How We Rate: Scoring Bands and Test Protocol
Most ratings in this industry are a number with nothing behind it. Ours is arithmetic on published thresholds, so the whole calculation can be reconstructed by a reader who disagrees with it.
This page is the operational detail. For what we test and whether we have done it yet, see the testing process. For editorial policy and the claims we refuse to publish, see the editorial guidelines.
How a Score Is Calculated
Seven categories, each scored out of 10 against the bands below, each multiplied by its weight, summed to a rating out of 10 and published to one decimal place.
(Licensing × 0.25) + (Payout × 0.20) + (Bonus × 0.15) + (Rails × 0.15) + (Games × 0.10) + (Mobile × 0.10) + (Support × 0.05)
Weights sum to 1.00. No rounding is applied to a category score before weighting.
Two rules constrain when the arithmetic may be run at all. At least four of the seven tests must be complete, and the licence must be verified on a regulator’s register. Where fewer than four tests are done, we publish no score rather than a partial one.
The Real-Money Deposit Test
Every test begins with real money from a personal account. Nothing is comped, and no operator is told a test is happening.
1. Register and fund
An account is opened under the tester’s own identity, using the market’s typical device rather than a desktop, and funded with a modest amount that a genuine first-time depositor would use.
2. Time the credit
The clock runs from the moment the payment provider confirms the transaction to the moment the balance updates on the site. That gap is the operator’s integration, not the payment rail.
3. Record the mechanics
Whether the cashier is a direct integration or a redirect through a third-party wallet, what authentication step was required, and what each side charged.
Recorded for every deposit test
- Payment method and the exact amount deposited
- Date, time and the tester’s name
- Device, operating system and browser
- Authentication step required, such as a PIN, an app approval or a one-time code
- Time from payment confirmation to balance update
- Fee charged by the operator, and fee charged by the payment provider, recorded separately
- Whether the integration is direct or redirects through a third party
- Whether the same method is available for withdrawal
| Score | Deposit score band |
| 10 | Credited in under a minute, no operator fee, direct integration, and the same method is available for withdrawal |
| 8 | Credited within five minutes, no operator fee, withdrawal available on the same method |
| 6 | Credited within thirty minutes, or a small disclosed operator fee |
| 4 | Credited within a few hours, or the method cannot be used for withdrawal |
| 2 | Manual review before credit, an undisclosed fee, or a failed first attempt that later succeeded |
| 0 | Deposit failed and was not recovered, or funds left the account without crediting |
The last row of the field list carries more weight than it looks. A deposit method that cannot receive a withdrawal is how money gets stranded: the reader funds an account by card or voucher, wins, and then discovers the exit runs on a rail they have not set up.
KYC Verification Checks
Identity verification is a legal requirement, not an obstruction, so we do not penalise its existence. We measure when it happens and how cleanly it runs, because those are the operator’s choices.
Documents are submitted on registration day, before depositing. Testing verification this way rather than at the point of withdrawal isolates it from the payout clock and reveals the operator’s actual policy.
Recorded for every KYC check
- Date and time documents were submitted, and by which route
- Exactly which documents were requested
- Time to a verified status
- Whether anything was re-requested, and whether the re-request was specific or generic
- Whether verification was demanded at registration or only triggered by a withdrawal
- Whether source-of-funds evidence was required, and at what threshold
- Whether the account was usable while verification was pending
| Score | KYC score band |
| 10 | Verified in under an hour, first submission accepted, requested at registration |
| 8 | Verified the same day, no re-requests |
| 6 | Verified within two days, no re-requests |
| 4 | Verified within two days but with one re-request |
| 3 | Three to five business days |
| 0–1 | Beyond five business days, or repeated generic re-requests with no stated reason |
Two Penalties That Apply Regardless of the Clock
An operator that only demands identity documents once a withdrawal is requested is capped at 5 for this test; however quickly it then moves. Verifying after a win rather than before a deposit is a structural choice, and it is the most common reason a payout that looks fast on paper takes a week. Separately, a document re-requested after it was already supplied and accepted costs two points, because in practice that is where payouts stall.
Withdrawal Time Tracking
This is the measurement the rest of the rating is arranged around, and the one most often reported in a way that hides the answer.
A single withdrawal figure conflates two different things. Stage one is operator processing: approval, the pending window, and any compliance hold. Stage two is the payment rail moving the money. The operator controls the first entirely and the second not at all, so we time them separately and score them separately.
| Stage | Clock starts | Clock stops | Controlled by |
| One – operator processing | Withdrawal submitted | Marked approved and sent to the rail | The operator |
| Two – payment rail | Sent to the rail | Funds usable in the destination account | The rail |
Recorded for every withdrawal test
- Timestamp the withdrawal was submitted, to the minute
- Timestamp the operator marked it approved
- Timestamp the funds became usable in the destination account
- Stage one duration, stage two duration, and the total, all stated separately
- Method and exact amount
- Whether any document was requested at this point, and whether it had been supplied before
- Any fee deducted, and any tax withheld
- Whether the withdrawal was reversible or cancellable, and whether the account nudged reversal
- Whether the amount received matched the amount requested
| Score | Stage one: operator processing |
| 10 | Approved in under an hour |
| 8 | Approved within six hours |
| 6 | Approved within twenty-four hours |
| 4 | Approved within two days |
| 2 | Approved within three days |
| 0 | Beyond three days, or the payout required chasing to complete |
Stage two is not scored against the operator, because the settlement speed of a bank transfer is not something an operator can change. What we do score is the choice of rails offered: an operator that supports an instant-settling withdrawal rail in its market is doing something an operator that only pays back to a card is not.
| Score | Stage two: rail availability |
| 10 | At least one instant-settling withdrawal rail offered, matching the deposit method |
| 7 | Instant rail offered but not on the method used to deposit |
| 4 | Only same-day or next-day rails available |
| 2 | Card or bank transfer only, with multi-day settlement |
| 0 | No withdrawal rail available in the market the operator advertises to |
The payout velocity category combines the two, weighted 70 percent to operator processing and 30 percent to rail availability. Two further adjustments apply to the combined figure: a reversible-withdrawal design that invites you to cancel a pending payout costs one point, and an amount received that does not match the amount requested, after disclosed fees and tax, costs three.
Why the split changes the answer
- An operator approving in twenty minutes but paying only to a card is slower to your money than one approving in three hours onto an instant rail.
- Reported as one blended number, the first operator looks better while the second one pays you sooner.
- Splitting the stages is also the only way to tell a slow operator from a slow rail, which is the thing a reader actually needs to know before choosing a deposit method.
The Remaining Four Bands
The other categories are scored against the same kind of published thresholds, summarised here.
| Category | Scores 8–10 | Scores 0–3 |
| Licensing and safety | Licence number resolves on the regulator’s own register, operating company named, parent ownership traceable, dispute route stated | No number published, no company named, or a number that does not resolve where the regulator maintains a public register |
| Bonus integrity | Wagering base stated explicitly, multiplier at or under 35× bonus-only, no maximum cashout on deposit-match winnings, max bet stated in the offer itself | No wagering base published anywhere, or a maximum cashout that caps winnings below the advertised bonus value |
| Payment rails | Multiple local rails, withdrawal available on the deposit method, fees published, limits proportionate to the market | Advertised methods absent from the cashier, or a minimum withdrawal set well above the minimum deposit |
| Game engineering | Default RTP configurations, providers confirmed as genuine integrations, RNG certified by a recognised laboratory | Downgraded RTP variants without disclosure, provider logos with no working games behind them, no certification evidence |
| Mobile and UX | Cashier completes a deposit in a mobile browser, reference codes copy correctly, no layout failure blocks a withdrawal | Any failure that prevents completing a deposit or withdrawal on a phone |
| Support efficiency | Human contact within five minutes out of hours, accurate on bonus terms, competent on responsible gambling | No human contact reached, or a materially wrong answer about terms or responsible gambling tools |
Support is graded hardest on the responsible gambling question rather than the payments one. An agent who cannot explain how to set a deposit limit is telling you what the operator has trained for.
Gates That Void the Score
Three conditions replace the arithmetic rather than adjusting it, and no combination of high category scores overrides them.
| Condition | Effect |
| Licence not verifiable on the regulator’s own register | No score, and the operator is not ranked at all |
| Withdrawal test not completed | Score capped at 6.0, with the cap and the reason stated in the review |
| Documented pattern of voided withdrawals | No score, removed from ranking, and considered for an investigation instead |
| Fewer than four of seven tests completed | No score published |
A Worked Example
Numbers below are illustrative, to show the arithmetic rather than to describe any operator.
| Category | Score | Weight | Contribution |
| Licensing and safety | 9 | 0.25 | 2.25 |
| Payout velocity | 7 | 0.20 | 1.40 |
| Bonus integrity | 4 | 0.15 | 0.60 |
| Payment rails | 8 | 0.15 | 1.20 |
| Game engineering | 7 | 0.10 | 0.70 |
| Mobile and UX | 6 | 0.10 | 0.60 |
| Support efficiency | 5 | 0.05 | 0.25 |
| Final rating | – | 1.00 | 7.0 |
The interesting part of that example is the bonus score of 4 dragging a well-licensed, fast-paying operator down by nearly a point. That is the weighting working as intended: an operator that will not publish its wagering base is withholding the number that decides what its headline offer is worth.
What a Score Does Not Mean
- It is not a prediction about your session. Nothing on a review page changes the mathematical margin built into the games.
- It is not a safety guarantee. It records what we verified on the dates we verified it, and operators change processors, terms and ownership.
- It is not transferable between markets. The same brand can score differently in two countries because rails, licences and terms differ.
- It has a shelf life. Any rating older than six months is re-tested before it is cited again, and a stale score is withdrawn rather than carried forward.
- It is not a recommendation to gamble. If you are looking for a reason to stop, our responsible gambling page is more useful than any rating here.
What We Have Scored So Far
Nothing, yet. The bands on this page are the standard we will apply, and our testing ledger currently holds no completed payout timings, so no page on this site publishes a score.
Publishing the thresholds before the results is deliberate. It means the standard cannot be adjusted after the fact to suit an operator we happen to earn commission from, and it means the first scores we publish can be checked against a rubric that was fixed before the tests were run.
Current testing status, including which operators have been assessed from published material only, is on the testing process page.
Category weights
| Licensing & safety | 25% |
| Payout velocity | 20% |
| Bonus integrity | 15% |
| Payment rails | 15% |
| Game engineering | 10% |
| Mobile & UX | 10% |
| Support efficiency | 5% |
Each is scored out of 10, then weighted. Three gates can void the total.