Our testing methodology

Overview and philosophy

The methodology described on this page has evolved over three years and eight testing quarters. It rests on a single premise: a casino recommendation is only useful to a reader if the casino has been tested in the way the reader will actually use it. That means real deposits from real bank accounts, real bonus terms accepted at face value, real support conversations conducted in the reader's language, and real withdrawals timed with a stopwatch from click to arrival on the destination account.

We test five pillars because casino experience is compositional: an operator that scores high on four pillars but fails on the fifth will hurt a reader precisely when the failing pillar becomes the deciding one. A casino with excellent games but slow withdrawals is a bad casino for a reader who wants to withdraw a win. A casino with fast withdrawals but hostile support is a bad casino for a reader whose first bonus does not credit correctly. Averaging the pillars into a single number hides these compositional failures, so we track and publish each pillar separately, then use the composite score for ranking order only.

An operator that scores below 7.5 on any single pillar does not appear on our recommendation lists, regardless of how high the other four pillars score. This is the rule that produces our shorter ranking lists relative to competitor publications, and we consider it non-negotiable.

Pillar 1: Licence verification

Every casino we cover has been checked against the register of the regulator whose licence the operator claims. The check involves finding the licence number in the regulator's public register, confirming the date of issue, confirming the current status is active, and confirming that the entity name on the register matches the entity operating the casino brand. This last check is the one that catches the most operator claims: a licence held by a master-licence holder that has since transferred is not a licence held by the current operator.

For Curaçao licences issued from 2024 onward, we verify against the Curaçao Gaming Control Board register at gaming-curacao.com. For Malta MGA licences we verify against the MGA operator register. For UK Gambling Commission licences we verify at gamblingcommission.gov.uk. For French ANJ licences we verify against the ANJ operator register. For each other jurisdiction our published pages cover, the specific register URL is documented in our internal reference and available on request.

An operator whose licence cannot be verified in the regulator's own register does not appear on 8day.gg, regardless of how prominently the licence is displayed on the operator's website. Aisha Halima personally completes this check for every new operator entering our coverage.

Pillar 2: Deposit testing

Every operator on 8day.gg has received real deposits from us via a minimum of two payment methods. For markets where multiple methods are common we test three or more. The methods tested vary by market: for France we test carte bancaire, Paylib (where supported), and USDT on the Tron network. For the Netherlands we test iDEAL and Bancontact. For Denmark we test MobilePay and Trustly. For Thailand we test PromptPay, TrueMoney Wallet, and USDT.

Each deposit is timed from the point of confirmation on the payment provider side to the point of funds appearing in the casino account balance. This measurement is repeated three times on separate days to average out network conditions. The published deposit time is the median of the three baseline measurements, not the fastest and not the operator's advertised time.

Deposits that fail or are declined are documented but not used as a scoring signal in themselves, because payment declines can originate at the issuing bank rather than the operator. Where declines occur, we retry via alternative methods and document the pattern. Repeated declines across methods for a specific market are a scoring signal against the operator.

Pillar 3: Bonus work-through

The welcome bonus of every operator on 8day.gg has either been worked through to completion, worked through to abandonment with the abandonment reason documented, or refused at the deposit stage with the refusal reason documented. This is the most time-consuming part of the test cycle because welcome bonuses at reasonable wagering requirements take 30 to 90 hours of active play to complete.

During bonus work-through, we track four separate measurements: the effective wagering requirement multiplier, the maximum bet permitted during work-through, the game contribution percentages, and the elapsed time to complete. Each of these can turn a nominally generous bonus into a mathematically unattainable one, and we publish each separately rather than aggregating them into a single "bonus quality" score.

When a bonus term reads reasonably on the terms page but produces an unexpected behaviour during work-through (such as a bet being flagged as non-contributing when the terms did not exclude that game), the incident is documented and forms part of the operator's scoring on this pillar. Repeated incidents of this type across the sample size are a strong signal against the operator.

Pillar 4: Support timing

Every operator has been contacted through their published live chat interface at least three times during the test cycle, at varying times of day and days of the week. The three interactions cover a simple question (typically about a promotion or a payment method), a moderately complex question (typically about bonus terms or KYC requirements), and a complex question (typically about a fictitious dispute or edge case).

Each interaction is timed from the moment the chat is initiated to the moment a substantive first response arrives (not the automated acknowledgement). We record whether the responder is a bot or human, the language quality, the accuracy of the answer, and whether the complex question required escalation. The published median support response time is the median across the three timings.

For markets where language support is a claimed feature (Thai support at FinnBet, French support at Kyngs, Dutch support at DiceSpin), we specifically test the availability and quality of native-language responses. Automated translation delivered as native support is documented as such and treated as a scoring negative on this pillar.

Pillar 5: Withdrawal chronometry

Every operator on 8day.gg has processed a real withdrawal for us, timed from the moment the withdrawal is initiated in the casino cashier to the moment the funds appear on the destination account. This is the single most consequential measurement in the entire methodology because withdrawal experience is where operator quality most directly affects the reader.

Withdrawals are tested with amounts ranging from small (equivalent to €50) to substantial (equivalent to €5,000) to characterise whether the operator processes larger withdrawals differently. Each withdrawal is timed separately from KYC review (if triggered), from operator internal review, from payment provider processing, and from final arrival. The published withdrawal time is the median across at least three baseline measurements.

Withdrawals held for reasons that we consider legitimate under the operator's published terms (documented KYC review, first-withdrawal enhanced review) are not treated as pillar failures. Withdrawals held for reasons not documented in the operator's terms (repeated document requests after prior approval, unexplained delays past the operator's published cut-off) are pillar failures.

Score computation

Each of the five pillars receives a score from 0.0 to 10.0 based on the tests described above. The composite score published on ranking pages is a weighted average of the five pillars, with weights that we have kept constant since 2024: licence 15%, deposit 15%, bonus 20%, support 20%, withdrawal 30%. Withdrawal is weighted highest because it is the pillar most closely correlated with reader outcomes.

An operator that scores below 7.5 on any single pillar is excluded from ranking pages regardless of the composite score. An operator that scores between 7.5 and 8.0 on multiple pillars appears with a lower ranking position. An operator that scores above 9.0 on all five pillars is rare and typically becomes an editor's pick on the relevant hub.

Scores are recalculated every quarter as we re-test operators. Movements up or down between quarters reflect actual changes in operator performance, not editorial reweighting.

Publication cadence and updates

Hub pages are updated quarterly with new test results. Sub-pages within a market are updated on the same quarterly cycle. Individual review pages for operators are updated whenever a material change is detected: a payment method added or removed, a licence status changed, a bonus term revised, or a significant complaint pattern emerging in public dispute resolution channels.

Each updated page carries a "last updated" date visible in the author block. Corrections resulting from reader tips, factual disputes, or our own error detection are documented in a changelog visible at the bottom of the updated page.

Special protocols

Three verticals require testing protocols that differ from the standard five-pillar cycle.

Crypto casinos receive additional testing on payment rail confirmation counts, network fee accuracy, address family compatibility, and settlement latency across chain congestion conditions. Raymond Delacroix owns this protocol.

Live dealer coverage requires testing of table availability at published claimed hours, dealer language quality, video and audio quality across desktop and mobile, and stake ranges at claimed VIP tables. This is the primary supplementary test cycle for pages such as France live casino.

Lottery and lottery-adjacent products require testing of result verification against upstream sources (Thai Government Lottery Office for สลากกินแบ่งรัฐบาล, stock exchange data for หวยหุ้น, Lao Lottery Office for หวยลาว, and official lottery operators for EuroMillions, Powerball, MegaMillions). Aisha Halima owns this protocol.