How We Test Online Casinos
Reviews on this site are produced from a real test of the operator, not from press releases or marketing material. That distinction matters because the gap between an operator’s self-description and its actual behaviour is often where the useful information lives. The process below sets out, stage by stage, what is checked, what is recorded, and how each stage feeds the rating that appears on the published review.
A full test of one operator runs over five to ten days — long enough to put a real deposit through, claim and use the welcome bonus, request a withdrawal and time it end-to-end, and put the customer-support team under specific product questions. Costs are real: each test involves a real deposit, real wager and real exposure to losses on the games tested. The results of a single test are not extrapolated; if a payout window varied between two test runs, both observations are noted in the review.
Stage 1 — Pre-analysis
Before any account is opened, an operator is assessed on the desktop side. The work in this stage answers a single question: is this operator credible enough to put real money into for the test?
The licence is the first item checked, against the issuing regulator’s public register rather than the operator’s own footer. Curaçao master and sub-licences, Malta Gaming Authority records at mga.org.mt, UK Gambling Commission entries at gamblingcommission.gov.uk and equivalent registers in other jurisdictions are consulted at source. ACMA’s Australian register at acma.gov.au is also checked for any record of complaints, investigations or blocking action against the operator under the Interactive Gambling Act 2001. A licence number that does not return a current, active record on the regulator’s register fails the pre-analysis. The operator’s ownership and corporate history are looked up: who runs it, what other brands the same operator has run, whether previous brands have closed under disputed circumstances. Independent watchdog platforms — AskGamblers, Casino Guru, Trustpilot, established Reddit gambling communities — are scanned for recurring complaint patterns, with attention to pattern frequency rather than individual complaints. Terms and Conditions are read in full at this stage, looking specifically for clauses that would make a payout dispute hard to win: dormant-account confiscation rules, “at our discretion” clauses, bonus terms that materially extend wagering through deposit-plus-bonus base, max-bet rules during bonus play that are easy to breach accidentally.
Pass criteria: licence verifies on the regulator’s register; ownership traces to a parent group with at least two years of operational history under that group; complaint patterns on independent platforms are at or below the industry average for offshore-licensed operators serving Australia; the T&C contain no clauses that would obviously preclude a fair payout dispute.
Stage 2 — Registration
An account is created on the operator’s site as an ordinary new player — no special credentials, no negotiated terms. The signup flow is timed and the screen path recorded.
What is measured: total time from clicking “Sign up” to a logged-in account ready for deposit; number of mandatory fields; presence and clarity of welcome-bonus opt-in (a bonus that is auto-applied without the player understanding the wagering attaches is a problem, not a feature); presence and clarity of identity-verification requests (whether they are triggered at signup or at first withdrawal — both are valid, but the cleaner pattern is upfront KYC); availability of self-restraint settings during the signup flow. Errors, broken fields and any forced upsells are recorded against this stage.
Pass criteria: registration completed in under five minutes for a new player without prior account history; bonus opt-in is explicit and the wagering disclosure is in the signup flow rather than buried; deposit and self-exclusion limits can be set before any deposit is made.
Stage 3 — Deposit
A real deposit is made — typically between A$30 and A$100 — through more than one payment method. Where PayID is offered, it is the first method tested, because it is the dominant Australian-specific channel and a major differentiator between offshore operators serving Australian audiences. Visa and Mastercard, or their bank-issued debit equivalents, are typically tested as second methods. Where cryptocurrency is offered, a small Bitcoin or Ethereum deposit is included to confirm the on-chain confirmation and credit timings.
Each method is timed from initiation to credited balance. Minimum deposit, maximum single deposit, and any per-day deposit limits are recorded. Currency conversion fees, where the operator banks in a currency other than AUD and converts on receipt, are calculated and recorded; some operators absorb these and some pass them through. The cashier interface itself is assessed for clarity: an Australian player should be able to find the deposit screen, choose a method, and confirm the transaction without re-reading any element. Friction in this flow is recorded as a con on the published review even when the deposit eventually goes through.
Pass criteria: at least five deposit methods supported; minimum deposit at or below A$20 for the cleanest entry point; PayID supported with crediting under sixty seconds; at least one zero-fee method available across all major channels; clear AUD denomination throughout the cashier.
Stage 4 — Bonus mechanics
If the welcome bonus is accepted, its mechanics are worked out in numerical detail rather than described in marketing language.
The wagering multiplier is read from the operator’s own bonus terms, including whether it applies to bonus only or to bonus-plus-deposit (the second is materially more expensive in expected loss; the difference between “wagering 35x bonus” and “wagering 35x bonus-plus-deposit” is enormous and is one of the most common ways operators make a flat headline figure look worse than it reads). The maximum bet during bonus play is recorded; a bet over the cap is normally voided, and the cap on offshore casinos is often A$5 per spin, which is restrictive on high-volatility slots. The game-weighting schedule is recorded: slots typically count 100% toward wagering, table games often 10% to 20%, live games sometimes 0%. Cashout caps on bonus winnings — some operators cap the maximum payout from bonus winnings at a multiple of the bonus, which can produce a worse-than-expected outcome on a lucky run — are recorded.
The expected cost of completing wagering is calculated against a representative game. For a 100% bonus up to A$500 with 35x wagering on bonus-only at a slot with 96% RTP, the expected loss across the wagering volume is approximately 4% of A$17,500, or A$700 — before variance, which can be material in either direction. The arithmetic is shown explicitly in the published review where it adds clarity, not as an exercise.
Pass criteria: wagering at or below 35x on bonus-only base; minimum 14-day window to complete wagering; max bet during bonus at A$5 or above; cashout cap on bonus winnings either absent or set at a sensible multiple of the bonus; full disclosure of the game-weighting schedule on the operator’s own bonus page.
Stage 5 — Gameplay
Three categories of game are tested in every full review: slots (a minimum of ten different titles across at least four providers), table games (roulette, blackjack, baccarat, with at least one variant of each), and live casino (where offered). The set is chosen to overlap with what an Australian player is likely to look for: at least three Pragmatic Play slots including Sweet Bonanza or Gates of Olympus; at least one Hacksaw Gaming title for the high-volatility category; one Nolimit City title for similar reasons; at least one classic NetEnt slot such as Starburst or Gonzo’s Quest; and the live-casino lobby tested across at least three Evolution tables.
What is measured: load time for each game (target under three seconds on a stable connection); whether the demo mode is available without registration (useful diagnostic for operator caution); stability under continued play (any disconnects, freezes or replay glitches are recorded); accuracy of the game catalogue against the operator’s marketing claims; presence of any provider conspicuous by its absence (for example, a casino marketing a strong table-games offering with no Playtech or Microgaming live tables raises a question worth answering in the review); video quality and dealer professionalism on the live tables.
Pass criteria: total catalogue of at least 500 games with at least 10 distinct providers; live casino present and stable; load times under three seconds for the great majority of titles tested; no significant gap between marketed catalogue and actual lobby.
Stage 6 — Withdrawal
This is the highest-weighted stage of the test. It is where most of the meaningful difference between operators surfaces, and where the review’s findings are usually most useful to readers.
A withdrawal is requested for an amount that includes some of the original deposit and some of the winnings, on the same payment method used for deposit (where the operator’s rules allow). The full timing is recorded: from clicking “Withdraw” to the funds appearing in the receiving account, broken into the operator’s internal pending period (the time during which the request can still be cancelled) and the network processing time after the operator releases the payment. Any additional verification requested at this stage is recorded: many operators run their full KYC at first withdrawal rather than at signup, which is acceptable in itself but is a friction point that the player should know about. Any per-transaction or per-day or per-week withdrawal caps are recorded, along with any reverse-withdrawal mechanism (the ability to cancel a pending withdrawal and put the funds back into the play balance, which is a known harm-amplifying feature and weighs against the operator on the responsible-gambling layer).
Pass criteria: total time from request to received funds under 24 hours for PayID and crypto, under three working days for card and bank transfer; no fees for ordinary withdrawal volumes; weekly cap of at least A$5,000 (operators that cap withdrawals lower than this are flagged as a structural issue regardless of how the rest of the review reads); no reverse-withdrawal feature, or, where one exists, a clearly accessible “lock” option to disable it.
Stage 7 — Customer support
Support is contacted through every channel the operator offers — live chat first, email second, telephone if available. Each channel is timed for response. The questions asked are deliberately specific, designed to test product knowledge rather than producing scripted answers: the wagering treatment of split bets on blackjack during a bonus, whether reality-check intervals can be set below 30 minutes, whether the operator honours BetStop entries placed during account life, the actual processing window for a withdrawal at 11pm on a Sunday in Melbourne. Generic, scripted responses score low; substantive answers from agents who understand the product score high.
Pass criteria: live chat available 24/7 with a response time under three minutes during business hours; email response under 24 hours; English language available natively; agents able to answer specific product questions without redirecting to terms-and-conditions pages.
Stage 8 — Mobile and security
The operator is tested on mobile (iOS and Android) at every stage above, not as a separate after-the-fact pass. Cashier flows, bonus opt-in flows, KYC document upload and live-game streaming are all checked on a phone-sized viewport. Where a native app exists, it is installed and tested on iOS and Android; where one does not exist, the responsive web experience is tested as the primary mobile path.
Security is reviewed on the desktop side: certificate validity (TLS 1.2 or above), absence of mixed-content warnings, presence of two-factor authentication for the player account, presence of session-timeout behaviour, completeness of the privacy and cookie policies on the operator’s own site, presence and quality of responsible-gambling tools (deposit limits, self-exclusion, reality checks). Operators without native two-factor authentication on the player account are flagged as a structural concern; that is a low bar and operators who have not crossed it should know better in 2026.
Pass criteria: full functionality on mobile including cashier and live games; valid TLS certificate with no warnings; two-factor authentication available on the player account; full set of self-restraint tools accessible from the player account, not buried in support tickets.
After the test
Once all eight stages are complete, the testing log is converted into a rating against the eight-criterion framework on the How We Rate page. The published review surfaces the rating both as a headline number and as a per-criterion breakdown, so a reader can see whether a high overall rating is being supported by every component or carried by a few stand-out criteria. The pre-publication fact-check, described on the Editorial Policy page, runs after this stage to confirm that the operator’s current published terms still match the testing log.
Reviews are then re-tested every three to six months, or earlier if a triggering event — a change of ownership, a new licence, a material bonus rewrite, a wave of complaints on independent platforms — arrives in the meantime. Each re-test repeats the relevant subset of the eight stages and updates the rating. The mechanism for those updates is part of the Editorial Policy.