Editorial standards

1. A development is not reported without a primary source

   Every story names the filing, rule, release, dataset, prospectus or court document that evidences it, and links to it. Where the only available evidence is another outlet's reporting, the story says so explicitly and attributes the claim to that outlet rather than asserting it ourselves.

2. Announcements are reported as announcements

   A stated intention to launch, list, tokenize or authorise is recorded as an intention. It is not reported as the thing itself, and the record distinguishes the two.

3. Every outlet we read is credited

   Story pages list every publication whose reporting informed the summary, with the headline, the outlet and the date. The desk writes its own summary; it does not reproduce anyone else's copy, and it does not read an outlet without crediting it.

4. Where sources disagree, the disagreement is printed

   When two credible sources give different figures, dates or characterisations, both are stated and the discrepancy is named as unresolved. We do not average them, pick the more dramatic one, or quietly follow the more recent one.

5. A missing number is shown as missing

   Gaps in a series are drawn as gaps. No interpolation between observations, no carrying the last value forward, and never a zero standing in for an unknown. A class nobody has sized is listed without a size rather than estimated.

6. Nothing is labelled live unless it refreshes

   The word live is reserved for series with an endpoint the reader's browser reaches and refreshes in the page. Everything else states that it is the value as of the last build, with the build time attached.

7. Every figure carries its provenance

   A published figure carries the source, the source URL, the method of compilation, the unit and currency, the cadence, the as-of date, the coverage it claims and the exclusions it does not. If one of those is unknown, the figure is not published.

8. Stale data is flagged, not hidden and not replaced

   Each series has a stated interval after which an observation is treated as stale. Past that point the figure keeps the last verified number and carries a visible flag. We do not remove the figure and we do not substitute an estimate.

9. We do not compile a number nobody published

   Where two providers measure the same quantity differently, we publish each provider's figure with its method attached. We do not reconcile them into a third number, because a compiled figure is indistinguishable from a measured one once it is on a chart.

10. Charts do not mislead by construction

    One vertical scale per figure, never two. Category order is fixed and does not change with the data. Colour identifies an entity, never its rank. A figure with two or more series always carries a legend, and every figure can be read as a table.

11. Reference pages carry a review date

    Every concept and jurisdiction page states the date the desk last checked it against current practice and the primary source. Nothing on this site is dated evergreen, and a page nobody has reviewed says so.

12. Analysis is separated from reporting and from other people's research

    Each longer piece is labelled at the top as analysis, a data feature or a summary of research published elsewhere. A research summary names the authors and links the paper, and their finding is never presented as ours.

13. Interests are disclosed in the story, not only on the About page

    The editor is a co-founder of Busha, a digital-asset exchange. Where Busha, or any institution with which the publication has a commercial relationship, is a party to a development, that is stated in the story. No coverage on this site is paid for, and no placement in any index, register or timeline is for sale.

14. Corrections are made in place and stated

    A material error is corrected on the page with a note recording what changed and when. Where a development was later reversed, the original entry and the reversal both stand; the record is not edited to look prescient.

15. We publish what we cannot yet establish

    Where a story turns on a fact we could not verify, the page carries an open question saying what is not settled and what would settle it. An unanswered question printed plainly is more useful than a confident sentence built on an assumption.

How to hold us to this

If a page on this site breaks one of the rules above, that is a defect and we would like to know about it. Write to desk@cryptoeconomics.com naming the page and the rule. Corrections are made in place with a note recording what changed.