Measurement
Methods for turning raw chain data into a defensible number, and for telling a broad market from a narrow one that only looks broad.
What this area is about
Measurement is where chain data becomes a claim. The data itself is neutral: signatures, slots, balances before and after. A claim requires choices — which events count, over which window, grouped by what, compared against which baseline. This area is about making those choices explicitly and writing them down, so a reader can disagree with the method rather than merely doubt the conclusion.
Two practical problems recur. The first is the missing baseline: activity is measured only after something has already changed, leaving nothing to compare against. The second is the missing denominator: a total is quoted without the participant distribution that would say whether it came from many independent wallets or a handful cycling the same balance.
Both problems are solvable with public data and ordinary arithmetic. Neither is solvable retroactively, which is why the protocol note insists on recording the baseline before anything runs.
The standard applied here
A measurement published by this desk states its source, window, timezone and grouping rule, or it is not published. Where a test is a heuristic rather than proof, the note says so and describes the innocent explanations that produce the same signature. Structural context for these methods sits in market structure, and the depth figures they lean on are covered in liquidity and depth.
Field notes in this area
- 07 Organic Volume vs Manufactured Volume What each pattern looks like in raw transaction data: wallet distribution, timing regularity, size clustering, round trips — and the honest grey zone in between.
- 09 Measuring a Volume Campaign A measurement protocol: define the objective, take a baseline before anything runs, pick metrics that survive scrutiny, and report windows that match the activity.