We do not give gambling companies a score out of ten. Our responsible gambling operator rating is a written evidence record, not a star rating. For each company we check five things, and we only use documents that you can open and read yourself: the law, the rules in its licence, the public record of regulator action against it, and the figures the company publishes in its own reports.
This page explains that method. We are publishing it before we publish any assessment of a named company, so you can judge the method first and the verdict second.
Why the method comes first
Most pages that rank gambling companies on “safety” are paid for by those companies. They are built to send you to a sign-up page. A method that is written after the ranking can be bent to fit it. A method published first cannot.
There is a second reason. Gambling is one of the few industries where the company is often asked to police itself. The World Health Organization is blunt about how that has gone. Its gambling fact sheet of 2 December 2024 says industry self-regulation “is a common feature of gambling regulation” but that “there are many examples of the failure of this approach”. The same fact sheet says responsible-gambling measures are “typically ineffective, particularly where the uptake of measures is optional”, and that the framing can end up “effectively blaming those who experience harm”.
So we start from a sceptical position. A company telling us it cares is not evidence. A regulator fining it, or a published number it cannot easily take back, is.
What counts as evidence
We sort every source into one of four tiers. Higher tiers beat lower ones. If a company’s own page contradicts a regulator’s decision, the regulator wins.
| Tier | What it is | Example | How we use it |
|---|---|---|---|
| 1 | Law and licence rules | Great Britain’s SR Code 3.4.3; the Malta Gaming Authority’s Player Protection Directive (Directive 2 of 2018) | Sets the floor. Meeting it earns no credit. |
| 2 | Regulator findings and decisions | The Gambling Commission’s register of regulatory actions | Strongest evidence of real behaviour. |
| 3 | Standards and audits run by someone else | GamCare’s Safer Gambling Standard; the European standard EN 18144 | Useful, but voluntary. |
| 4 | The company’s own published reports | Flutter’s Positive Impact reporting | Quoted with the limits named. |
Four things never count as a source of fact on this site: another affiliate or comparison site, a clinic or rehab marketing page, a news summary of a report we have not opened ourselves, and any page that appears to be machine-generated. Press releases count only for the figures inside them, and only when the company also publishes the underlying report.
The five areas in a responsible gambling operator rating
Each assessment answers the same five questions in the same order. We write what we found and we name the document we found it in. Where a company publishes nothing, we say so plainly. Silence is a finding.
| Area | What we check | Where we check it |
|---|---|---|
| 1. Regulatory record | Fines, warnings, suspensions, licence conditions added, and what the regulator said the failing was | Regulator public registers and enforcement notices |
| 2. Account tools | Whether deposit, loss, time and session limits exist, how fast a cut takes effect, and how much friction an increase has | The live account settings and the regulator’s technical standards |
| 3. Harm detection | Which behavioural markers the company says it monitors, and whether that list matches a recognised standard | Company disclosures compared against EN 18144 and SR Code 3.4.3 |
| 4. Disclosure | Whether the company publishes any number that could embarrass it, and whether last year’s number is still there | Annual and sustainability reports |
| 5. Money and governance | Who pays for harm prevention, and whether the company still chooses where that money goes | Levy rules and company accounts |
Area 1: the regulatory record
In Great Britain the Gambling Commission publishes a register of regulatory actions for the last three years. It lists the company, the type of action, the decision date and the outcome. It does not show the penalty amount, so we read the separate enforcement notice for that.
Two points of fairness. A three-year window means an older failing drops off the register, so we say when a finding is older than the register and link the original notice. And an absent record is not proof of good behaviour: it can also mean the company is small, new, or licensed somewhere that publishes less. We write which of those we think it is. You can see how much regulators differ in our guide to gambling regulators worldwide.
Area 2: the tools on the account
We care whether a tool works, not whether it is advertised. A deposit limit that only starts at the next calendar month is a weaker tool than one that bites immediately, even though both are called a deposit limit.
Two things carry extra weight. The first is how long a cut takes to apply, and how long an increase takes. The second is whether the company is in the national self-exclusion scheme. In Great Britain that is not optional: SR Code 3.5.5 says licensees “must participate in the national multi-operator self-exclusion scheme”, which is GAMSTOP, run by The National Online Self Exclusion Scheme Limited. The Commission has enforced it: on 3 April 2020 it suspended two operators’ licences for missing the 31 March 2020 deadline, saying the failure “placed vulnerable consumers at risk”.
The detail of each tool is covered in our guide to responsible gambling tools, and the technical floor sits in the Commission’s remote gambling and software technical standards, last updated on 31 October 2025.
Area 3: harm detection
Every large company now says it uses data to spot players in trouble. The useful question is which signals it watches, because that list is now written down by other people.
Great Britain’s SR Code 3.4.3 names seven categories of indicator that remote licensees must consider: customer spend, patterns of spend, time spent gambling, gambling behaviour indicators, customer-led contact, use of gambling management tools, and account indicators. Most of the provision took effect on 12 September 2022, with the main customer-interaction requirements following on 31 October 2023.
Europe now has a voluntary standard as well. The European Committee for Standardisation approved EN 18144 in September 2025 and the standard became available in May 2026. According to the European Gaming and Betting Association’s Annual Activity Report 2026, which began the standardisation work in 2022, it sets out nine core behavioural indicators: staking volume, speed of play, depositing behaviour, cancelled withdrawals, player-initiated contact, time spent, gambling products, safety tool usage and losses.
We compare what a company says it monitors against those lists. A company that monitors cancelled withdrawals is watching one of the clearest warning signs there is. A company whose published list is shorter than the legal minimum has told us something important about itself.
Area 4: disclosure
A number that could embarrass a company is worth more than a page of warm words. Flutter, for example, reports on its own sustainability pages that 47.3% of its global active players used its Play Well tools in 2025, in its 2025 Positive Impact Report.
We quote figures like that, and then we name their limits, because there are three.
- The definition is the company’s own. “Used a tool” can mean setting a deposit limit, or it can mean once opening an activity statement. Those are very different things.
- The numbers are not comparable between companies. Different definitions and different customer mixes mean a league table of these percentages would be meaningless. We never build one.
- Disclosure can disappear. Kindred Group used to publish, every quarter, the share of its revenue that came from high-risk players. Its old website now redirects to FDJ United, following the takeover, and that page does not carry the measure. A metric that can be withdrawn was never a guarantee.
Independent certification sits between a company’s word and a regulator’s finding. GamCare’s Safer Gambling Standard assesses operators against ten areas, covering matters such as sharing best practice, advertising and promotion, and the protection of children and young adults. We record whether a company holds it. We do not treat it as a substitute for a clean regulatory record.
Area 5: money and governance
For years, British gambling companies chose how much to give towards research, prevention and treatment, and often which charity received it. The statutory levy ended that choice. The Gambling Commission’s guidance says the leviable rates “range from 0.1 percent to 1.1 percent” depending on the licence sector.
The government’s response to the levy consultation sets the split: 20% to research, 30% to prevention, and 50% to treatment, with that treatment half going to NHS England and the appropriate bodies in Scotland and Wales to commission services. Research is led largely by UK Research and Innovation.
This matters for a rating. “We fund treatment” used to be a claim a company could make about its own generosity. Under a compulsory levy it is mostly a description of paying a bill. So we give little weight to levy payments, and more weight to what a company does that the levy does not require.
What we will never publish
- Lists or links for sites that sit outside a national self-exclusion scheme. People search for these at their lowest moment, and the only honest answer is that such a site has removed the one protection that was working.
- Affiliate links, tracking links, sign-up pages or bonus pages. We name companies. Where we link one, it is to its own responsible-gambling or corporate page, marked so search engines do not pass credit.
- Strategy, systems or how-to-win content of any kind.
- Quizzes that claim to diagnose you.
- Counts of “addicts” worked out from screening scores. The Problem Gambling Severity Index is a screening questionnaire, not a diagnosis. The Gambling Commission’s own guidance on using its survey statistics lists “as a measure of addiction to gambling” among the uses its figures are not for, and says the PGSI “should not be confused with a measure of gambling addiction”.
That last point also governs how we handle numbers. The same guidance warns against using its statistics “to provide direct comparisons with results from other gambling or health surveys”, because “differences in survey design, methodology, and sampling can lead to misleading comparisons”. Every figure on this site carries its survey name, year, instrument and mode, and we explain the method before we put two numbers near each other. Our articles on global gambling harm in numbers and how many people have a gambling problem set out why that caution matters.
Who pays for this site
We take no money from gambling companies, from their affiliates, or from treatment providers. We carry no gambling advertising and no affiliate links. No company can pay to be assessed, to be assessed sooner, or to have an assessment changed. Our funding and independence page sets out the detail in full, and our editorial policy explains how pages are written and reviewed.
We expect to get things wrong. Rules change, helpline numbers change, and companies restructure. When we find an error we correct the page, date the correction and keep the old wording visible, as described in our corrections policy. If you spot something out of date, please tell us.
What this method cannot tell you
Three honest limits.
We cannot see inside the system. We can read what a company says its risk model watches. We cannot see the thresholds, how many accounts it flagged, or what happened next. Nobody outside the company and its regulator can.
Open companies look worse than closed ones. A company that publishes an uncomfortable number gives us something to criticise. A company that publishes nothing gives us nothing. We correct for this by treating non-disclosure as a finding in its own right, but it remains the weakest part of any method built on public documents.
The evidence is unevenly spread. British, Maltese and Nordic regulators publish a great deal. Others publish very little. So our assessments are better evidenced in some markets than others, and we say which.
None of this tells you whether a company is safe for you. No rating can. If gambling has stopped being something you choose, the useful step is not comparing companies but blocking all of them: start with our guides to self-exclusion and how to get help. If you want the background to the term itself, see what responsible gambling means and our explainer on gambling addiction.
This methodology is reviewed every year, and sooner if a regulator changes a rule it depends on. Every assessment we publish will link back to this page, and will name the version of the method used.
Last reviewed: 1 October 2026. Sources checked on this date.