Non GamStop casinos compared on 22 measurable dimensions

What we compare and how we compare it

Meridian Bets Journal is a comparison publication. Every page you read on this site is built from the same underlying dataset, a working spreadsheet in which we hold twenty-two columns for each of the operators we currently track. This section explains what that dataset holds, what it does not, and why the framing matters when you read a matrix that puts one operator next to another.

Why comparison is the only honest editorial mode

The non-GamStop sector is a working ecosystem of roughly a hundred customer-facing operator brands, most of which share back-end infrastructure with a smaller set of platform providers. Writing about any one of them in isolation is a mug's game, because features that read as generous in a single-brand review very often turn out to be median or below-median once you compare them across a peer set. Editorial framing without a peer benchmark is not editorial, it is copy. We built the twenty-two-column dataset precisely so that every claim in our copy has a comparator behind it.

When we say a welcome bonus is generous, we mean it sits in the top quartile of the twenty-two brands we scored in our sample. When we say a withdrawal SLA is slow, we mean it sits below the sample median. Aisha Rowland, who runs the desk, worked at the Financial Times' data team for four years before Meridian, and the editorial ethic she brought across is a simple one. If you cannot state your comparator, you cannot state your claim. The rest of this homepage walks through what our comparators are and how we treat them.

The twenty-two operators in our current sample

Our current working sample is twenty-two live brands that accept UK residents outside the GamStop database. Selection was not random. We started from a longlist of eighty-plus brands surfaced through search, then filtered on three practical constraints. Brands had to hold a live and verifiable licence with one of the six regulators we recognise on this site. Brands had to accept GBP as a deposit currency, either natively or via a wallet layer. Brands had to have a customer-facing English-language site that scored above thirty on a basic readability floor.

The filter cut us to thirty-one. We then scored the survivors on the twenty-two dimensions that follow, and set aside a further nine brands that failed at least one integrity check during the scoring pass. That left twenty-two brands in the working sample. The sample is refreshed roughly every six weeks. When we publish an updated matrix, the composition of the sample is disclosed on the money page. This is not a ranked list of recommendations. It is a data view.

What our comparison tables never claim

A comparison table can only ever compare the columns you have chosen to compare. We are transparent about this. Our tables do not claim to score the aesthetic quality of an operator's slot lobby. They do not score how much fun it is to play at a given brand. They do not score whether a given operator is safer than another in any legally protective sense, because operators outside the UKGC perimeter carry different consumer protections and those protections are not fungible across jurisdictions. What our tables do claim, and this is the harder and more useful claim, is that on the twenty-two dimensions we scored, here is where each operator sat on the day we measured it. That is a data claim. It is not a recommendation. If you want a recommendation, we do not publish one. If you want a decision framework, that is what the matrices are for.

The twenty-two dimension scoring rubric

The heart of Meridian's editorial approach is a scoring rubric of twenty-two dimensions, each of which returns a normalised zero-to-five score for every operator in our sample. This section explains what those twenty-two dimensions are, why each one carries the weight it carries, and how we resolve the inevitable trade-offs when a brand ranks strongly on one dimension and weakly on another. Aisha Rowland rebuilt the weighting scheme in March of this year after feedback that the earlier version over-indexed on welcome bonuses.

The twelve structural dimensions

Twelve of our twenty-two dimensions are what we call structural, in that they measure something reasonably stable about how an operator is set up. These are licence tier, welcome bonus size, wagering multiplier, game count, live dealer provider count, withdrawal SLA in hours, KYC threshold in GBP, minimum deposit, minimum withdrawal, deposit method count, withdrawal method count, and the presence of a formal VIP scheme. Structural dimensions are the ones we can measure by looking at the operator's own terms and infrastructure without needing to test the site. Because they are stable, they carry the largest share of the composite score. Together the twelve structural dimensions account for sixty percent of the weighted total.

The seven experiential dimensions

Seven further dimensions measure the operator's user-facing behaviour, and these we call experiential. They include mobile UX score (measured against a rubric of nine sub-checks), support hours per week, first-response time in minutes on live chat during the working day, terms-and-conditions readability, complaint route (do they belong to a recognised ADR provider), responsible gambling toolkit completeness, and average operator lifespan under the current licensing entity. Experiential dimensions carry the next tranche of weight, twenty-eight percent, because although they matter enormously to the practical experience of playing at a site, they are harder to measure reproducibly. We time-box our checks and disclose the date of measurement.

Aisha Rowland has written elsewhere on the desk about how difficult support-hours scoring becomes when an operator's advertised hours diverge from its actual staffing. We measure both. If a brand advertises 24/7 live chat and our test contacts on a Tuesday at 03:00 UTC hit a bot with a ninety-minute human handoff, we score the actual not the advertised.

The three commercial dimensions

The final three dimensions are what we call commercial, meaning they reflect the ongoing value proposition to a returning customer rather than the acquisition offer. These are weekly cashback percentage where offered, tournament frequency (freerolls plus paid), and crypto-support depth (defined as the number of natively supported chains, not just the number of tokens). Commercial dimensions account for the remaining twelve percent of the composite score. We keep them at the smallest weight deliberately, because the incremental value of a weekly cashback compounds slowly and readers should not over-index on it when picking between brands. It also reflects a broader editorial view at Meridian – that the flashy differentiators between brands matter less than the boring structural ones, and our scoring should reflect that.

How GamStop shapes the offshore landscape

Before you can compare non-GamStop operators, you have to understand why the category exists in the shape it does. GamStop is not a regulator, it is a self-exclusion register operated by the National Online Self-Exclusion Scheme Ltd, and its scope is defined by the UK Gambling Commission's licence conditions. The offshore sector is what sits outside that scope, and its shape reflects that fact almost mechanically. In this section we explain the perimeter, the brand strategies that have grown around it, and how our scoring interacts with the regulatory geography.

The GamStop perimeter and where it stops

GamStop is a condition of UKGC licence, which means every operator holding a UKGC remote licence is contractually required to check the GamStop register before opening an account. Operators without a UKGC licence have no such obligation, because they are not licensed by the UKGC. That is the perimeter. It is not a loophole. It is a jurisdictional consequence. Around forty thousand UK residents were on the GamStop register as of the last published figures, and the offshore sector's addressable market for the UK includes both those on the register and the much larger group of UK residents who have never enrolled but prefer offshore terms for other reasons – higher deposit ceilings, crypto rails, larger bonuses.

Our comparison sample is built entirely from operators outside the GamStop perimeter. That is a definitional constraint, not an editorial preference. If a brand accepts UK residents and is required to check GamStop, it does not qualify for our sample.

The six regulator families we recognise

Across the twenty-two brands in our current sample, licences resolve to six regulatory jurisdictions. In descending order of sample share, these are the Curacao Gaming Control Board (with the master-licence and B2C-licence sub-categories introduced by the LOK reforms), the Anjouan Gaming Authority, the Malta Gaming Authority for a small number of brands that hold both an MGA and a Curacao licence for different markets, the Kahnawake Gaming Commission based in the Mohawk Territory in Canada, the Isle of Man Gambling Supervision Commission for the very small number of brands that hold IoM licences but not UKGC, and the Gibraltar Gambling Commissioner. We treat licence tier as an ordinal variable in our scoring, not a nominal one, and the ordering reflects our reading of the enforcement track record in each jurisdiction. Our subpage on licences goes into detail on why.

Why brand proliferation matters when you compare

A striking feature of the non-GamStop sector is the density of brand proliferation on top of a smaller number of platform providers. It is not unusual for a single platform to host fifteen or twenty customer-facing brands, each with different palettes, different welcome offers, but very substantially similar back-office infrastructure. When you compare brands from the same platform, you are usually comparing marketing offers layered on top of an identical technical stack. When you compare brands from different platforms, you are comparing more fundamentally. Our sample deliberately includes brands from at least six distinct platform providers so that structural differences show through. Aisha Rowland's editorial line on this has been consistent for eighteen months. Comparison across platforms is more useful than comparison within a platform, because within-platform differences are largely cosmetic.

Licence tiers side by side

Of the twenty-two dimensions we score, licence tier carries more weight than any single other dimension, because a licence is the first proxy for consumer protection in the absence of the UKGC. This section walks through the four tiers we assign, how we assign them, and how the tier system feeds into the composite score. If you want the deeper legal walk-through, our subpage on non-GamStop licences goes further.

How we tier the six jurisdictions

Meridian's licence tier is an ordinal score from one to four. Tier one is reserved for MGA, Gibraltar and IoM. Tier two is Kahnawake. Tier three is post-LOK Curacao B2C licences issued directly by the Curacao Gaming Control Board under the National Ordinance on Games of Chance regime. Tier four is legacy master-and-sub Curacao licences that predate the LOK reform, and Anjouan. The tier reflects our read of two things – the depth of ongoing supervision the jurisdiction publicly discloses, and the availability of a functional complaint route for the customer. We do not claim tier one is safer in any absolute sense. We claim it scores higher on our rubric for the reasons stated.

What the tiering looks like across our sample

In the current sample, tier four is the most common by count. Fourteen of the twenty-two brands sit at tier four, reflecting the historical dominance of the pre-LOK Curacao ecosystem in this market. Four brands sit at tier three under the post-LOK Curacao framework. Three brands sit at tier two on Kahnawake. One brand in our sample holds an MGA licence for European markets alongside its Curacao licence for offshore markets, and we score it at tier one for the MGA-scoped activity and tier three for the offshore-scoped activity. That kind of tier bifurcation is unusual but not unique. Where a brand holds multiple licences, we score against the licence under which UK-resident play is booked.

The gap between tier one and tier four

Because licence tier is weighted more heavily than any other single dimension in our composite, the gap between tier one and tier four is the single largest structural swing a brand can experience. A tier one licence contributes five points to the licence-tier column. A tier four licence contributes one point. Multiplied by the licence-tier weighting of ten percent of composite, that is a four-point swing on the hundred-point composite scale. To make up a four-point deficit elsewhere, a tier-four brand needs to score meaningfully above the sample median across ten or more other dimensions. In practice, that is rare. The tier-one brands in our sample tend to sit above seventy on composite. The tier-four brands tend to sit between fifty and sixty-five. This is by design, and we do not apologise for it. If we scored licence tier equally with tournament frequency, the composite would be actively misleading.

Welcome bonus benchmarking across our sample

Welcome bonuses are the single most heavily marketed feature of any casino brand, and by some distance the most inflated single variable in the sector's marketing. Our approach at Meridian is to strip the headline number back to something comparable. This section explains how we do that and where our sample sits.

Headline offer and effective value

We record two figures for every operator's welcome offer. The headline is the maximum GBP-equivalent bonus a new depositor can theoretically claim, before any wagering restrictions bite. Across our twenty-two-brand sample, the headline welcome runs from around a hundred pounds at the low end to fifteen hundred pounds at the top end, with a sample median of five hundred pounds. The effective value is our own calculation. It multiplies the headline by the probability of extraction, which is a joint function of the wagering multiplier, the minimum bet weighting, the maximum bet cap during wagering, and the expiry window. Where the headline says fifteen hundred pounds, the effective value we calculate typically lands between two hundred and four hundred, because the wagering requirement absorbs the great bulk of the theoretical value in expected-return terms.

The wagering multiplier as the single biggest lever

Wagering multiplier is the single largest determinant of effective value on a welcome bonus. In our current sample, wagering multipliers on the deposit-plus-bonus base run from twenty times at the low end to fifty times at the high end, with a modal value of thirty-five times. A fifty-times wagering requirement on a hundred-pound bonus means turning over five thousand pounds in stake before withdrawal, which sounds a lot but is not unusual by sector standards. At a slot with a ninety-six-percent RTP, the expected loss on five thousand pounds of turnover is two hundred pounds, which is more than the bonus itself. This is why the effective value is almost always well below the headline. Priya Deveraux at our sister publication has written on this in more detail, but the number to hold in your head is that the sample-median wagering multiplier of thirty-five times, at ninety-six percent RTP, extracts roughly forty percent of the bonus's face value in expected loss before it converts to cash.

Cashback, no-wager offers and reload cadence

Welcome bonuses are only one form of promotion. We separately track reload frequency, cashback percentage, and the presence or absence of a no-wager stream. Six of the twenty-two brands in our sample publish a weekly cashback of between five and fifteen percent of net losses, subject to a minimum-loss threshold. Four brands offer a no-wager welcome stream that trades a smaller headline against zero wagering. Nine brands publish a reload calendar with an offer at least twice a week during the working week. The rest run promotions on an irregular cadence advertised via email. Reload cadence, incidentally, is one of the dimensions we do not include in the twenty-two-dimension composite, because it correlates too strongly with welcome-bonus size and would double-count in the scoring.

Wagering multipliers and the mathematics of turnover

If you take one number away from Meridian Bets Journal, take the effective wagering multiplier, because it is the piece of arithmetic that governs everything else about how a welcome offer converts into withdrawable cash. This section walks through the maths, using the sample-median values we hold in our dataset, so you can carry the arithmetic into any comparison you make yourself.

The formula, plainly stated

The wagering requirement is a multiplier applied to a base, and both parts of that pairing matter. The base is usually the bonus alone (bonus-only, denoted B) or the deposit plus bonus (D+B). In our sample, twelve of the twenty-two brands calculate wagering on D+B, eight on B, and two use a hybrid where the deposit-portion wagers at a lower multiplier than the bonus-portion. Turnover is then multiplier times base. If you deposit one hundred pounds and receive a one hundred percent match bonus of one hundred pounds, and the site's wagering is thirty-five times on D+B, your required turnover is thirty-five times two hundred, which is seven thousand pounds. If the wagering is thirty-five times on B alone, your required turnover is thirty-five times one hundred, which is three thousand five hundred. That is a two-times difference, on the same bonus, driven only by the base choice.

Game weightings and the effective multiplier

Wagering is almost never applied at one hundred percent to every game. Slots typically contribute one hundred percent. Table games often contribute ten to twenty percent. Live dealer contributes anywhere from zero to twenty. This means the effective wagering multiplier you experience depends on what you play. If you spend all your bonused turnover on live dealer at ten percent contribution, your thirty-five-times multiplier effectively becomes three hundred and fifty times in slots-equivalent terms. Across our sample, the median contribution of slots is one hundred percent, of table games is ten percent, and of live dealer is ten percent. Only a handful of brands allow live blackjack to count at full weighting, and those brands typically pair that concession with a higher headline multiplier on the offer itself, so the net effect washes out.

The maximum-bet cap during wagering

Every serious operator caps the maximum single bet you can place while a wagering requirement is outstanding, and the cap matters because it slows down how fast you can move through the turnover requirement. In our sample, maximum-bet caps run from two pounds at the low end to fifteen pounds at the high end, with a sample median of five pounds. At the median cap, turning over seven thousand pounds requires at least fourteen hundred individual bets. If you play at a slot that spins in three seconds, that is seventy minutes of continuous unattended play. In practice most players spread wagering over multiple sessions, but the point is that a low bet cap is not a minor detail. It materially changes how many sessions the offer takes to clear.

Withdrawal SLA as the deciding factor

Nothing decides a returning customer's opinion of an operator faster than how long the first withdrawal takes. Across every brand we track, withdrawal SLA is the single most consistent predictor of user sentiment, which is why we score it heavily and why we measure both advertised and actual. This section walks through how the SLA scoring works and what the sample looks like.

Advertised versus measured SLA

Advertised SLA is what the operator says. Measured SLA is what we observed on a test withdrawal in the sample month. The two agree less often than you might hope. Across our twenty-two brands, six had a gap of more than twenty-four hours between advertised and measured on at least one of the payment rails we tested. Two had a gap of more than seventy-two hours. We publish both figures on the money page, and the score uses measured, not advertised. If an operator advertises instant crypto withdrawal but our test transaction was pending after twelve hours, the score reflects twelve hours. This is unforgiving on paper, but it aligns the score with the reader's experience, and we think that is the right trade.

How SLA breaks down by rail

Rail matters more than brand for typical withdrawal speed. Across our sample, the median measured withdrawal times were roughly five minutes on Bitcoin and other layer-one crypto after network confirmation, ninety minutes on Ethereum ERC-20 stablecoins after network confirmation, four hours on e-wallets such as Skrill or MuchBetter, twenty-four hours on card withdrawals via Visa Direct or Mastercard Send, and three business days on plain SEPA or Faster Payments bank transfer. Layered on top of those figures is the operator's internal processing window, which typically adds between zero and forty-eight hours. Where the operator processes withdrawals only during working hours, the internal window can stretch to a weekend delay if you happen to withdraw on a Friday evening.

The KYC gate and where SLA really lives

For most brands in our sample, first-withdrawal SLA is dominated not by the payment rail but by the KYC gate. If the operator does not require documents at deposit, they will require them at withdrawal, and the review of those documents is the single largest variable in first-withdrawal timing. Second-withdrawal timings are usually much closer to the rail median, because the KYC pack has already been cleared. This is why we report first-withdrawal and steady-state withdrawal as separate numbers on the money page. Aisha Rowland has argued repeatedly on the desk that reporting only the steady-state figure understates the friction a new customer will actually experience, and we now default to publishing both.

Game library depth and provider mix

Every non-GamStop operator markets its library size, but the number that appears in the marketing is almost always the total across every supplier the operator aggregates, which means it is only weakly informative. Meridian scores game library on depth as well as breadth, and on the mix of providers as well as the count. This section explains what we look at.

Slot counts across our sample

Sample-median slot count across our twenty-two brands is around four thousand three hundred titles. The range runs from just under nine hundred titles at the smallest brand in the sample to over eleven thousand at the largest, with a distribution that is heavily skewed to the top end. This is because most brands aggregate the same fifty or sixty supplier catalogues, and once you are past three or four thousand titles you are counting increasingly obscure supplier back-catalogue rather than genuinely differentiated titles. We score slot count on a log scale to prevent the top-end brands from dominating the column, and the log scaling brings the sample-effective range down to something more comparable across brands.

Live dealer provider mix

Live dealer is a more discriminating dimension than slot count because the sector is dominated by a much smaller set of studios. Evolution accounts for the majority of live dealer tables across our sample, followed by Pragmatic Play Live, Playtech Live, Ezugi, Vivo Gaming, and a very small number of dedicated boutique studios. We score live dealer provider count and specifically the presence of Evolution as a subscore. Brands without Evolution score meaningfully below the sample median because the Evolution catalogue is the reference set for the genre, and its absence tends to correlate with other supply-side constraints on the operator. Of the twenty-two brands in our sample, seventeen carry Evolution, three carry Pragmatic Play Live without Evolution, and two carry a limited boutique offering only.

Slot suppliers beyond the majors

Beyond the majors – Pragmatic Play, NetEnt, Play'n GO, Hacksaw Gaming, Nolimit City, BGaming, Yggdrasil and a handful of others – the interesting supplier variation happens in the mid-tier. Some brands aggregate niche crypto-native studios such as Spribe or Turbo Games alongside the majors. Others exclude specific studios for regulatory or commercial reasons – Nolimit City in particular is absent from a meaningful share of our sample because some jurisdictions have historically pressed operators to remove certain of its higher-volatility titles. Aisha Rowland ran a supplier-diversity supplement earlier in the year that mapped the twenty most-and-least-carried studios across our sample, and the results are used to weight the game library score toward diversity rather than raw count.

Payment coverage matrix

How you get money in and out of an offshore operator is where the mechanics really differ from a UKGC brand, and the payment coverage matrix is one of the most requested pieces of our editorial dataset. This section summarises the current sample. The subpage on payment methods goes rail-by-rail through the detail.

Card, e-wallet, crypto, bank transfer

The four rail families that matter are cards (Visa, Mastercard), e-wallets (Skrill, Neteller, MuchBetter, ecoPayz), crypto (Bitcoin, Ethereum, USDT, plus a growing tail of chain-specific stablecoins), and traditional bank transfer (Faster Payments, SEPA). Across our sample, twenty of the twenty-two brands accept card deposits from UK-issued cards, though only fourteen support card withdrawals via Visa Direct or Mastercard Send. E-wallet acceptance is somewhat higher than card in withdrawal support, at around eighteen brands. Crypto is close to universal on the deposit side, with twenty-one of twenty-two accepting Bitcoin and nineteen accepting at least one stablecoin. Bank transfer is the rail with the most variance, accepted at deposit by only twelve brands and at withdrawal by fewer still.

Paysafecard and prepaid options

Paysafecard and other prepaid vouchers are a distinct segment because they carry no withdrawal path back to the same instrument. If you deposit with Paysafecard, you withdraw to a different rail, and most operators require that rail to be verified independently under their KYC policy. In our sample, sixteen of the twenty-two brands accept Paysafecard at deposit, but the accompanying withdrawal channel is typically e-wallet or bank transfer, and the operator will insist on documentary verification of the destination account. We score Paysafecard acceptance as a positive on deposit-method count but neutral on withdrawal-method count, because the withdrawal has to happen elsewhere.

Deposit and withdrawal minimums

Minimums vary by rail. Card and e-wallet deposits usually start at ten to twenty pounds. Crypto deposits often carry a minimum of the equivalent of twenty pounds, though a handful of crypto-first brands set that as low as one pound. Bank transfer minimums are the highest, typically fifty pounds at deposit and one hundred at withdrawal. Withdrawal minimums are generally higher than deposit minimums, in the range of twenty pounds on crypto and e-wallet up to a hundred pounds on card and bank transfer. Where an operator sets a low withdrawal minimum on crypto, it is usually because the operator's own settlement cost on the crypto rail is lower than on the card rail. This is one of the reasons crypto-first brands tend to have friendlier small-value withdrawal policies, and it is a systematic pattern in the sample, not a coincidence.

Mobile UX and interface scoring

Roughly seventy percent of sessions across our sample happen on mobile, which makes the mobile experience arguably more important than the desktop one. Meridian scores mobile UX on a rubric of nine subchecks, and the sample distribution is one of the most spread out of any dimension we track. This section walks through what we look at and why.

Native app or PWA or plain web

Non-GamStop operators cannot list a real-money casino application in the Apple App Store or Google Play, both of which prohibit real-money gambling apps from unlicensed jurisdictions. Some operators sideload an Android APK from their own domain. Others provide a progressive web app that installs to the home screen. Others simply provide a responsive web site. Across our sample, three brands offer a sideloaded Android APK, ten offer a PWA install prompt, and the remaining nine are responsive-web only. We score PWA installation as roughly equivalent to a native experience for practical purposes, because the PWA lifecycle is well supported on iOS and Android alike. Responsive web scores lower only where the responsive implementation fails one of our other sub-checks.

The nine mobile sub-checks

Our mobile UX sub-rubric checks the following. First, does the site load below three seconds on a throttled 4G connection. Second, does the primary CTA fit above the fold on a standard six-inch device. Third, is the game grid infinite-scroll or paginated. Fourth, does the sticky footer navigation obscure content on iOS Safari. Fifth, is search available from every screen. Sixth, is the deposit modal single-flow or multi-flow. Seventh, can withdrawal be initiated from the account page without going through cashier. Eighth, is live chat accessible from mobile without switching context. Ninth, is the responsible gambling toolkit accessible from the mobile primary menu. Each returns a binary pass or fail. Total score is the sum of passes, expressed as a zero-to-nine value and then normalised into the composite.

Where the sample struggles

The single most common failure across our sample is sub-check nine – RG toolkit accessibility from mobile. Fourteen of twenty-two brands hide the RG toolkit two levels deep in the account menu, requiring an average of four taps to reach the deposit-limit setter. That matters because friction against setting a deposit limit is exactly the kind of friction the toolkit is meant to remove. The second most common failure is sub-check four, iOS Safari footer interference, which affects nine of twenty-two brands on the version of iOS we tested. The third is sub-check one, load time on throttled 4G, which fails at seven brands, most of them running heavy hero animations or ungated auto-play video on the landing lobby. These are all fixable issues, and where a brand has moved on any of them between our sample refreshes, we flag it in the changelog.

Customer support hours and response benchmarks

Customer support is what you contact when the product fails, and its quality is a function of hours, channels, and the speed and skill of first response. Meridian scores support on three sub-dimensions. This section explains what we found in the current sample.

Advertised versus staffed hours

Every brand in our sample advertises live chat, and eighteen of twenty-two claim 24/7 coverage. Our test contacts told a different story. Of the eighteen 24/7 claimants, twelve had a genuine human agent responding within ninety seconds at 03:00 UTC on a Tuesday. The other six routed the query to a bot with a human handoff of between forty minutes and four hours. We score the staffed hours we measured, not the advertised hours. Where a brand runs a bot-first triage during off-hours, we score that separately as a channel note and reduce the staffed-hours count accordingly. Fourteen of twenty-two brands in the sample have genuinely staffed 24/7 live chat by our measurement.

Channels and their relative usefulness

Live chat is the dominant channel across our sample, followed by email, followed by an in-account ticketing system. Only four brands offer telephone support, and of those four, only two have a UK-facing number. WhatsApp support is offered at three brands, mostly targeting non-UK markets but accessible from the UK. Telegram support is offered at two brands in the crypto-first subset. The most useful channel for a first-response question is almost always live chat, because it forces synchronous engagement. The most useful channel for a complex account or KYC issue is email or ticketing, because it forces the exchange to leave a written trail. Aisha Rowland's editorial view, articulated in our support scoring memo, is that the presence of both a synchronous and an asynchronous channel is a positive signal, and the absence of either drags the score.

Escalation and complaint routes

Beyond first-line support, the more consequential question is what happens when the first-line agent cannot resolve a query. Across our sample, only nine of twenty-two brands publish a documented complaint escalation path within the operator itself. Of those nine, eight further name an external Alternative Dispute Resolution provider to which the customer can escalate if internal escalation fails. The named ADRs vary by licence jurisdiction. Curacao brands tend to name the Antillephone or Gaming Curacao ADR routes. MGA brands name the Malta Gaming Authority's own dispute service. Kahnawake brands name the KGC's Fair and Responsible Gaming Standards process. This is one of the areas where the licence-tier ordering we use is validated by observable behaviour rather than just published policy.

KYC thresholds compared across jurisdictions

The KYC threshold, or Know Your Customer verification threshold, is the deposit or withdrawal value at which the operator escalates from streamlined onboarding to a documentary check. Where the threshold sits, and how strictly it is applied, matters both to how quickly you can play and to what the operator does with your data. This section explains our approach.

What the threshold is and why it varies

KYC thresholds are set by a combination of licence conditions, payment-provider terms, and the operator's own risk policy. At the licence level, some jurisdictions (Malta, Gibraltar, Isle of Man) require KYC at account opening, so the threshold is zero. Others (Curacao, Anjouan) allow the operator to set its own threshold subject to broad anti-money-laundering norms. At the payment-provider level, cards and bank transfers usually attract lower thresholds than crypto rails, because the payment provider itself has stricter KYC obligations. At the operator level, the risk policy is where the real variation lives, and this is why the sample spread is so wide. In our current sample the KYC threshold, expressed as cumulative lifetime deposit before mandatory documentary check, ranges from two thousand pounds at the low end to five thousand pounds at the high end, with a median around two and a half thousand.

What documents get requested and when

The standard KYC pack across our sample is proof of identity (passport or driving licence), proof of address (utility bill or bank statement dated within three months), and where card payments are involved, a redacted image of the deposit card showing the first six and last four digits. Some operators additionally request a selfie holding the identity document, or a video verification via a third-party provider such as Sumsub. Enhanced Due Diligence, triggered at higher thresholds usually around five thousand pounds cumulative deposit, adds source-of-funds evidence, which can include payslips, tax returns, or bank statements covering three to six months. Every operator we tracked in the sample was capable of triggering an EDD request, though only about a third did so during our test period.

How the KYC dimension is scored

Our KYC dimension is not simply the threshold value. It combines three things – the threshold as advertised, the transparency of the trigger (whether the operator tells the user in advance what triggers a check), and the time-to-clear once documents are submitted. A high threshold with an opaque trigger and a slow clear scores worse than a low threshold with a transparent trigger and a same-day clear, even though the high-threshold operator might feel more permissive at first. Across our sample the transparency sub-score is the single largest driver of variance in KYC scoring. Operators that publish a clear KYC policy on the terms page score meaningfully above those that leave the process implicit until the moment a check is triggered.

Responsible gambling toolkit comparisons

Outside the UKGC perimeter, responsible gambling tools are not mandated by the same conditions, and their presence and quality vary widely. This is one of the dimensions where the sample spread is largest, and it matters more than the score might suggest. This section walks through what we look at and what we found.

The core RG toolkit and what it should contain

A minimally sufficient RG toolkit contains four things. First, a deposit limit setter, ideally with daily, weekly and monthly bands, and ideally with a cooling-off period before an increase takes effect. Second, a loss limit setter, distinct from the deposit limit because it caps net loss rather than gross deposit. Third, a session timer or reality check that surfaces elapsed session time at fixed intervals. Fourth, a self-exclusion mechanism, ideally supporting periods of at least six months and up to permanent. Across our sample, twelve of twenty-two brands offer all four in a functional form. Six brands offer three of the four, typically missing the loss limit or the reality check. Four brands offer only self-exclusion and a deposit limit without loss limit or reality check.

Extended tools and their scarcity

Beyond the core four, we score for the presence of extended tools. Time-out for a specified short period (typically twenty-four hours to seven days) is a friendlier stepping stone toward self-exclusion, and eight of our sample brands offer it. Loss-history dashboards, which surface cumulative net win/loss across a specified period, are offered by only five brands. Reality checks that force a break at a set threshold (rather than merely displaying elapsed time) are offered by three. Integration with external blocking software – Gamban or BetBlocker – is offered by two, both of them Kahnawake-licensed. The rarity of extended tools reflects the sector's regulatory context, and where a brand voluntarily offers them we score it upward, because voluntary provision beyond the licence floor is an editorial positive.

How to use the tools that exist

Whatever operator you play with, the practical guidance from Aisha Rowland and the desk is consistent. Set a deposit limit before the first deposit, not after the first loss. Use the shortest cooling-off period on limit increases you can find. Enable session timers even if they feel intrusive, because the intrusion is the point. If you find yourself exceeding limits or breaking cool-off periods, treat that as a signal, not a nuisance. GamCare is contactable on 0808 8020 133 for a free, confidential twenty-four-hour helpline. BeGambleAware offers a chat service via the National Gambling Helpline. Where the operator's toolkit is inadequate, external blocking software (Gamban, BetBlocker) can layer over the top. The lack of a comprehensive toolkit at a given operator is not a reason to play there without any tools at all.

Terms and conditions readability index

The terms and conditions document is where the small print of a welcome bonus lives, where withdrawal caps are defined, where the KYC and AML policies are specified. It is also, at most operators, actively hostile to actually being read. Meridian's terms readability index is our attempt to make that hostility measurable.

How we measure readability

We use the Flesch-Kincaid grade level as the underlying readability metric, applied to the full terms-and-conditions document as published on the operator's own site. Across our twenty-two brands the sample-median grade level is fifteen, which in plain English means the terms are pitched at a first-year university reading level. Six brands come in above grade seventeen, which is postgraduate territory. Two brands come in below grade twelve, which is broadly what we would consider a fair floor for a document meant to be read by an ordinary customer. Where an operator's terms-and-conditions grade level is above sixteen, we deduct meaningfully from the readability sub-score, because a document that is not read is a document that cannot serve its stated protective function.

The clauses that matter most

Grade level is only half the story. The other half is how the terms document handles the six clauses that most commonly bite the customer. These are the maximum-bet cap during wagering, the game-weighting table, the maximum cashout on bonus wins, the dormant-account policy (including any account-fee schedule), the withdrawal reversal or pending window, and the operator's stated right to close accounts. We check each of these six clauses for two properties – is it clearly signposted with a heading or table, and is it stated in numerical rather than editorial terms. A term that says maximum bet during wagering is subject to reasonable limits at our discretion is worse than a term that says maximum bet during wagering is five pounds per spin. We score for the number of the six clauses that are specified numerically, out of a possible six.

Where the sample tends to be weakest

Across our sample, dormant-account policy is the clause most commonly stated in vague rather than numerical terms. Only twelve of the twenty-two brands publish a specific dormancy trigger (usually six or twelve months of no activity) and a specific fee (usually five to twenty pounds per month). The rest reserve the right to charge reasonable or industry-standard fees, which is not a numerical commitment. The second-weakest clause is the operator's account-closure right, which most brands assert broadly and with limited procedural constraint. This does not mean operators routinely close accounts on flimsy grounds, but it does mean the reader has less protection on paper than they would under a UKGC licence, where the licence conditions themselves narrow the operator's discretion in this area.

How to read our comparison tables

A comparison table looks straightforward but reading one is a skill, and reading one well requires knowing what the columns are, how the values are normalised, and what the ranks and scores do and do not tell you. This section is a short guide to using Meridian's tables well, especially the larger matrices you will find on the money page.

Composite versus dimensional scores

Every operator in our sample has one composite score, out of one hundred, and twenty-two dimensional scores that feed into it. The composite is useful for ranking, and it is a fair summary of overall performance, but it can also obscure important trade-offs. A brand at seventy composite might be a tier-one licence at the top of the market on everything except tournament frequency, and another brand at seventy composite might be a tier-four licence with heroic scores on welcome bonus and cashback but weak scores on withdrawal SLA and support. These two brands sit at the same composite but represent very different offers. To choose between them intelligently, you have to read the dimensional scores, not just the composite. We publish both for exactly this reason.

Rank order versus score gaps

Ranking implies distance – if brand A is ranked first and brand B is ranked second, we tend to read that as A being the best and B being close behind. Composite scores are less flattering to that reading. The rank-one and rank-two brand in our current sample are separated by just over one composite point, which is well inside the noise band on our measurement. The rank-two and rank-three brands are separated by three points. Rank-three and rank-four by eight. What this means is that the top three brands are effectively tied and the fourth brand is meaningfully behind them, even though the ordinal ranking would suggest a smooth descent. Aisha Rowland has argued in the desk memo that composite gaps below three points should be read as ties, not as ordinal separations, and the tables you find on the money page reflect that convention with an explicit tied-cluster grouping.

What our tables do not measure

The tables measure the twenty-two dimensions in the rubric. They do not measure how any individual player will feel about a brand, how a specific game will fare on a specific day, or whether the offshore regulatory framework of a given jurisdiction will remain unchanged over the next twelve months. They also do not measure the operator's financial solvency in any deep sense. Meridian is a data publication, not a credit rating agency. If the sector experiences a wave of insolvencies in the coming year, our sample scores will not have predicted it, because insolvency risk is not in our rubric and we do not have the underwriting data to score it if it were. This is a limit of the method, not a defect in it. Read the tables as a decision aid, not a decision.

1. 1Read the composite as a rough sort. Composites above 70 sit in the top quartile, composites below 55 in the bottom quartile.
2. 2Then look at the dimensional profile. Two brands at the same composite can have very different dimensional trade-offs.
3. 3Weight the licence tier column heavily. A tier-one brand at composite 70 is not the same reader experience as a tier-four brand at composite 70.
4. 4Cross-check the withdrawal SLA column. First-withdrawal SLA is the single most consistent predictor of whether a brand feels good or bad in ongoing use.
5. 5Check the sample-refresh date. Our data is not real-time. Brands move on individual dimensions between refreshes.
6. 6Treat composite gaps under three as ties. Ordinal ranking is convenient but the measurement noise is real.

Responsible gambling and where to seek help

Meridian Bets Journal exists to help readers make more informed decisions in a category where the marketing is loud and the information is sparse. That mission does not extend to encouraging play at any brand at any time, and this section is where we set out the responsible gambling guidance that underpins every page on the site.

The scope of GamStop and where it does not reach

GamStop is a self-exclusion register that applies to UKGC-licensed remote operators. Offshore operators are outside that scope, which is a factual statement about jurisdiction, not a recommendation. If you have joined GamStop as part of a considered response to problem gambling behaviour, we would not encourage you to seek out offshore operators as a workaround. The register exists because self-exclusion is a proven harm-reduction tool. Circumventing it on impulse is a decision most people go on to regret. If GamStop is on your account, the presence of offshore operators does not remove the reasons you enrolled in the first place.

GamCare, BeGambleAware and the National Gambling Helpline

GamCare runs the National Gambling Helpline, free and confidential on 0808 8020 133, available twenty-four hours a day. The helpline is a first port of call for anyone who thinks their gambling might have become a problem, or who is affected by someone else's gambling. BeGambleAware runs an information hub at begambleaware.org, with self-assessment tools, a chat service, and a network of local support organisations across the UK. Gamban is a paid device-level blocking application. BetBlocker is a free equivalent maintained by a charitable organisation. Both work independently of any operator's own toolkit and can be layered over any device. Aisha Rowland has argued repeatedly that reader-facing publications in this category have a duty to signpost these services on every page, and Meridian's pages are structured to comply.

Practical steps if you are worried

If you are worried about your own gambling, or someone else's, the practical first step is to talk to someone. That someone might be a partner, a friend, or a helpline advisor. The second step is to reduce access – deposit limits, session timers, blocking software, or self-exclusion. The third step is professional support, which the National Gambling Helpline can refer you to through its network of NHS-linked and charity-linked providers. Financial harm from gambling is treatable, and the earlier the intervention the better the outcome. If you are reading this page because you are worried about someone else, the same helpline covers affected others. GamCare's advisors are trained to work with concerned family members as well as with gamblers themselves. There is no requirement to speak first to the gambler themselves.