Every rating on this site is a weighted average of seven measured areas, and the weights are published below rather than implied. The point of writing them down is that you can disagree with them. If you care more about the size of a game library than about whether a withdrawal arrives, our number will be wrong for you, and you will be able to see exactly why.
Nothing here is scored from a press release.
What a score on this site measures
A rating out of ten answers one question: how much of what a casino promises can be verified, and what happens to your money when something goes wrong. It does not measure how enjoyable a lobby is, nor predict that any individual session will go well.
That framing has a consequence worth stating early. A casino with 2,000 excellent games and no published regulator will score below a duller one that pays reliably and can be held to account, because the first is a taste question and the second is a money question.
The seven areas and what each one is worth
The weights total 100% and apply to every rated casino on the site, so two reviews can be compared against each other rather than only against themselves.
| Area | Weight | What it measures |
|---|---|---|
| Payouts and the cashier | 25% | Withdrawal limits, processing windows, fees, penalties and whether a cashout completes |
| Licensing and recourse | 20% | Whether a licensor is named and verifiable, and who hears a complaint |
| Bonus terms and fairness | 15% | Wagering, caps, weighting, expiry and whether the terms are published at all |
| Game library | 12% | Provider count, live coverage, demo availability, honest counting |
| Responsible-gambling provision | 12% | Deposit and loss limits, timeouts, self-exclusion, correct local help services |
| Support | 8% | Channels, availability, accuracy of what agents actually say |
| Transparency | 8% | What the operator publishes voluntarily, and what it declines to publish |
Payout behaviour carries the largest share because it is the only area where a failure costs you money you already own. A thin game library disappoints you. A withdrawal that stalls behind an unpublished limit removes a balance.
Payouts and the cashier, at 25%
This is the area with the most hand work in it. For each market a casino serves, the cashier is opened with the country selector set to that market and every limit read off individually: deposit floor, deposit ceiling, withdrawal floor, withdrawal ceiling, per method.
What earns marks here is a published ceiling you can plan around, a processing window stated in the terms rather than in marketing copy, no fee on an ordinary withdrawal, and parity between the way money goes in and the way it comes back out. What loses marks is a withdrawal cap lower than the deposit cap on the same method, a penalty for cashing out early, a monthly ceiling that turns a win into an instalment plan, and a processing desk whose hours fall overnight in the market being reviewed.
Evidence counts only if it is first-party or first-hand. The operator’s own published rule counts. A timed withdrawal from a funded account counts. A third-party review quoting a payout speed does not, and neither does a support agent’s verbal promise unless the terms back it.
Licensing and recourse, at 20%
Two things are checked, and they are not the same thing. First, is a licensor named, with a number, in a place a reader would find it. Second, does that number resolve in the regulator’s own register.
A licence claim that cannot be looked up scores no better than no claim at all. The full score in this area needs a named regulator, a working register entry, and a route to an independent complaints body or dispute scheme. An operator that publishes no licensor loses most of this weighting outright, because the consequence is concrete rather than theoretical: there is nobody above the operator to appeal to.
Conflicting third-party claims about a licence are reported as conflicting third-party claims. They are never resolved by picking whichever one sounds most plausible. How that plays out for this particular operator is unpicked in the what we found when we looked for a licence.
Bonus terms and fairness, at 15%
The first question is whether the terms exist in public. An operator that prints the wagering multiplier, the maximum bet under bonus, the weighting table and the expiry window scores well here even when those numbers are unflattering, because a reader can do the arithmetic before depositing.
The second question is what the numbers do. Wagering is read as a multiple of deposit plus bonus or of bonus alone, since the two are not comparable. Weighting is read against the headline multiplier, because 35x with slots at 50% is really 70x. Maximum-cashout caps are checked against the bonus they apply to. Expiry windows are checked against anything else in the terms that runs on a clock, which is where collisions turn up.
Marks come off for anything that voids a balance rather than a bonus, for a bonus dispute clause that makes the operator the final arbiter, and for headline totals assembled from several deposits and advertised as one.
Game library, at 12%
Studio counts come out of a casino’s machine-readable site index, one entry per provider, tallied by hand. A front-page claim of “200+ providers” is not evidence of anything; a list of named pages is, because it can be counted while a claim can only be believed.
Live coverage, roulette and blackjack variant counts and jackpot totals come from the on-site category counters, which are reported as the operator’s own figures. Demo availability without an account is checked by loading games, since it is the cheapest way for a reader to audit a lobby before spending anything. A library this size always has titles excluded from bonus play, so the excluded list is read too, and a breakdown of what sits in the lobby lives on the lobby and studio notes.
Responsible-gambling provision, at 12%
This area is scored on tools rather than on tone. A page of advice about playing within your means is worth very little; a deposit limit you can set yourself in the account area is worth a great deal.
The checklist is short and specific: deposit limits, loss limits, session limits, reality checks, a cooling-off or timeout option, and self-exclusion you can trigger without emailing anyone. The local help services listed are checked against the market, because an operator pointing New Zealand players at a UK charity has not thought about who is reading. Where an operator’s own page falls short, the right services for this market are gathered on the help and limits guide.
Support and transparency, at 8% each
Support is scored on channels, availability and accuracy. Live chat is opened before registration where that is possible, and the agent is asked questions whose answers can be checked against the published terms. An agent who answers quickly and wrongly scores worse than one who answers slowly and correctly.
Transparency is the residual category, and it catches the things an operator could publish and chooses not to: the company name, the dormancy policy, the withdrawal ladder, the VIP thresholds, the auditor behind a segregated-funds claim. Credit goes both ways here. An operator that prints an unflattering withdrawal table gains marks for printing it.
What each band means
A number only helps if the bands behind it are stable, so here is what each one is intended to mean.
| Score | Reading | Typical profile |
|---|---|---|
| 8 and above | Recommended | Verifiable licence, real self-exclusion tools, withdrawal rules that favour nobody in particular |
| 7 to 7.9 | Solid with caveats | Licensed and functional; one area, usually bonus terms or payout ceilings, is weak |
| 6 to 6.9 | Proceed informed | Something structural is missing or unverifiable; the product itself may be good |
| 5 to 5.9 | Not recommended for most | Multiple failures, or a term that can cost a player their balance |
| Below 5 | Avoid | Evidence of forfeiture, no recourse, no meaningful player protection |
A six is not a pass with a shrug. It means the casino may function perfectly well and that a specific, named thing about it cannot be checked, which is a risk a reader is entitled to accept knowingly.
The rules that bind what gets written
Scoring rules are easy to publish. These are the ones that constrain the sentences.
- A figure that appears nowhere in an operator’s own pages is labelled unpublished. It is never estimated, averaged from competitors or inferred from a pattern.
- A third-party claim is attributed to the party making it, every time, and never promoted to a fact because it has been repeated often.
- A score can be revised, and the review carries the date it was last checked so you can see how fresh the number is.
- Commercial relationships do not move a score. Operators reviewed here may pay a commission on referrals; that arrangement buys placement and nothing else, and it is disclosed in the footer of every page.
- Where the evidence says a casino, a bonus or a payment route is a bad idea for a particular reader, the page says so plainly. A review that never declines to recommend anything is not a review.
- Marketing adjectives have to earn their place with a number or a named entity in the same sentence, which is why you will not read that a cashier is fast without being told how fast and against what.
How a casino gets tested here
The process is the same each time, in roughly this order.
- Open and fund a real account in the local currency, at the published minimum, so the deposit floor is tested rather than quoted.
- Pull the cashier limits by hand for every market under review, switching the country and currency selectors separately to catch a localisation that is only a layer.
- Read the full terms, start to finish, noting the version number and date. The sections on verification, withdrawals and bonus weighting get read twice.
- Claim a bonus where one applies, and track the wagering window against every other clock in the terms.
- Time a withdrawal cycle from request to arrival, recording which day of the week it was lodged and what the processing clause says.
- Complete identity verification, including whatever documents are demanded, and record how long the review took against the window the operator promises.
- Message support with questions whose answers are already known from the terms, then compare.
- Tally the studios from that machine-readable index instead of from a marketing badge.
- Load games on desktop and on both test phones, in demo mode and with real money.
- Write down what could not be verified, and publish that list rather than quietly omitting it.
A score is revised when any of that changes: new terms version, a moved limit, a withdrawal that behaves differently from the last one, or a tool appearing or disappearing from the account area. The published date moves with it. Figures given in New Zealand dollars are read from a New Zealand account, and euro figures from a euro account, because converting one into the other hides exactly the asymmetries worth reporting.
About the reviewer
Reviews here are written by Erin Caldwell, Lead Reviewer, with nine years spent reviewing online casinos and part of that working in payments support at a licensed gaming platform. Testing is done with real-money accounts in New Zealand dollars and, for the Finnish pages, a second account in euros.
No employer is named, and that is deliberate rather than coy. A reviewer’s claimed CV is the easiest thing on a site like this to invent and the hardest thing for a reader to check, so this site publishes only the claims that affect how the work was done. The same logic rules out a photograph, a university and a list of social profiles.
What can be judged is on the page: the limits, the clause numbers, the version date of the terms and the arithmetic are reproducible by any reader willing to open the same cashier. More about who publishes this site sits on the about page, and the applied version of everything above is the rated Alawin review.
Frequently asked questions
Why does payout behaviour outweigh the size of the game library?
A weak library costs you enjoyment. A cashout that fails removes money already sitting in your name. Payouts and the cashier carry 25% of the score; the game library carries 12%.
Can an operator pay to raise its rating?
No. Referral commissions fund the site and are disclosed on every page, but they have no input into the seven weightings or the figures behind them. A paying operator can and does score in the sixes.
What makes a score get rewritten?
A new terms version, a changed limit, a different withdrawal outcome, or a responsible-gambling tool appearing or being withdrawn. The date on the review moves whenever the number does.
Where do the provider counts in your reviews come from?
From a casino’s own machine-readable index of pages, counted one studio at a time. A badge on the home page quoting a round number is not treated as evidence.
What does it mean when a number is marked unpublished?
That it appears nowhere we could find on the operator’s own pages. It is not an estimate and not a gap to be filled from a competitor, and the review says so rather than guessing.
Why is the reviewer’s former employer not named?
Because the claim could not be checked by a reader, and an unverifiable credential has no business sitting beside a rating built on verification. Only claims that bear on the testing are published.
Do you score the same casino differently for different countries?
The weightings stay fixed, but the evidence is market-specific. Limits, payment rails and help services are read from the local cashier, so the same casino can lose marks in one market and keep them in another.
Sources
- The current Alawin rulebook, version 1.12: an example of the primary document every rating here is read against, clause by clause.
- the New Zealand gambling regulator’s own published remit: the regulatory baseline used when judging whether an offshore operator offers any comparable recourse.
- Safer Gambling Aotearoa and its 24/7 helpline: the local service every operator help page gets measured against.