Policy
The methodology and the disclosure boundary are published first
Before there is any data to publish, the definitions used to produce it and the boundary of what will never be published are both public and versioned.
Most measurement projects publish results first and definitions later, if at all. That order is convenient and it is why five public studies of the same phenomenon can report five different numbers without any of them being wrong.
We are doing it the other way round. The definitions are published before the first reading, stamped with a version, and frozen at v1.0 before any entry depends on them. Changing a definition afterwards costs a republication and says so on the record.
What is public
The full set of definitions used to produce a published figure: what counts as an endpoint, how an interval is measured, how magnitude is assigned, which cases are excluded and why, and the step-by-step procedure a reader follows to recompute an entry from public chain data. All of it is on the methodology page.
Alongside it, the boundary itself: the categories of information that will not be published at any point, and the reason each one is withheld.
What is not, and why we say so
We publish what we measured. We do not publish how we decide where to measure.
Those are different things, and the second one is the only part worth protecting. Publishing the first without the second is what makes a reading checkable; withholding the second is what keeps the instrument from being trivially copied. The methodology page argues this boundary explicitly rather than leaving readers to notice the omission.
No byline
Everything here is published under the company name. There is no author credit on any entry, any study, or any post — including this one.
That is not modesty. Verifiability is the substitute for attribution: an entry that anyone can recompute from public data does not need a name attached to be trusted, and an entry that cannot be recomputed is not made trustworthy by one.