The six-stage test protocol
Every test follows the same six stages, in the same order, using the same approximate deposit amounts. Consistency is the point — we're not trying to capture every possible scenario, we're trying to produce tests that are directly comparable across operators. If one casino pays out in 14 hours and another in 72, the difference should reflect the operator, not the tester's behaviour.
Fresh account creation
A new account registered under genuine Australian identity details. No shared accounts, no VPNs, no prior history with the operator. The account is funded from a clean payment source that has no prior relationship with any casino in our testing pool.
Deposit between AUD 100 and AUD 200
We deliberately stay within a test amount range that triggers baseline KYC rather than enhanced due diligence. This produces results that reflect the experience of most casual Australian players, not high-rollers who sit in dedicated VIP review queues.
Real play session of 40+ minutes
We wager the full deposit amount at minimum, across eligible slots, without claiming bonuses unless we're specifically testing the bonus cycle. Minimum 40-minute session duration to avoid tripping rapid-turnover fraud heuristics.
Withdrawal request to the fastest documented method
We test the operator at its best rather than its worst. Whichever method the casino advertises as fastest for Australian players — Skrill, POLi, Bitcoin, USDT — is the method we use for the primary test, with additional methods tested when resources allow.
Timestamp and document every stage
We record the email timestamps, cashier status changes, support transcripts, and screenshots at each transition. Every published timing figure can be traced back to a specific piece of recorded evidence archived against the review.
Cross-check against 100+ player reports
Our test result is one data point. We cross-reference it with at least 100 verified player reports from Trustpilot, CasinoGuru, and AskGamblers to identify the range of experiences rather than the outlier. If our test falls outside that range we investigate before publishing.
The scoring rubric
Every casino is scored on six weighted criteria that sum to 100 percent. The weights reflect what we believe matters most to players — speed is the largest single factor, but process transparency and evidence of honest payment handling together outweigh it. Here's the full rubric:
| Criterion | What it measures | Weight |
|---|---|---|
| Payout speed | Median withdrawal time from request submission to funds received, measured against our own test and a minimum of 100 verified player reports. | 30% |
| KYC reasonableness | Is the verification process proportionate to the withdrawal amount? Does the operator trigger enhanced due diligence at appropriate thresholds, or at every cash-out regardless of size? | 20% |
| Payment method support | Does the operator support the methods Australian players actually use? Is there symmetry between deposit and withdrawal methods, or does the cashier create traps? | 15% |
| Terms transparency | Are the withdrawal-relevant terms clearly disclosed before the first deposit, or buried in fine print? Is the relationship between bonuses and cash-out rights explicit? | 15% |
| Complaint resolution | How does the operator handle disputes at third-party mediators? What percentage of player complaints resolve in the player's favour when the case is legitimate? | 12% |
| Fee structure | Are withdrawal fees disclosed clearly? Are there hidden inactivity fees, currency conversion fees, or minimum-balance penalties that reduce effective payouts? | 8% |
| Total | 100% | |
Each criterion is scored on a scale of 0 to 10. The weighted sum produces the final rating out of 10 that appears in the verdict box on every review. We don't round to whole numbers — a 7.8 and a 7.4 are meaningfully different ratings in our system, and we treat the decimals as signal rather than noise.
Worked example: how RocketPlay got 7.77
To show the rubric in action, here's the full scoring breakdown for RocketPlay from the current test cycle. This is not a summary — it's the actual calculation that produced the published score.
RocketPlay final score, criterion by criterion.
The single biggest drag on RocketPlay's score is the terms transparency criterion — specifically the withdrawal reversal feature that creates the player-harm pattern we documented in the full review. On every other criterion RocketPlay scores in the high 7s or 8s. If we removed the reversal trap from the equation, the score would rise to roughly 8.2. The full 0.43-point difference is exactly the cost, in our scoring system, of one poorly-disclosed cashier feature.
How we measure timing
Timing is the most important criterion in our rubric, and it's also the one where sloppy methodology produces the most misleading numbers. To make our timings comparable across operators, we split every withdrawal into six named phases and record the boundary timestamps for each.
| Phase | Code | Starts when | Ends when |
|---|---|---|---|
| Submission | E1 | User submits withdrawal request in cashier | Casino sends "pending" confirmation email |
| KYC trigger | E2 | KYC request email from casino | User uploads all requested documents |
| KYC review | E3 | User completes document submission | Compliance team approves the documents |
| Payment queue | E4 | KYC approved, withdrawal moves to pending payments | Payments team marks withdrawal as "processing" |
| Settlement | E5 | Casino dispatches payment to processor | Funds visible in user's receiving account |
| Total | E6 | User submits request (E1 start) | Funds received (E5 end) |
The "payout time" figure in every review is the E6 total — start of E1 to end of E5, inclusive. When we compare operators, we compare like-for-like on E6. The individual phases matter when we're diagnosing why a particular operator is slow: an operator with a fast E5 but a slow E3 is bottlenecked on compliance, not on payment rails.
The 90-day retest cycle
Every operator on this site is scheduled for a full retest every 90 days, or sooner if significant changes are reported. We re-run the complete protocol — fresh account, fresh deposit, fresh withdrawal — and publish any change to the rating along with the date the retest was completed. If a review hasn't been updated in more than 90 days, a visible banner appears at the top of the page warning readers that the data may be stale.
The quarterly rhythm exists because casino terms, KYC policies, and payment processor relationships change constantly. A review that was accurate in January can be outdated by April. Our first retest cycle, in August 2025, led to two score revisions downward and one upward — the downward revisions were due to operators slowing their payment queues, the upward revision was due to an operator simplifying their KYC protocol. Without the retest discipline, those changes would have been invisible to readers until a player complained publicly.
What this methodology doesn't capture
Every methodology has blind spots. Here are ours.
- Single-tester bias. Our tests are run by one researcher with a specific risk profile, playing style, and payment method preference. A player whose profile differs meaningfully from ours may have a different experience at the same operator.
- Small sample per test. One $150 withdrawal doesn't capture the full range of possible timings at any given casino. We mitigate this with 100+ cross-referenced player reports, but the primary timing figure is still a single data point.
- Baseline tier only. We test from fresh, unverified accounts to capture the experience most players actually have. We don't test VIP or high-roller queues, which often pay faster but are not representative of typical users.
- Crypto volatility window. For crypto-native operators, the "payout time" we measure includes the crypto-to-AUD conversion step we recommend, but price volatility during that window is not separately scored. A 3% price move during the conversion can meaningfully change the effective payout.
- Jurisdiction drift. The regulatory status of Australian-facing online casinos is not static. Operators that are accessible today may be blocked tomorrow, and vice versa. Our reviews are accurate at the time of publication but we cannot guarantee continued access.
- We can't test what we can't access. Some operators restrict new registrations from certain regions or require invitation codes. We note these limitations explicitly and do not attempt to circumvent them.
Methodology changelog
The methodology on this page is a living document. Here's every meaningful revision since the initial publication:
Added crypto-to-AUD conversion timing
Introduced specifically for the Stake review, the crypto conversion workflow is now counted as part of the total payout time for crypto-native operators. This is a more honest comparison with fiat-rail operators, where the "time to AUD in bank account" is the functional measurement.
Six-criterion weighted rubric formalised
Replaced the earlier three-criterion rubric with the current six-criterion system. The additions were terms transparency, complaint resolution, and fee structure — all of which were implicit in v1 scoring but inconsistently weighted between reviewers.
E1–E5 phase timing framework introduced
Added the five-phase breakdown of withdrawal timing that lets us diagnose where a slow operator is actually bottlenecked. Previously, reviews only reported total time, which made diagnostic comparisons difficult.
90-day retest cycle codified
Formalised the quarterly retest discipline that had been running informally since launch. Every operator is now explicitly scheduled for retest and the stale-review banner was added to the template.
100-report cross-reference requirement
Added the explicit requirement that every published review must cross-reference our test result against at least 100 verified player reports from multiple platforms. Previously the cross-reference was done but not standardised.
Initial methodology published
The original three-criterion rubric (speed, KYC, support quality) and six-stage test protocol, drafted during the publication's founding week. Most of the current methodology descends directly from this version.