The head all three addresses share

The fact itself, measured here

The table below is not typed in. It is produced by comparing the three strings printed at the top of this page, character by character, at the moment the page is built. If the set ever changed, the table would change with it and nobody would have to remember to edit a sentence.

The first column is identical on all three lines and is 8 characters long. The bold character is the first one that differs. Both figures are measured from the strings above at the moment this page is built, not typed in by hand.

Why a set can have a shared head at all

An onion address comes out of a key, and you do not get to choose what it says. What an operator can do is generate keys repeatedly and keep one whose address happens to begin with the characters they wanted. Longer runs of chosen characters take disproportionately more attempts than shorter ones.

So a shared head across a set is a statement about work. Somebody wanted the members of the set to look like members of a set, and was willing to do the same search more than once to get it.

The one thing it settles

It settles negatives quickly. A string offered as a WeTheNorth market url that does not open with the same head as the ones above is not part of the published set, and you can put it down without further thought.

Fast negatives are rare and worth having. Most checks available to a reader are slow, inconclusive or both. This one takes a couple of seconds and does not need any software.

The much larger thing it does not settle

A matching head proves nothing. The head is public. It is printed on this page and on every other page that lists these strings. Anybody who can read it can go and generate their own address that begins the same way, and the reason they would bother is that the head is the part readers check.

This is the trap that plurality creates and a single address would not. The shared opening looks like a family resemblance, and a family resemblance is exactly the impression a fourth string wants to give.

Which end of the string carries the information

The head is shared, so it carries almost no information about which member of the set you are holding. Everything that distinguishes one from another is in the part after it, and the divergence starts immediately: the bolded character in the middle column above is the first one that differs, and it is the very next character.

The practical consequence is that reading the first eight characters and feeling satisfied is the wrong habit. If you are checking a string by eye, the tail is where the check has to be done, character for character, against text you already hold.

How to use it without over trusting it

1. Use the head as a filter that throws things out, never as a reason to accept anything.
2. If the head does not match, stop. Nothing further is needed.
3. If the head does match, that means you are now at the beginning of the check rather than the end of it.
4. Compare the remainder against a copy you already hold, not against memory, which is not able to hold fifty six characters reliably.

What the measurement above does not depend on

The table is produced by comparing the strings on this page with each other. It does not depend on any claim about who generated them, when, or why, and it would produce the same figures for any three strings that happened to share an opening.

That is deliberate. A measurement that only works if you already trust the source is not a measurement, it is a restatement of the trust. This one is checkable by eye, by anybody, in a few seconds, which is the highest standard anything on this subject reaches.

What this study is not. It is not a way of telling whether an address is genuine, and a matching head is the beginning of a check rather than the result of one.