Five figures this site deliberately does not publish

Five figures that appear routinely elsewhere are absent from this site: order-book depth, holder concentration, unlock schedules, fund flows and realized capitalization. Each is omitted for the same reason, stated in the methodology page as a rule — a number is published only when its definition, its feed and its observation time can all be stated, and when a competent outsider could in principle reproduce it. The five below fail that test in five different ways, and the way each one fails is more instructive than the figure would have been.

Order-book depth

Market depth is the quantity of resting orders within some distance of the mid price on an order book. Published as a single figure — depth within two percent, summed across venues — it implies something it cannot support. Depth is an instantaneous snapshot of intentions that can be withdrawn in milliseconds, and much of it belongs to market makers who cancel on any sign of adverse selection, so the depth visible before a large order is frequently not the depth available during it. It is also not additive across venues: liquidity on one exchange cannot absorb an order routed to another, and the same market maker's inventory may be quoted in several places at once.

A site that publishes depth once a day is publishing one arbitrary sample of a quantity that changes continuously, aggregated by a rule that assumes venues combine when they do not. The concepts remain worth understanding — slippage, price impact and the bid-ask spread all have entries in the glossary — but the site publishes the measurable neighbors instead: 30-day volume and turnover, both with the caveat that venue-reported volume has a documented history of inflation through wash trading. Structural liquidity concerns are handled in words on the risk section rather than in a false number.

Holder concentration

The claim that some number of addresses control some percentage of supply is the most widely repeated statistic in this field and one of the least sound. An address is not a holder. The largest addresses on most chains are exchange omnibus wallets holding customer assets, custody providers, bridge contracts, staking contracts and burn addresses. One exchange address represents millions of people; one person may control thousands of addresses. On chains using UTXO accounting the count is looser still, because a wallet routinely spreads one balance across many outputs.

Correcting for this requires address clustering, which is a set of heuristics — common-input ownership, change detection, exchange labeling — that are probabilistic, differ between vendors and degrade whenever a service changes its wallet architecture. Two providers can publish concentration figures that disagree by a wide margin with no way for a reader to adjudicate.

What can be measured is concentration among identifiable, labeled parties. The Nakamoto coefficient counts the minimum number of entities required to compromise consensus, which is observable from the validator or mining set rather than inferred from balances, and it sits alongside validator count on the staking section. Supply held by exchange-traded products and supply held by corporate treasuries are published because those holders disclose their positions in filings. Where insiders received allocations at inception, the site describes the arrangement through the insider allocation entry rather than asserting a present-day percentage.

Unlock schedules

A dated calendar of future token releases looks precise and rarely is. The terms come from founding documents that can be amended, from private agreements that are never published, and from contracts whose administrator can extend or accelerate a vesting cliff. Allocations move over the counter before they unlock, so the economic transfer often precedes the on-chain event that a calendar would mark. Some schedules are enforced by contract and genuinely verifiable; many are enforced by promise. A calendar that presents both in the same format misrepresents the second kind.

The site publishes the aggregate instead, where it is observable on-chain: the issued share of supply, the pending share, dilution overhang and the ratio of diluted to circulating valuation. Those describe the size of what has not yet reached circulation without inventing a date for it. The unlock schedule and unlock overhang glossary entries explain what to look for in an asset's own primary documents, and the calendar lists only dated events with a verifiable source, such as protocol upgrades and scheduled parameter changes.

Fund flows

Statements that a given sum flowed into or out of an asset over a week are usually reconstructions rather than measurements. On-chain, a flow is inferred by labeling addresses as belonging to exchanges and netting transfers across that boundary — the quantity described by the exchange netflow entry. The labeling is the weak link: exchanges rotate wallets, move balances between hot and cold storage, and hold assets for other exchanges, so a large internal reorganization can register as a flow that no customer initiated. Off-chain, flows are estimated from volume and price, which measures trading rather than net movement of capital; every purchase has a matching sale, and no dollar amount enters an asset in the sense the phrase implies.

The one place where a genuine flow figure exists is the regulated product wrapper, because share counts are disclosed. Coins held and assets under management for an exchange-traded product come from the issuer's own daily disclosure, and changes in them reflect real creations and redemptions executed by authorized participants in creation units. Those series appear on the exchange-traded product section and are labeled as holdings, with the change derived from them, rather than as a market-wide flow.

Realized capitalization

Realized capitalization values each unit at the price when it last moved on-chain rather than at the current price. The idea is defensible and the measurement is not portable. It depends on a chosen price source for every historical moment, on a rule for handling units that have never moved and may be permanently lost through key loss, and on the assumption that a transfer represents a change of ownership — an assumption broken by exchange rebalancing, wallet migrations, custodial internal transfers and contract interactions. It is defined naturally on UTXO chains and awkwardly on account-model chains, where balances are commingled rather than tracked as discrete units, so a single figure cannot be compared across the two designs.

Different vendors publish different values for it under the same name. Rather than pick one convention and present the result as a fact, the site publishes the observable inputs on the supply section and the current market capitalization with its formula attached, and treats cost-basis estimation as analysis belonging in the research section, where a method can be stated and argued with.

The general rule, and its cost

Omitting these five has a price, and it is worth naming: readers who want them will find them elsewhere, and a page without a concentration figure looks less complete than one with a wrong concentration figure. The trade is deliberate. A number that cannot be defined, sourced and timestamped cannot be checked, and an unverifiable number in a table of verifiable ones lowers the credibility of everything beside it. Where a quantity matters but cannot be measured well, the site describes the mechanism in the glossary and the specific exposure in the risk section, which is a weaker claim honestly made.

The natural next step is the final article in this track, which walks through the feeds behind the figures the site does publish, followed by the data sources page for the current list and the methodology page for the rules that govern additions to it.