A measurement and a judgement are different things
Everything in this track produces measurements; a view about worth is a judgement, and keeping the two apart is what the whole method depends on.
A measurement states what was observed, by what method, over what period, with what known error. A judgement adds assumptions about a future nobody has seen and weights them according to someone's priorities. Both are legitimate activities. Presenting the second in the clothing of the first is how most weak analysis in this sector is produced.
This site publishes measurements. It does not publish judgements, and this article sets out where the line falls and why it is drawn there.
What makes something a measurement
Four properties, all of which can be checked by a reader. A stated object: which quantity, for which asset, on which chains. A stated method: which contracts were included, which addresses excluded, which window used. A stated period: an exact interval rather than a vague recent. And a stated uncertainty: whether the series has been subject to a revision, whether the source is a disclosure or an estimate produced by a data provider, and whether coverage is complete.
A figure carrying all four can be checked, disputed and reproduced. A figure missing any of them cannot, and the missing element is usually the one doing the work. The methodology and data sources pages exist so that every number here carries all four, including the awkward ones.
What a judgement adds
A judgement takes measurements and adds at least one of: a forecast, a weighting, a threshold, or a tolerance. It might assume a fee stream persists, or that a governance vote passes, or that a regulatory position holds. It might decide that supply concentration matters more than protocol revenue. It might set a level at which a ratio becomes worth attention. It might accept a risk another person would refuse.
None of those additions is observable. They are properties of the person making the judgement, not of the asset. Two people looking at identical, correct measurements can reach opposite conclusions without either having made an error, which is not true of measurement disagreements, where at least one party is wrong about something checkable.
| Statement | Type | What it depends on |
|---|---|---|
| Fees over the last 30 days totaled a stated amount on a stated boundary | Measurement | The pipeline and the boundary definition |
| Fee growth over 30 days was negative while market cap rose | Measurement | Two series and one window |
| That divergence will close | Judgement | An unstated model of what drives each series |
| Sixty percent of supply is unissued on a published schedule | Measurement | The schedule and the supply definition |
| The schedule matters more than the fee stream | Judgement | A weighting the reader has chosen |
The words that smuggle a judgement in
Certain vocabulary converts a measurement into a conclusion without announcing it. Describing a ratio as low implies a reference level that does not exist. Describing supply concentration as concerning imports a threshold nobody has stated. Describing a fee decline as temporary is a forecast wearing the grammar of an observation. Describing an asset as one the market has mispriced asserts knowledge of a correct price that no method in this track can produce, and quietly assumes the market capitalization on the page is wrong in a knowable direction.
The disciplined alternative is mechanical. State the number, state the comparison it is being made against, and leave the adjective to the reader. This site's screener and compare tools are built to that rule: they rank, filter and display, and they do not label.
Why the distinction is load-bearing here specifically
In equities, a judgement is disciplined by an audited denominator and a long record of outcomes. Neither exists for digital assets. That means judgements in this sector rest on thinner ground, and correspondingly more of the work is done by the assumptions rather than by the evidence. A framework that blurs the line will therefore mislead further here than the same framework would elsewhere.
The record supports the caution. The 2022 collapses of Terra, Celsius, Three Arrows and FTX were preceded by dashboards showing measurements that were largely accurate. Terra's on-chain figures were real. The mechanism binding its stablecoin to its peg was an algorithmic design whose failure mode was structural rather than a data problem, and the risk of a depeg was a property of that structure rather than a number on a chart. The measurements did not fail. The judgements built on them did, because they treated the absence of a number for a fragility as evidence of the absence of the fragility.
Keeping the two apart in practice
Writing them down separately is the whole technique. A note recording the number, the metric key, the date observed and the method version is a measurement record that will still be interpretable in a year. A separate note recording what the writer concluded, which assumptions the conclusion required, and what would falsify it is a judgement record that can be revisited honestly when the situation changes.
Kept apart, a revision to a measurement does not silently invalidate an old conclusion, and a change of mind does not require rewriting history. Merged, neither can be audited, including by the person who wrote them, which is the failure mode that matters most because it is invisible from the inside.
The last piece is the willingness to record that a question has no answer available. Where an asset has no fee stream, where a boundary is contested, where a series contains a structural break, the accurate entry is that the measurement does not exist. That entry is more useful than a number nobody can defend, and it stays useful, because it does not decay into a false memory of having known something.
The metric catalog gives the definition and window behind every figure on this site, the methodology pages give the boundaries, and the research pages show how the measurements are assembled without crossing into a view. Read alongside the glossary, they are the whole of what this site claims to provide.