Each rating published here is the output of seven weighted areas, and the heaviest of them, at 25%, asks one question: when something goes wrong with your money, who is obliged to listen? Below are those weights, what each band means in ordinary words, the evidence a page here may rest on, and the three things this approach cannot tell you.
The evidence register
Coverage of 7Bit Casino on this site is documentary. That word carries the whole method, so here is exactly what it admits and what it shuts out.
Admitted: the operator’s own published pages, read from start to finish and filed under the date they were pulled. The terms and conditions, the bonus terms, the cashier tiles, the twelve-level loyalty ladder, the play-safe page, the help-centre answers and the fair-games explainer. All of it was pulled on 20 October 2026 over a consumer-grade Canadian internet connection, because an online casino serves different currencies, ceilings and offers according to where a request appears to come from, and a reader in Winnipeg or Victoria needs the version Canada is shown rather than one converted out of euros.
Shut out: anything needing a funded account. None was opened, nothing was staked, no payout was requested, and no message went to the support desk. The verbs available here are read, captured, counted, compared, added up and checked. Arithmetic counts as reading: when a page states that four welcome tiers total 325% and C$10,800, that is addition done on figures the operator printed, and it says so.
Where the operator publishes nothing, the page names the absence rather than lifting a figure off a rival review or an affiliate feed. Six figures had no published value at all on that date, and carry a label saying so wherever they arise:
- the point thresholds separating one loyalty level from the next
- any pending or review window between a payout request and a payment
- payout times method by method, as distinct from deposit times
- the testing house behind the random number generator
- return-to-player percentages, per game or per category
- which of two conflicting rollover multipliers governs a live bonus
An empty cell in a table on this site means the operator left it empty. It never means we guessed and rounded.
The seven weighted areas
Weights are set before a page is drafted, which is the only workable guard against a score being reverse-engineered from a conclusion somebody already favoured. They are the same for every operator covered.
| Area | Weight | What moves it |
|---|---|---|
| Licensing and recourse | 25% | A licence number in the format its named regulator actually issues, a company standing behind it, and an appeal route that the operator does not control |
| Payments and limits | 20% | Fees, floors, ceilings, instalment rules, and every condition attached to asking for money back |
| Bonus terms and their arithmetic | 15% | What a headline resolves to tier by tier, the rollover multiplier, stake caps, expiry and game weighting |
| Game catalogue and availability | 15% | What can be counted, what the studios permit in this market, and whether fairness is demonstrated or asserted |
| Responsible-gambling provision | 10% | Which controls exist, whether they function on every funding rail, and whether the help offered reaches this country |
| Account and verification process | 8% | Document lists, timing commitments, and whether one clause of the terms sits awkwardly beside another |
| Support and disclosure | 7% | Published channels and hours, plus how much a company volunteers about itself before being asked |
Those seven add to 100, and the ordering is a judgement a reasonable person could argue with. Plenty of competing sites put catalogue size and bonus generosity at the top, because that is what readers type into a search box. The view here is that a large bonus at a company with no appeal route is a promise with nothing underneath it, while a thin catalogue at a well-regulated one costs a player an evening rather than a balance.
What a number means
Scores run to one decimal place, which is spurious precision unless somebody says what the decimals are for. Bands do that job.
| Band | Reading in plain words |
|---|---|
| 8.0 and above | A real complaints route, terms that agree with each other, and no clause a reader would be angry to find later. Nothing covered here currently sits in this band. |
| 7.0 to 7.9 | Sound, with limitations that are disclosed up front and cost money rather than protection. |
| 6.0 to 6.9 | Usable by somebody who reads the terms first. Something material is wrong, contradictory or missing, and the pages name it. |
| 5.0 to 5.9 | Problems to work around rather than budget for: punishing conditions, poor disclosure, or protection that stops where it is most needed. |
| Below 5.0 | We would tell a friend to keep their money in their chequing account. |
Guides and legal pages carry no score at all, because there is no product in them to rate.
Three things this method cannot tell you
This is the part most review sites omit, and omitting it is how an afternoon with the paperwork gets written up as though somebody had been playing. The edges fall exactly where readers care most.
How fast money actually arrives. Deposit speeds are published and quotable. Payout speeds are not published at all, and nothing here measured one. Any figure on this site about how long a withdrawal takes is either the operator’s own marketing claim, attributed to it, or a general industry range, labelled as a range. No stopwatch sits behind either.
What the support desk is like. Channels, advertised hours and the written help-centre answers can all be assessed from outside. How a ticket gets handled once a balance is in dispute cannot, and that is the moment support matters.
Whether a term is applied the way it is written. A clause demanding three times’ turnover on any deposit ahead of a cash-out request is quotable, and it is quoted. Whether it binds every account, is waived for most, or is invoked selectively stays invisible from outside. The clause gets reported; how it is used in practice is not something anybody here knows.
Those are gaps in the method, not findings about the operator. Reading them as findings would be as unfair as pretending they were measured.
Corrections
Figures on gambling sites move without notice, and some of what is published here will be stale before it is read. Anything factual can be challenged, by a reader or by the operator, at [email protected], and both get handled identically.
The sequence is fixed. The disputed number goes back to its source first, meaning the operator’s live page, since a reader in one province may be served a different version from the one captured. If the page here is wrong, it is corrected in place, with a line recording what changed, and the date on the page changes with the number. A re-check that finds nothing still updates the date, because “looked again and it had not moved” is useful information about a company that edits its terms quietly.
What no request can change is an opinion. A low weighting on catalogue size, or a verdict that one bonus tier is bad value, is an editorial position, and it stays unless the arithmetic behind it is shown to be wrong.
Where the money sits, and why it is not an input
This publication is paid by affiliate commission when a reader registers with an operator after following a tracked link. That creates a conflict, and the sensible response is to say where the conflict is fenced off rather than to pretend it is absent.
It stays out of the seven weights, written down before any page was drafted and containing no commercial term, and out of the band table, published here rather than applied privately. Commission arrives on a registration and never on a rating, so no mechanism exists by which a higher number pays better. No operator has seen a draft, approved a score or suggested a heading, and there is no price at which a disclosed contradiction comes off a page. The full commercial position sits on the disclaimer; who writes all this is on the about page.
None of that is auditable from outside, which is the honest limit of the assurance. The one test a reader can actually run is whether these pages read like reporting.