For every casino we cover, we show what we know, how we know it, when we checked it and what we could not verify. This page explains the full method behind Smarter Verified, our evidence levels, Data Confidence and the Smarter Score.
- Methodology version
- 1.6
- Locked
- 2026-10-10
- Principle
- Receipts, not reviews.
Principles
- What we know. Every fact on a casino profile has a value, an evidence level, a source and a check date. No source, no fact.
- How we know it. We say whether a fact comes from a regulator register, the casino's binding terms, the casino's own claims, a third party or player reports.
- When we checked it. Every fact has a date and a re-check interval.
- What we could not verify. Gaps are shown as gaps ("—", "Not researched", "Not published"). We never fill them in.
We have not deposited, played, completed KYC or withdrawn at the casinos we cover. "Verified" never means that we did. We only describe something as independently measured when a documented SmarterCasinos measurement exists for exactly that fact.
Missing information is never treated as positive. If a casino's terms are silent on a withdrawal limit, that is not "no limit". If we have not found complaints because a source was unavailable, that is not "no complaints".
Verification
Smarter Verified means that key casino information was independently checked by SmarterCasinos against identified sources. It is a statement about our process, not about the casino.
The 8 checks
| Check | What we check |
|---|---|
| Operator | The legal entity behind the casino. |
| Licence | The claimed licence in the regulator's official registry. |
| Exact domain | Whether the exact domain we review can be tied to that licence. |
| Terms | The full terms and conditions. |
| Payments | Methods, limits, fees and stated processing times. |
| KYC policy | The published identity-verification policy. |
| Bonus terms | The terms of the bonus programme (or confirmation that there is none). |
| Restricted countries | The casino's published list of restricted countries. |
Each check ends with one of these results: Confirmed (primary source), Claim recorded (only the casino's own statement), Finding recorded (a confirmed unfavourable fact, for example "domain match not confirmed" or "no KYC policy published"), Not applicable, Unresolved (an open question, conflicting or unavailable information) or Open (not checked yet).
Verification states
| State | Meaning |
|---|---|
| Verified | All 8 checks are complete, no material conflict is open and no check is past its re-check date plus grace period. |
| Pending | At least one check is open or unresolved, or a material conflict is open. |
| Expired | All checks were completed, but at least one is now past its re-check interval plus the 14-day grace period. |
| Not verified | Nothing has been checked yet. |
Verified means the verification process is complete with no open check. It never means "safe", "good", "trusted" or "recommended", and it is not a regulator certificate.
Confirmed negative findings do not block Verified. If we confirm that a domain is not in the licence registry, that check is complete; the finding is shown on the profile and affects the score, but the casino can still be Verified. What blocks Verified is an unanswered question, not an unfavourable answer.
Evidence levels
Every fact carries one evidence level. The level describes how a fact is known, not whether the casino is good.
| Level | Used when | Typical sources |
|---|---|---|
| Verified | We checked it against a primary or authoritative source. This includes the content of a casino's binding terms ("the terms state 30× wagering"). | Regulator registers and notices, company registers, binding terms, our own documented observation without gambling (for example an account settings page) |
| Casino says (official claim) | The casino says something about itself that we cannot confirm independently, especially about performance ("withdrawals within 1 hour"). | Casino pages, support statements, marketing |
| Third-party data | It comes from an identified independent source. | Datasets, blockchain explorers, credible media, a complaint platform's own published resolution |
| User reported | Players report it. These are signals, not facts. | Complaint platforms, forums, messages to us |
| Unknown | It could not be reliably determined. | — |
| Independently measured | Only with a documented SmarterCasinos measurement (date, amount, steps, receipt). | Our own measurement record |
- A claim is never upgraded to "verified" because it sounds plausible or because several casinos say the same thing.
- A user report becomes "verified" only when a primary source, such as a regulator, confirms it.
- Withdrawal times are kept in four separate channels (official claim, third-party, user reported, independently measured) and are never merged into one number.
Missing-data states
When a fact is missing, we say why.
| State | Meaning | Effect |
|---|---|---|
| Not researched | Nobody has looked yet. | Treated as missing: earns no points and counts against coverage. |
| Unknown | We looked but could not determine it. | Treated as missing: earns no points and counts against coverage. |
| Unavailable | The source could not be reached (blocked or offline). | Treated as missing: earns no points and counts against coverage. |
| Incomplete | Research started but was not completed under our method. | Treated as missing: earns no points and counts against coverage. |
| Not published | We looked and confirmed the operator does not disclose it. | Counts as assessed, earns 0 points. |
| Not applicable | It does not apply (for example, a confirmed absence of any bonus programme). | Left out; weight is redistributed. |
Data Confidence
Data Confidence tells you how strong the evidence behind a whole profile is: High, Medium or Limited. On casino pages it is shown as Research confidence: High, Medium or Low (Low means Limited). It is separate from the Smarter Score and says nothing about the casino's quality.
We look at 11 core facts, each with a fixed weight: operator 15, licence 15, domain 10, restricted countries 10, terms 10, KYC policy 10, withdrawal processing time 10, withdrawal limits 5, fees 5, bonus 5, complaint research 5 (total 100). Each fact contributes according to its evidence level (verified or independently measured 1.0, third-party 0.7, official claim 0.5, user reported 0.3, unknown 0) and its freshness (current 1.0, past its re-check date 0.8, past the grace period 0.5, never checked 0). A confirmed "not published" counts as a well-evidenced fact.
| Level | Requirement |
|---|---|
| High | Evidence strength of at least 80% and operator, licence and domain established in a primary source and not stale. |
| Medium | Evidence strength of at least 55%. |
| Limited | Below 55%. |
Data Confidence is limited to Medium at most when:
- any conflicting-information item is open or documented as unresolvable;
- complaint research is not Complete (see Reputation research);
- the licence is only listed in the issuing body's own registry (see Licensing).
Smarter Score
The Smarter Score rates the durable, documentable properties of a casino from 0 to 100, calculated from a published rubric. Our researchers record facts with evidence; the points are derived from those facts by fixed bands. Researchers do not choose points. A manual override is only possible with a written note and sign-off by a second reviewer, and it is labelled "Override" in the score explanation.
Categories and weights
| Category | Weight | What it covers |
|---|---|---|
| Safety & Licensing | 33% | Licence in the registry, regulator tier, operator identified, two-factor authentication |
| Withdrawals & Payments | 27% | Firm withdrawal-time statement, monthly withdrawal limit in the binding terms, casino-side fees |
| Terms & Fairness | 20% | Risky or unfair clauses (full-text review), dated and publicly accessible terms |
| KYC & Transparency | 13% | When verification is required, documented thresholds |
| Responsible Gambling | 7% | Limit tools, cooling-off and self-exclusion, help resources |
The weights keep the order and ratios of our earlier rubric for these five categories (25 · 20 · 15 · 10 · 5), rescaled to 100. Bonus offers and SmarterCasinos deals are not part of the score – they change often and can be commercial, so they are shown in their own section. Complaint-platform data is not part of the score either – it is shown in the separate Reputation section (see Reputation research). Abusive bonus or account clauses still count through the terms review.
Earned, documented, not verified yet
Every point on the rubric is in exactly one of three places, and together they always add up to 100:
- Earned from sources – points backed by evidence at or above the minimum level the criterion needs. This is the score. Some criteria accept the casino’s own statement as that minimum (responsible-gambling tools and resources, two-factor login, a stated processing time); points from such statements are earned but remain labelled “Casino says”, and the breakdown shows how many points come from casino statements.
- Documented findings – not earned because of something we documented: a risky clause, a regulator tier, a fee, or information the casino does not publish.
- Not verified yet – we do not hold the evidence yet (for example a page we could not read, or an individual item such as a limit type that we could not find or determine). It earns nothing until verified, and it is never presented as a problem of the casino.
Next to every score we show Evidence %: the share of the rubric, by weight, for which we hold evidence. Only verified evidence earns points. Because missing information can never earn points, less or weaker evidence can never raise a score – we test this automatically for every change (more than 2,300 mutations of the real records).
When a casino gets a number
A score is shown when the fundamentals are in place (the Hard Safety Gate):
- The operator, licence, domain and terms checks are complete and not overdue for re-checking (a confirmed negative finding is allowed – it caps the score instead).
- The full binding terms have been reviewed for risky clauses.
- The regulator's records and public records have been searched, so that an official negative finding cannot be hidden.
- No official negative finding is under review (an unverified regulatory finding or reported licence revocation blocks the score until it is verified or dismissed).
Otherwise the profile shows "Score not yet available" with the reasons. Information still missing elsewhere does not block the score; it simply earns no points and lowers Evidence %. A criterion affected by an unresolved material contradiction earns no points until it is resolved.
Withdrawal-time statements
We record every statement a casino makes about withdrawal times, word for word, with its source and type: guaranteed (a firm maximum), typical, majority ("most …"), percentile ("90% …"), target ("we aim / try …") or conditional ("with few exceptions", "in some cases longer"). The governing statement is taken from the highest-ranked document that makes one: binding terms, then a payments table, then the help centre, then marketing pages. Between documents of the same rank, the more cautious statement governs. Only a governing guaranteed statement counts as a firm maximum – and only once every higher-ranked document has been reviewed. General KYC, anti-money-laundering and security reservations are not time statements; they are assessed under KYC and the terms review.
Claim ceilings
Where a criterion can only be supported by the casino's own statement, an official claim earns at most part of the points:
- Withdrawal processing time: a firm statement by the casino earns at most 60% of the criterion. A third-party or independently measured median has no ceiling.
- Two-factor authentication: an official claim earns at most 67% of the criterion; our own documented observation (for example, creating a free account without depositing) earns the full points.
Why two casinos score differently – real examples
| Casino | Score | Safety & licence | Withdrawals & payments | Terms & fairness | KYC & transparency | Player protection | Evidence |
|---|---|---|---|---|---|---|---|
| Mirax Casino | 65 | 25/33 | 11/27 | 13/20 | 10/13 | 5/7 | 90% |
| Betway | 49 | 21/33 | 0/27 | 18/20 | 10/13 | 5/7 | 83% |
| BC.Game | 29 | 8/33 | 0/27 | 13/20 | 3/13 | 4/7 | 89% |
Points earned out of the category weight, from the published records as of 12 Oct 2026. The highest score today is 65; nobody gets points for what we cannot verify. Where a regulator cap applies, the score is the lower of the points earned and the cap.
Why Mirax Casino (65) scores higher than BC.Game (29)
- Safety & licence: Mirax Casino earns 25 of 33 points, BC.Game earns 8 – most of the difference is documented findings – what the casino’s own documents say or leave out about licence status, regulator tier, operator identification or two-factor evidence.
- Withdrawals & payments: Mirax Casino earns 11 of 27 points, BC.Game earns 0 – most of the difference is not verified yet (withdrawal-time statements, limits or fees).
Why Betway (49) sits in the middle
- Withdrawals & payments: Mirax Casino earns 11 of 27 points, Betway earns 0 – most of the difference is not verified yet (withdrawal-time statements, limits or fees).
- Safety & licence: Mirax Casino earns 25 of 33 points, Betway earns 21 – most of the difference is documented findings – what the casino’s own documents say or leave out about licence status, regulator tier, operator identification or two-factor evidence.
What the number is, and what it is not
The Smarter Score is the share of rubric points a casino earns from the facts we could establish, after caps. A score of 42 means the casino earned 42 out of 100 points on our published rubric, based on the evidence on its profile on the check date.
- It is not a probability: 42 does not mean a 42% chance of being paid or of anything else.
- It is not "% safe", and no score means "safe".
- It is not a recommendation and not a guarantee.
- It is not for sale: commercial data cannot affect it (see Commercial separation).
Caps and disqualifiers
Some problems cannot be outweighed by good bonuses, games or support. When one of these conditions applies, the final score cannot exceed the cap. If several apply, the lowest cap wins. Every applicable cap is shown with its reason.
| Maximum score | Applies when |
|---|---|
| 59 | The casino claims no licence. |
| 59 | The claimed licence could not be confirmed in the registry. |
| 59 | The licence was found, but the reviewed domain could not be matched to it ("Domain match not confirmed"). |
| 59 | The licensing regulator is in Tier D. |
| 59 | The terms contain a verified critical clause (see Terms review). |
| 49 | A credible, ongoing pattern of unresolved payout problems is confirmed by an authoritative source: a regulator or ADR decision, a court or official record, or the operator's own written confirmation. |
| 39 / 49 / 59 | Verified official regulatory actions (warning, fine, suspension, revocation) – graded, see below. |
Regulatory history
Only verified actions count. A decision that was later annulled stays visible on the profile but does not count. The cap depends on how recent and how frequent the actions are:
| Maximum score | Applies when |
|---|---|
| 39 | An action is younger than 24 months, or an action concerns offering gambling in a market without the required licence (this does not age out). |
| 49 | Two or more actions within the last 60 months. |
| 59 | One action within the last 60 months. |
| no cap | All actions are older than 60 months. |
Independently of the cap, every counted action (at most two) lowers the Safety & Licensing score by 10 points – permanently, so a score never falls only because time passes, and an old action stays visible in the number. The cap date is shown on the profile.
User accusations alone never trigger a cap. A payout pattern reported only on complaint platforms does not cap the score; it is shown prominently in the Reputation section and in Good to know, and it blocks sign-up links and deals until it is clarified. Gaps in our own research never cap a score.
Not recommended
A casino is shown as Not recommended, with no number, when:
- an authority has made a verified finding of severe fraud or manipulation; or
- the official registry shows the casino's licence as revoked or suspended.
The profile stays online with its facts. We never show sign-up links for a casino marked Not recommended.
Reputation research
Player complaints are signals, not facts. We write "players have reported …", never "the casino doesn't pay".
Source types
Our method is defined by types of sources, not by specific websites. If a website closes, changes access or becomes unreliable, we replace it, or the research level drops. The method itself does not depend on any single site.
| Type | What it covers |
|---|---|
| T1 Regulatory, ADR and dispute bodies | Decisions, warnings and sanctions of the licensing regulator; decisions of dispute-resolution bodies named in the casino's terms or recognised by the regulator. |
| T2 Public records and credible media | Court and insolvency records, official notices, credible industry and general media. |
| T3 Independent complaint platforms | Platforms with a mediation process, public case history, case IDs and fixed dates, and no ownership link to the casino. |
Forums, social media and review sites are signals only and never count toward a cap on their own. Anonymous posts without dates or details, competitor or affiliate content, and AI-generated or search-result summaries are not used as complaint evidence.
A complaint qualifies when it is identifiable (case ID or link), specific, within the last 24 months, not a duplicate, about the reviewed brand or domain, and not obviously caused by the player breaking the rules.
Research levels
| Level | Requirement | Shown as |
|---|---|---|
| Complete | T1 searched (or not applicable), T2 searched, at least 2 accessible independent T3 platforms searched for the full 24-month window, and at least 12 months of documented operating history. | Complete data |
| Partial | T1 and T2 searched and at least 1 accessible T3 platform searched for the full window, but Complete not reached. | Limited data |
| Insufficient | Anything less, including research not done. | Insufficient data – and no Smarter Score, because T1/T2 are part of the Hard Safety Gate |
Reputation is separate from the score
- Complaint-platform data never changes the Smarter Score – neither positively nor negatively. A platform we cannot access is a gap in our research, never a finding about the casino.
- Authoritative negatives always count: regulatory warnings, fines or suspensions (graded cap 39 / 49 / 59), confirmed payout problems (cap 49), fraud findings or a revoked licence (Not recommended).
- Credible payout patterns reported on complaint platforms are shown prominently and block sign-up links and deals until clarified.
- With Limited data, not finding complaints earns nothing in the reputation assessment, and operator responsiveness is not credited – more known complaints can never improve the result.
- Data Confidence is at most Medium, no "Best for" label is shown, and the casino is listed after casinos with Complete research in the same verification state.
The 12-month history rule
A casino with less than 12 months of documented operating history, or with unknown history, can reach Partial at most, whatever sources were searched. A short history is not evidence of good conduct, and unknown history is never treated as positive. These rules ensure that less available information can never improve a casino's result.
Licensing
Regulator tiers
We rate each licensing regulator, not the casino, on five factors, each scored 0–4 from observable conditions:
The outcome of this check for every profiled casino is published as an open dataset: licence register check (holder, register status, domain match, finding; CSV/JSON). The withdrawal clauses we read are published the same way: withdrawal limits.
| Factor | Weight | Question |
|---|---|---|
| Verifiability | 25 | Can anyone confirm the licensee, licence number and status in a public, official register? |
| Domain attribution | 20 | Can a specific website be tied to the licence from official records? |
| Oversight & enforcement | 25 | Is there real supervision, with published sanctions, suspensions or revocations? |
| Player protection & recourse | 20 | AML and KYC duties, responsible-gambling duties, protection of player funds, testing, a complaint or dispute route. |
| Transparency | 10 | Published law, decisions, reports and accessible records. |
| Tier | Quality score | Additional condition | Meaning | Points (licence tier criterion, max 30) |
|---|---|---|---|---|
| A | 80 or more | Verifiability 4 and oversight at least 3 | Strong regulatory protection | 30 |
| B | 60–79 | Verifiability at least 3 | Established, verifiable regulation | 22 |
| C | 35–59 | Verifiability at least 2 | Limited regulatory protection | 10 |
| D | Below 35, or verifiability 1 or less | – | Unlicensed or not verifiable under our method | 0 (and the score is capped at 59) |
A tier describes the strength and verifiability of a regulatory framework. It is not a rating of any individual casino: a well-run casino can hold a Tier C licence, and a Tier A licence does not make a casino good. Tiers are researched from primary sources, dated, approved before use and reassessed every 12 months or after a regulatory reform.
How we check a licence
We check the licence in the regulator's own registry or tool, for the exact domain we review. A seal image on the casino's website is never enough; it must resolve to the regulator's own domain. Licences issued for one national market are checked for the reviewed domain and do not automatically cover a different, global website.
Licence states
| State | What it means |
|---|---|
| Confirmed in registry | The licence and the reviewed domain were confirmed in the regulator's official registry. |
| Domain match not confirmed | We found the licence, but could not tie the exact reviewed domain to it in official records. This describes what we observed; it is not an accusation. |
| Listed in issuer registry | The casino or domain appears in a registry presented by the issuing body itself. It does not mean we consider that framework independently verified or equivalent to higher tiers. |
| Licence claim not confirmed | We could not confirm the claimed licence in the registry. This is not an accusation of fraud. |
| Verification pending | The licence check has not been completed yet. |
| Conflicting information | Official sources disagree (for example, a certificate and a register show different statuses). The licence check stays unresolved until investigated. |
| Information unavailable | The registry was offline or could not be reached. This is treated as pending, never as "not confirmed". |
Terms review
We read the full terms and judge every clause in its full context, never as an isolated quote. Only a fixed list of 11 flags affects the score; each flag has a fixed severity that researchers cannot change. A clean result counts only after a documented full-text review.
| Severity | Flags | Effect on the terms criterion (max 70) |
|---|---|---|
| Caution |
|
−10 points each |
| Serious |
|
−25 points each |
| Critical |
|
0 points, and the total score is capped at 59 |
The remaining 30 points of the category reward clarity: 10 each for dated or versioned terms, terms readable without an account, and bonus terms that state wagering, maximum bet, maximum cashout and expiry.
Other noteworthy clauses are recorded as observations. They are shown on the profile but do not affect the score. New scoring flags can only be added in a new methodology version.
Countries and availability
We record which countries a casino says it accepts, which it says it restricts, and which are restricted by its licence conditions.
- A country restricted by the casino's terms or its licence is always shown as restricted.
- A country counts as accepted only if the casino lists it explicitly, or states that it accepts all countries except a published restricted list.
- Everything else is shown as unclear. We never infer acceptance from absence on a restricted list.
Availability affects only commercial elements, never facts, scores or verification, which are the same for every visitor. We do not show sign-up links when availability for your country is not confirmed, when your country is unknown, or in markets where we do not promote casinos. We never explain how to get around country restrictions.
Freshness
Facts go out of date. Each type of fact has a re-check interval. After the interval a fact is due; after a further 14-day grace period it is stale, and a Verified casino with a stale check becomes Expired.
| Fact type | Re-check interval |
|---|---|
| Bonus terms | 14 days |
| Licence, domain match, regulatory actions, complaints | 30 days |
| Restricted countries, terms, payments, withdrawal processing time | 60 days |
| KYC policy | 90 days |
| Operator, responsible gambling, account security | 180 days |
Every profile shows when each fact was last checked and when the next check is due.
Commercial separation
- At launch there are no affiliate links on SmarterCasinos. If commercial offers are added later, they are labelled as advertising.
- Commissions, CPA, revenue share, sponsorship and deal size never influence the Smarter Score, Smarter Verified, Data Confidence, evidence levels, watch-outs or rankings.
- This is enforced technically: commercial data lives in a separate store, and the scoring code rejects any casino record that contains commercial fields. It cannot read them.
- No paid rankings. No casino can buy a position, a label or a score.
- Even when commercial offers exist, we will not show sign-up links for a casino that is Not recommended or whose licence has not been confirmed in an official registry.
Changes
The methodology is locked. Weights, scoring bands, regulator tiers, caps, evidence requirements, confidence thresholds and complaint rules are not changed because a casino gets a surprising result. A genuine flaw is documented, its effect is shown across all affected casinos, and the fix is released as a new version. Earlier versions are never modified silently. How facts are sourced and corrected on the profiles is described in our editorial policy.
| Version | Date | Summary |
|---|---|---|
| 1.0 | 2026-10-05 | First locked methodology: regulator tiers, 22-criterion scoring rubric, caps, dual review, complaint research protocol and country model. |
| 1.1 | 2026-10-05 | Integrity fixes: missing information can never earn points; new licence state "Listed in issuer registry"; fixed list of 11 terms flags; percentile claims never treated as maximums; minimum evidence levels enforced. |
| 1.2 | 2026-10-05 | Complaint research graded Complete / Partial / Insufficient by source type instead of specific websites; Partial research allows a score capped at 49 with "Limited reputation data"; operating history under 12 months can never be Complete. |
| 1.4 | 2026-10-06 | Smarter Score redesign: five durable categories (33 · 27 · 20 · 13 · 7); bonus offers and complaint-platform data shown separately and never scored; Hard Safety Gate instead of coverage thresholds; Evidence % and the earned / documented / not-verified breakdown; structured withdrawal-time statements with a document ranking; the "limited reputation data" cap removed; caps only from authoritative evidence; operator-response credit fixed so that more complaints can never help. |
| 1.5 | 2026-10-07 | Individual responsible-gambling items recorded as “not found / not determined” are booked as Not verified yet instead of a documented finding, and no longer count as evidence. Scores are unchanged; only the breakdown buckets and Evidence % move slightly. Owner decision BC1. |
| 1.3 | 2026-10-06 | Withdrawal times: majority statements, non-binding targets and times with unquantified exceptions are never treated as an upper bound and earn no processing-time points (applied automatically from the wording). |