Methodology

PayOutIndex is a neutral, read-only payout observability registry for online casino withdrawals. It records declared payout conditions per operator, market, and payout method, with a source label and a timestamp on every visible field.

PayOutIndex does not rank, recommend, or endorse operators. It is not a review portal, complaint forum, affiliate site, or casino product.

Every record carries an origin label. The admin UI writes admin_declared and unknown. Brand accounts may submit brand_declared data through a moderation queue. user_reported is reserved and not currently active.

• admin_declared — entered by a PayOutIndex administrator from a primary or secondary source.
• brand_declared — submitted by an operator brand through the brand program and approved by an admin before publication.
• user_reported — reserved for a later phase; not currently active.
• unknown — origin unverified or not recorded.

Sources are never silently merged. Each (operator, market, method) cell holds one canonical row. Every public field shows the source_ref directly — URLs are rendered as outbound links, free-form notes are rendered inline.

Every public record exposes last_updated_at. Index tables also expose the most recent declared-row timestamp per operator, method, and market.

Missing data is rendered as Unknown. PayOutIndex does not infer, estimate, or fill gaps with marketing copy. Withdrawals and deposit availability use a tri-state scheme: yes, no, unclear.

Incidents are recorded with a short title, type (delay, disabled, degraded, unknown), status (active or resolved), source, and timestamps. There is no severity scoring, no editorial interpretation, and no recommendation language.

Scope of incidents. An incident may target a specific operator + market + method combination, or it may be scoped to operator + market only (no method specified). In V1, the per-row incidents_active_count shown in the matrix counts method-specific incidents only. Operator/market-level incidents still appear in the operator and market detail pages and in /incidents; they simply do not bump the method-specific counter on the matrix. This trade-off is accepted for V1 to keep the counter unambiguous.

• No rankings.
• No scoring.
• No affiliate links.
• No CTA buttons to operators.
• No use of words like “best”, “fastest”, “trusted”, or “recommended”.
• No open-text complaint threads.
• No publication of unreviewed brand or user data.
• Declared data is not treated as ground truth.

Where applicable, declared rows carry a confidence label — low, medium, high, or unknown — and an optional sample size. PayOutIndex does not show formulas or pseudo-precision.