Missing conflicting and stale evidence

Two providers can disagree because they measured different windows, not because one is deceptive. Picking the convenient value conceals the actual evidence problem.

Start with a conflict ledger. For each value record asset identity, chain, block/slot, raw unit, finality basis, pagination and retrieval time. Normalize compatible units while retaining raw responses. Values from different blocks can both be valid observations; they do not describe one common snapshot. A missing field is not zero, and a stale field is not refreshed by retrieving the same cache again.

When the records truly target the same state, test decoder version, token decimals, dropped events, pending/reorg handling and pagination. Do not average incompatible balances into a synthetic “consensus” figure. An independent primary state record can resolve a field conflict, but only if its scope and identifier match. Preserve the earlier conflict and explain the resolution rather than overwriting evidence.

The conclusion should identify which narrower claims are supported and what remains unavailable. UNKNOWN labels an unresolved field; INSUFFICIENT EVIDENCE means a proposed decision, such as a withdrawal accusation, cannot be supported from those fields. A tool’s blank panel may be a coverage failure, not a clean result.

Stale: a value whose measured state predates the claimed observation.

Conflict ledger: side-by-side source records with compatible-field checks.

Missing: unavailable field, distinct from numeric zero.

Resolution: evidence-backed explanation of disagreement.

ILLUSTRATIVE: provider A gives balance 100 at block 50 with 0 decimals. Provider B gives 9000 raw units at block 51 with 2 decimals: 90 displayed units. Provider C returns null at block 51 because its supported event type is unavailable. A claimed withdrawal of 10 units is proposed, but no transition receipt is supplied.

Evidence panel with A/B/C rows, raw/display amounts and block columns. Flag time mismatch and missing C; show withdrawal claim as unresolved, not a red fraud badge.

Caption: Fictional conflicting observations are reconciled before any event attribution.

SPECIFICATION_ONLY — the rendered visual has not been produced or independently reviewed.

A and B describe different stipulated block states; B converts to 90 units. C is missing, not zero. The apparent 100-to-90 difference is a candidate state change requiring transition evidence.

It does not prove a withdrawal event, controller intent, provider fraud or a current balance. Even a real balance difference could reflect several transitions or incompatible accounting.

Treating null as zero; averaging 100 and 9000; ignoring block mismatch; calling the difference a theft; losing the raw records after normalization.

Build the conflict ledger, compute B, reject one invalid aggregation, and write a conclusion about the proposed withdrawal. Specify the next decisive record to request.

• Align identity, unit, time and finality basis.
• Preserve null as unavailable.
• Reject averages of incompatible states.
• Name the evidence needed to resolve the specific claim.

Resolve the measurement basis first; if decisive transition evidence is absent, report the precise unresolved claim.

• Null is not zero.
• Different block states are not one snapshot.
• A balance difference alone does not name a mechanism or motive.

P10-L07

Use the named ZECOIN tool only as an evidence-reading context. This lesson creates no tool output, account session or entitlement. Offline exercise; do not sign, deploy, approve or fund anything.

• API: Esplora public API specification — GET routes, integer satoshis and indexed confirmation/outspend fields. Boundary: Mechanism documentation only; original illustrative values are stipulated, not observations. Historical records have separately frozen provenance and explicit collection gaps.

ILLUSTRATIVE — entirely fictional offline inputs; no current observation.

null — no real observation

1

null

needs_review