Editorial standards

Clearing a piece for publication and putting it in front of readers need different capabilities, held by different people more often than not. Publishing somebody else’s work needs a different capability again from publishing your own, so a writer who can send out their own reporting still cannot send out a colleague’s.

Nothing on this site publishes itself. There is no process that turns a scheduled piece live on its own, so going live is always a person’s decision at the moment it happens.

A piece with no byline cannot be published, and that refusal is in the software rather than in a habit. A byline is an ordered list of the people credited and what each of them did: lead writer, contributor, photographs, editing. There is at most one lead writer, and where there is one, their name leads.

Nothing is credited automatically. Creating a piece leaves it unbylined, because the person who opens a story is frequently not the person who will write it, and a name on an article nobody chose to put there is worse than a blank. The date a piece first went out is recorded once and cannot be rewritten afterwards, by anybody, which is what makes a citation of it dependable.

When news will not wait, a senior editor can take a piece straight to publication. That route has its own capability, it demands a stated reason, and it marks the piece for review after publication. The mark cannot be cleared by the person who used the route. It exists because deadlines are real, and it is built to be visible rather than quiet, because a shortcut nobody can see stops being exceptional.

A correction is a record attached to the article rather than an edit made to it, so what was wrong, when it was fixed and who fixed it all survive the fix. It is published on the piece it corrects, carries the date it was made, and stays there.

There are four levels, scaled to the error: a minor fix, a material correction with a visible dated note, a serious error that goes to senior editorial with a review of the sourcing, and a retraction. The database refuses to delete a correction or to reduce its severity afterwards, because a correction that could be quietly softened is a deletion wearing a disguise. Complaints and corrections is how to ask for one.

Where a central claim cannot be stood up, the piece is retracted: the text comes down, the explanation goes up, and the address stays live. Deleting the page would break every link to it and remove the record at the moment the record matters most. A retraction cannot be reversed, quietly or otherwise. It is the one state in the whole workflow with no way out of it.

A media release from a board, a club or a governing body is held as its own kind of piece and labelled as one wherever it appears: on the card, in the section, and in the related reading beside another story. Nobody should have to open a piece to discover that it is an announcement rather than CricZone’s reporting on that announcement.

CricZone sells advertising and sponsorship, and what is for sale is set out in public on the advertise page rather than in a media kit somebody has to ask for. Nothing is running at the moment: the advertising you see on CricZone today is CricZone promoting its own newsletter and membership, and no advertising network is connected to this site.

A sponsored piece is labelled as sponsored on the card, before you have opened it, because the card is where most readers decide. Inside a piece, a sponsor cannot be named without a sentence saying what the arrangement is: the software refuses to store the mark without the disclosure, so a logo with nothing explaining it is not something an editor in a hurry can produce.

Paid membership is not available at the moment, and when it is, a membership will remove advertising. It will not remove a disclosure. A member who has paid for an ad-free page still sees who sponsored the coverage they are reading, because taking that away would leave them with commercially supported work and the commercial part removed, which is the opposite of what they bought.

Every image in this system carries a declaration of what a model did to it, and there are four answers: no AI involvement, AI assisted, AI generated, and AI upscaled. An image nobody has declared is recorded as undeclared rather than assumed to be a photograph. That distinction is deliberate: an archive carried over from an older system contains pictures nobody has looked at, and quietly filing them as photography would be this publication asserting something it does not know. How CricZone uses AI is the fuller statement.

What CricZone has published stays published and stays reachable. When the address of a piece changes, the old one is kept and goes on sending readers to the current version, so a link somebody saved or cited does not quietly stop working. An address that has ever belonged to one article can never be given to another, which is what makes an old link unambiguous rather than a guess.