RollXO scores 3.4 out of 5. This page explains where that number comes from, so you can disagree with it precisely rather than vaguely.
A number is honest about being reductive. The alternative - "a solid choice for Aussie players!" - hides the same reduction behind adjectives. The number is only useful if the weights behind it are public, so they are below.
The score is not a recommendation. A 3.4 does not mean "play here". It means the operator is what it is, measured against a framework that does not change between reviews to suit the commercial relationship.
| Criterion | Weight | RollXO |
|---|---|---|
| Licensing and corporate transparency | 20% | 3.0 |
| Terms fairness (confiscation, caps, max-bet) | 20% | 2.5 |
| Payouts: speed, methods, ceilings | 15% | 3.5 |
| Bonus honesty: headline vs achievable | 15% | 2.0 |
| Dispute record | 10% | 3.5 |
| Game library and providers | 8% | 4.5 |
| Support | 7% | 4.0 |
| Responsible gambling tooling | 5% | 2.5 |
Licensing and terms carry 40% between them because they decide what happens on the worst day, and the worst day is the only day a review is worth anything. The game library carries 8% because every casino on this platform has the same one.
Licensing (3.0). The licence resolves in the regulator's register and the corporate identity checks out - that is the floor cleared. It loses points because the register's public edition shows a term that has lapsed into an automatic provisional conversion rather than a fresh assessment, and because the operator publishes no full licence anywhere a player can read it.
Terms fairness (2.5). Two clauses permit confiscation at the operator's discretion, and neither has a numeric threshold. The €50 cap on no-deposit winnings is punitive rather than protective. Progressive jackpots being exempt from the caps is the one genuinely generous term and it pulls this back from a 2.
Bonus honesty (2.0). The lowest subscore, and the reason is arithmetic rather than opinion: collecting the advertised $15,000 requires depositing $21,000 and turning over $750,000, and the resulting balance exceeds what the cashier releases in a month.
Payouts (3.5). Crypto inside an hour, no fees, weekend processing and a €20 minimum are all genuinely good. The ceiling and the unpublished, contradictory caps are what hold it at 3.5.
Each subscore multiplied by its weight, summed, rounded to one decimal. (3.0×0.20) + (2.5×0.20) + (3.5×0.15) + (2.0×0.15) + (3.5×0.10) + (4.5×0.08) + (4.0×0.07) + (2.5×0.05) = 3.4.
No hidden adjustment, no editorial thumb. If you disagree with the 3.4, disagree with a weight or a subscore - both are on this page, which is the point of publishing them.
Some findings cap the score regardless of the arithmetic. A licence that does not resolve in a register caps the total at 1.5. A documented pattern of confiscating verified winnings caps it at 2.0. Readmitting a self-excluded player caps the responsible gambling subscore at 2.5 no matter what tooling the account area advertises - which is precisely why that subscore sits where it does here.
The weights are a judgement, not a physical constant. Somebody who only ever plays $20 sessions and never wins more than a few hundred dollars would reasonably weight withdrawal ceilings at zero, and for them this score is too harsh. Somebody chasing a five-figure win would weight them higher, and for them it is too kind.
The larger limit is the one the method page admits: this framework scores documented evidence, not a live cashier. The payouts subscore rests on published processing windows rather than on a withdrawal we clocked ourselves. That is a real weakness and pretending otherwise would be worse than carrying it.
