Skip to crypto research
Crypto Trading Strategy GuideVenue / Contract / Settlement

Order-book depth is not your crypto fill

Keep snapshot continuity, displayed quantity, requested quantity and unique account execution separate.

A displayed book is one source object

A book feed describes price levels and displayed quantities under its selected snapshot/delta contract. A private execution record describes what a particular account actually traded. Their identities and clocks are different. Preserve feed sequence, snapshot basis and gaps before assuming continuity. The desk has not captured an order-book stream or reconstructed a real queue. Public documentation explains one feed mechanism, not a calibrated execution model or quantity guarantee. A visible favourable price is not proof that every intended unit filled there or that the source was available to the strategy.

Quantity changes the evidence request

A desired position of ten units and a displayed level of four units are not the same exposure. Actual matching may involve multiple prices, roles and partial quantities under product rules. The tutorial does not implement a matching engine, queue priority or depth simulator. It also does not extend the canonical research-fill window to search for a later convenient price. Record requested, acknowledged, executed and remaining quantities with original IDs. An authentic result needs the whole relevant execution population rather than a screenshot of the best visible level.

Cumulative status fields need careful interpretation

A state message can report cumulative executed quantity rather than a fresh fill amount. Repeated status messages must not be summed as independent executions. Selected venue documentation warns about particular race conditions; those details are scoped, not a universal account history. The invented outage drill keeps partial execution and cancelled remainder separate. It cannot prove actual account flatness, fill fees or an average execution price. Unique execution records and current position sources remain necessary for real economics, together with the applicable query and retention contract.

A stronger test needs licensed causal data

Historical depth, private fills and effective specification must be available under rights suitable for the intended research. Save original source and availability fields, missing episodes and any execution assumptions. Causal joins and a frozen rule precede evaluation; actual strategy economics still need independent master completion. Software tests of the amount lens supply no liquidity or financial sample. The fictional example below illustrates a quantity mismatch only. It reports no live spread, order-book level, queue success rate, customer experience or measured platform slippage.

Fictional crypto evidence case

An invented observation displays four units at price 100 while a request intends ten. There is no later book or private execution source. The narrow finding is incomplete quantity/execution evidence, not a ten-unit fill at 100 or a measured six-unit rejection. No venue, timestamp or real market data is involved.

ObjectRecordBounded finding
DisplayedFour assumed unitsBook object only
RequestedTen assumed unitsClient intention
ExecutedNo authentic sourceUnknown
Invented quote-settlement price change, entry fee, exit fee, funding and net cash
One invented quote-unit reconciliation, not real account cash or a probability distribution.

Primary references and scope

Source review: 14 October 2026. Official mechanism descriptions are not this desk's paid-account test, independently captured fills or vendor-return audit.

Related crypto questions