About
How We Review and Score Casinos
Six weighted sections, one licence check run in the same order every time, and a published account of what this research could not reach.
Worked example: FoggyBet
7.7/10
- Licensing & safety 25% of total
- 9.1
- Payments & payouts 20% of total
- 6.1
- Bonus terms 20% of total
- 6.5
- Games & providers 15% of total
- 9.1
- Mobile & UX 10% of total
- 7.1
- Support 10% of total
- 8.1
A licence that does not cover the player market makes the rest worthless.
What the operator publishes per method, and whether it publishes anything.
Wagering base, max bet and expiry, not the size of the headline number.
Named providers visible in the lobby, and a game count only where we counted it.
The public mobile site as it loads, without an account behind it.
Channels and languages the operator advertises, and whether the claim holds up.
What a score on this site is
A score is a weighted mean of six section scores, each out of 10, using the weights below. The panel above is drawn from the same weights the reviews are scored against, so display and arithmetic cannot drift apart. Nothing is added by hand at the end, and no editor lifts a 5.8 to a 6.0 because a brand looks like it deserves the rounder number. The six inputs are the part worth arguing with.
The six weights
| Section | Weight |
|---|---|
| Reputation and licensing (safety) | 0.25 |
| Payments and payouts | 0.20 |
| Bonus fairness | 0.20 |
| Games and providers | 0.15 |
| Mobile experience | 0.10 |
| Customer support | 0.10 |
Safety takes a quarter because a licence nobody can evidence makes the other five sections academic. Payments and bonus fairness take 0.20 each, the two places where a clause decides whether a win turns into money. A lobby of 4,000 slots matters less than a cashier that pays back the way it took money in, so games, mobile and support carry the rest.
What each section measures, and what it does not
Reputation and licensing. The register entry behind the licence number: licensee, status, listed domains, plus the terms setting the dispute route and the scope of a self-exclusion. Not how long a brand has advertised, and not how many badges sit in the footer.
Payments and payouts. Deposit list against withdrawal list, minimums, per transaction ceilings, fees, dormancy charges, and the processing time the operator publishes, quoted as its claim. Not how fast money arrived. Nobody here has taken money out of these casinos.
Bonus fairness. The wagering multiplier and what it applies to, the maximum bet clause while a bonus is live, game weighting, expiry, excluded titles, any cap on winnings. Not the headline percentage. A 350% offer at 30x on deposit plus bonus is worse than a 100% at 25x on bonus only, and the scoring treats it that way.
Games and providers. Which studios are visible in the lobby and how the catalogue is organised. A counted total appears on two of the 20 reviews. Elsewhere there is no total: counting a lobby properly needs an account, and repeating the operator’s number would be advertising with our name on it.
Mobile experience. The public site on a phone browser: whether cashier limits are reachable, terms readable at that width, safer gambling controls findable. Not a native app nobody installed.
Customer support. Channels advertised, published hours, published languages, and whether the FAQ answers a single payment question. Support was tested on one brand, SlotRoyals. On the other 19 the section scores documentation rather than responsiveness, and says so.
How the overall is worked out
Multiply each section by its weight, add the six products, round to the nearest fifth of a point. Take the Nords Casino numbers: safety 7.6, payments 4.2, bonus 5.2, games 7.0, mobile 5.6, support 6.0. That gives 5.99, published as 6.0.
Rounding to fifths is why every score here ends in .0, .2, .4, .6 or .8, and never .1, .3, .5, .7 or .9. Section scores are judgements about evidence, and evidence rarely supports a distinction of a tenth of a point between two operators. Fifths are as fine as the inputs allow, so fifths are what gets printed.
The licence check, in order
The starting point is the licence number printed on the site, not the logo beside it. A badge is an image, and an image proves nothing. Three questions then run in the same order, every brand.
- Is the entry live? Not lapsed, not surrendered, not expired, and not a number that returns nothing.
- Does the licensee match the operating company? The name on the register has to be the company named in the terms, the one taking deposits.
- Does the register list this domain? A live licence often covers domains that do not include the site a reader is being sent to.
Registers opened by name: the Malta Gaming Authority authorisation register; the Anjouan Gaming licence register at anjouangaming.com/license-register/; the Curacao Gaming Authority at cert.cga.cw plus its published PDF register; the Tobique Gaming Commission at thetgc.ca/license-holders/; the Kahnawake Gaming Commission.
Six of the 20 brands did not survive that sequence. A regulator is named on the site and nothing in a public register confirms it. Each of those six reviews states it in the first paragraph rather than near the bottom.
What we did not do
- No accounts were opened. No registration, no deposit, no play, no withdrawal, at any of the 20 brands.
- No payout was timed. Every payout figure here is the figure the operator publishes, attributed to the operator. Where an operator publishes nothing, the review says nothing is published.
- Support was tested on one brand. SlotRoyals. On the other 19 the review lists the advertised channels and says they were not contacted.
- Game totals were counted on two brands. The other 18 name the studios visible in the lobby instead of quoting a marketing figure.
Everything else came from public pages: landing pages, cashier pages, terms, bonus terms, FAQs and registers. That is a real limit, it is printed on every page it applies to, and our editorial policy covers what happens to a draft that ignores it.
Why the average is low, and why that is the honest answer
Scores run from 6.1 to 7.7. The mean is 7.03, the standard deviation 0.44, and 18 of the 20 sit under 7.6. There is no floor, no minimum publishable score, and nothing was quietly dropped for scoring badly.
That distribution follows from the weights rather than fighting them. Weight safety at a quarter, find six brands whose licence cannot be evidenced, and the arithmetic goes where it goes. Score bonus fairness on the wagering basis instead of the headline, and generous looking offers land in the sixes. A site wanting every brand to look recommendable would raise the floor, cut the low scores, or move the weights. Publishing the weights first makes that hard to do quietly.
Spotted a wrong figure? Report it. Wondering who pays us and what that does not buy? Next page.

Alana Hippensteele
Managing Editor since 2024
Alana Hippensteele is the Managing Editor of LSM99 and the person whose name sits behind the site's editorial standards. She has run the newsroom since 2024. The job is unglamorous and mostly administrative in the best sense: she assigns every casino review and market page, sets the brief for each writer, and signs off the final fact-check before anything goes live. Nothing publishes on LSM99 without passing across her desk. She owns the editorial policy, the scoring methodology and the corrections process, which means when a rating is wrong she is the person who has to answer for it.