How we score: six criteria, published weights, three that move

The rubric in one paragraph

Six criteria, one hundred points, fixed weights, identical for every operator. Regulatory standing in your province carries 25 points; the complaint and escalation route 20; safer-gambling controls and self-exclusion 20; player fund protection 15; game fairness verification 15; and transparency of terms and ownership 5. Three of the six change with your province, and together they are 65 of the 100.

Why the first three carry sixty-five points

Because they are the ones that decide what happens when something goes wrong, and everything else is a description of how pleasant the site is when nothing does.

A regulator that can act, a complaint route that reaches it, and an exclusion that binds across a market are the mechanisms that convert a promise into an obligation. An operator with excellent games, fast payouts and no regulator behind it has given you a good experience and no recourse. We weight for the bad day rather than the good one.

What each criterion actually measures

Regulatory standing (25). Is this operator on the register that governs play where you live? Read from the register itself, not from the operator's claim. Full marks for a provincial register entry; a small residual score where no provincial framework exists at all, because nobody can earn what is not on offer.

Complaint and escalation route (20). When support says no, is there a body that can compel a remedy? Full marks where a provincial regulator sits above the operator. Partial credit for an offshore licence with a genuine dispute process — the Kahnawà:ke Gaming Commission staffs a full-time dispute officer, and Malta runs an adjudication route. Very little where escalation ends with the operator.

Safer-gambling controls and self-exclusion (20). Twelve of these points sit with a binding province-wide exclusion register, and eight with the operator's own deposit, loss and time limits. The split is a judgement: an exclusion that covers a market is worth more than a good limit screen at one site.

Player fund protection (15). Are balances ring-fenced from operating capital, and is that a licence condition or a promise? The distinction is the whole criterion.

Game fairness verification (15). Independent testing by a named accredited laboratory, with a certificate we can open. Reading a claim about testing is not the same as reading the certificate.

Transparency (5). Who holds the licence, and are withdrawal terms stated where a reader would look rather than buried. The smallest weight, and the easiest for an operator to fix.

Where the numbers come from

Primary documents only: provincial register entries, testing laboratory certificates, and the operator's own published terms. We do not score from other review sites, and that is a practical rule rather than a fastidious one — a false claim that the testing house eCOGRA closed in 2021 circulates widely in casino content, and any rubric assembled by reading affiliates would have absorbed it and penalised every operator using them.

Evidence is scored in three tiers. A document we have opened earns the full weight of a criterion. A binding requirement we can read in a regulator's published standards, together with a register entry proving this operator is bound by it, earns most of the weight but never all of it. An operator's own assurance, or nothing at all, earns zero.

That is why no fairness figure on the site reaches fifteen. Certification by an AGCO-registered laboratory is a written condition of Ontario registration, so a registered operator sits in the middle tier at eleven; the top four points wait on a certificate we can open, and none of the registered laboratories publishes one openly. We would rather publish an honestly incomplete score than a flattering assumption.

What a protection ceiling means

On province pages you will see a ceiling as well as operator scores. The ceiling is a statement about the province rather than about any company: the highest total anyone could reach for a player living there.

In Ontario and Alberta the ceiling is 100, because all six criteria are attainable. In a province with no commercial framework it stops around 60, because the regulatory and escalation points are unavailable to everyone and most of the exclusion points go with them. No operator can close that gap by being better run.

One result that looks like a mistake and is not

An unregistered operator scores three of the twenty-five regulatory points in a province that has a framework, and five in a province that does not. So the same operator can score slightly lower in Alberta than in Manitoba, and the first time you notice it, it reads like a bug.

It is deliberate. In Manitoba nobody can be registered, so being outside a framework is the only available state and says nothing about the operator's choices. In Alberta a framework exists, other operators went through it, and this one did not — it is trading to Albertans from outside a system it could have joined. That is a worse position for the player than the same operator in a province where no system exists, because in Alberta there was a registered alternative and the unregistered site is competing against it without accepting the obligations.

The effect is small on purpose: two points, not ten. It is not a punishment, it is the honest observation that "unregistered" means something different where registration is possible. If you would rather it did not, the safety score explorer lets you set the regulatory weight to zero and see what the table looks like without it.

What we refuse to score

Offer sizes, promotional generosity and loyalty schemes. Not because they do not matter to players, but because they change monthly and say nothing about safety — and in Ontario, advertising them to residents is not permitted at all.

Observed payout speed, as a ranked table. Keeping one honest would require continuous re-measurement, and a stale speed claim is worse than none. We publish stated windows with the date we checked them instead.

Reputation, brand recognition and how long an operator has traded. These measure marketing budget and survivorship rather than protection.

Game selection and software quality. Real preferences, not safety.

What we are least confident about

The weights are a judgement and could reasonably be set differently. The most arguable decision is the twelve-to-eight split inside the safer-gambling criterion: we treat a province-wide exclusion register as worth more than an operator's own tools, which penalises a genuinely careful operator in a province that has not built a register. We think that is right, because the register is the thing that works when a player is not thinking clearly. It is also the decision we would most expect to be challenged.

The second is giving regulatory standing a full quarter of the total, which means a registered but mediocre operator outscores an excellent unregistered one in Ontario. That is intentional and it is still a choice.

If you disagree with either, the weights are published precisely so you can see how much of the total your disagreement moves. The masthead says who owns them and who signs a correction. Tell us — that is a more useful message than a report that a score feels wrong.

When a score changes

A new or lapsed laboratory certificate, a changed licence condition, a rewritten withdrawal term, or a registration gained or lost in a province. Scores also change when we verify something we previously could not, which currently is the most common reason.

We publish corrections rather than editing quietly, so a score that moves has a visible reason. See corrections and our sources.