Method
Show the working.
In a domain full of confident charts and unreproducible numbers, the differentiator is not having data — it is being checkable. This page states exactly how that works here.
Where data comes from
- On-chain (primary). Subnet state is read directly from Bittensor’s Finney chain — pool reserves and spot prices off a live node, and historical state off an archive node at named block heights. These are the figures we trust most, because anyone can re-run them.
- External (labelled). Some context comes from third-party analytics and news sources. When it does, it is named in the piece, and its refresh behaviour is disclosed.
The rules we hold ourselves to
- One number, one source. No figure appears without a described origin.
- Freshness stated on the figure. Where a source republishes on its own schedule, the figure shows when that source last changed — not when we last looked.
- No silent estimation. If something is approximated, interpolated, or assumed, it says so in the text.
- Reproducible snapshots. Report figures come from a dated data snapshot, exported alongside the piece so the analysis can be re-run.
- Corrections are public. When we get something wrong, the correction is published — not quietly edited away.
What this means for you
You should never have to take our word for a number. Read it, check the source, re-run the query. If it doesn’t hold up, tell us — that is the entire point of doing it this way.
Verification tooling and the published snapshots that back each report are being built. Until a figure carries a snapshot reference, treat it as provisional.