator by the same two hundred and fifty-six, and vanished.

Items that cannot exclude anybody multiply the numerator and the denominator equally. They are arithmetically invisible, which is exactly why they are visible to everybody reading the page.

Why the ladder descends by one

Each rung of the ladder corresponds to a core of biting items of a given size: eight for one in two hundred and fifty-six, seven for one in one hundred and twenty-eight, and so on.

A core of eight can be split into a compulsory part and a fully-required optional part in nine ways, and then extended with unrequired optional items in several more. A core of seven can be split in eight ways. One fewer split, one fewer member, each step down.

So the ladder is a counting result about splitting a core, not a pattern anybody arranged.

What this rules out

It rules out the hope that equal-admitting shapes differ in some subtler way the fraction is missing.

They do not. Under the record's conventions, two shapes with the same fraction are the same filter in every respect the arithmetic can see — same intake, same proportion excluded, same everything. Any difference between them lives entirely outside the arithmetic, in how they read.

That is a strong statement and it is worth being precise about its scope. It holds under the coin convention. A model that weighted items differently could easily separate two shapes this record calls identical, and the page on what the arithmetic cannot say takes that up.

A second case, worked

Take seven compulsory items with five optional, three of the five needed, and compare it with four compulsory items and four optional with all four needed.

The first: twelve items, so four thousand and ninety-six profiles. Five optional items give sixteen profiles meeting three or more — ten, five and one. Seven halvings over that give sixteen of four thousand and ninety-six, which is one in two hundred and fifty-six.

The second: eight items, two hundred and fifty-six profiles, one passing configuration. One in two hundred and fifty-six.

Twelve lines against eight, a threshold on one and not the other, and the same answer. The agreement is not approximate and not a rounding artefact — both reduce to the identical fraction.

Why no class is larger than twenty-one

Because the number of routes to a value is bounded by how many threshold sums land on the right power of two, and binomial coefficients are sparse.

A value of one over two to the power of some number can be reached by halvings alone, by a conjunction in any mixture, or by a threshold sum that happens to equal a power of two. The first two routes are plentiful; the third is rare, because most running sums of binomial coefficients are not powers of two.

One in two hundred and fifty-six collects the most routes because eight is the cap on the compulsory part, which puts the meeting point of the two plentiful families exactly there. Every other value draws on fewer, and the largest class in the record is therefore twenty-one shapes rather than anything more.

The usable version

If two lists reduce to the same fraction, stop comparing them on strictness. They are equally strict, exactly, and no amount of staring at the items will change that.

Compare them instead on the things that differ: what they tell a reader, how many lines they occupy, whether their rule is stated, and who they will put off. Those are real differences and the arithmetic has cleared the ground to see them.