An eight-criterion weighted framework, where every score is what a worksheet returns rather than a hunch. This page sets out the weights, the criteria, and what actually distinguishes a 7.5 from an 8.2 in practice.
Handing out a star rating is easy; defending one is hard. The number crowning the KingBet9 review — 4.3 out of 5 — is the weighted average of eight subscores, each built from the test data collected over the two-week cycle documented at how we test casinos. The weights are set in advance and don't shift per review, and the criteria are the same from one operator to the next, which is how the system stays comparable over time.
Locking the weights down before the test matters. Were they free to move review by review, a reviewer could push a borderline casino either way simply by loading weight onto the criterion it happens to handle best. Fixed weights are the chief reason the cons list on the KingBet9 page reads as plainly as it does.
| Criterion | Weight | What it measures |
|---|---|---|
| Safety & Licensing | 20% | Licence validity, TLS, 2FA availability, T&C fairness, dispute route |
| Withdrawals | 15% | Processing time, rails available, caps, consistency across tests |
| Bonuses & T&C | 15% | Wagering math, max-bet rule, max-cashout cap, game contribution transparency |
| Game Library | 12% | Provider mix, title count, live dealer breadth, mobile parity |
| Payments (deposits) | 10% | Number of rails, AU-specific methods (PayID), deposit speed, fees |
| Customer Support | 10% | Live chat wait time, agent knowledge, hours, channels |
| Mobile Experience | 8% | Browser performance on iOS and Android, mobile cashier, load times |
| Responsible Gambling | 10% | Deposit limits, self-exclusion flow, session alerts, enforcement |
Safety and licensing carries the single biggest weight, because the worst outcome at an unlicensed or abusively licensed operator — a seized balance with nowhere to escalate — is far worse than a slow withdrawal or a thin bonus. Withdrawals and bonuses share second place, since those are the two points where an offshore casino most often skims value from players who skipped the fine print in the T&C.
Responsible gambling holds a real 10% weight, because a site that deliberately buries self-exclusion isn't one worth pointing readers toward. This criterion is also the chief reason the framework is held against the current Google Quality Rater guidelines and the AU responsible-gambling support structure.
Weighed against four inputs. Is the licence active on the regulator's register? Is the corporate licensee the entity that actually runs the site? Do the T&C hold any red-flag clauses (unilateral T&C changes with no player notice, confiscation for "bonus abuse" left undefined, dormancy fees inside 90 days)? Is the dispute escalation route written down? The rating drops steeply — a full point or more — if any of those come up short.
This is test data, not marketing. Approved-to-account time is clocked to the minute on at least two rails. A site that pledges "within 24 hours" in the T&C and delivers in 2h 39min on PayID takes full marks here; one that pledges the same and crawls to 36 hours loses the subscore whatever the cover copy promises. Ceilings count too — a A$10,000 weekly cap costs a fraction of a point for a site aimed at mid-volume players.
Scored on the math, not the headline. A 125% match to A$1,500 at 40× wagering on the bonus portion demands A$5,000 of turnover before a withdrawal. At 96% RTP that's a A$200 expected theoretical loss — larger than the bonus itself — and the score owns that rather than hide it. Sites running 50× on the combined deposit-plus-bonus fare worse, while zero-wager free spins — like the Dragon Pearls spins logged in the bonus section of the KingBet9 review — pick up a real positive adjustment.
Not the headline count — the mix. A library of 3,000 titles from second-tier studios scores under 1,500 titles spanning Pragmatic Play, Play'n GO, Hacksaw, Nolimit City and a working Evolution Live feed. Evolution or Pragmatic Live in the lobby is a genuine lift; their absence a genuine knock. Mobile parity — the same catalogue on phone and on desktop — also feeds in here.
AU-specific rails count. Two-way PayID is a positive worth real weight — rare on Curaçao offshore sites and the fastest domestic option when it works. The lack of Apple Pay and POLi is noted, though POLi's 2022 shutdown is no fault of the operator. Crypto support is a plus for readers who want it, and Neosurf running deposit-only is flagged for hemming in the withdrawal path when players aren't careful.
Agent answers specific T&C questions without reading a script? Plus. Agent pastes a link and hands off? Minus. 24/7 live chat genuinely 24/7? Plus. Phone line missing entirely? Small minus — most AU players do not miss a phone line, but some do, and the option is a real one.
Tested on two real devices over two networks. A pokie that loads in 3 seconds on home Wi-Fi and 5 seconds on 4G is acceptable. Anything over 8 seconds on 4G is a problem, because a lot of AU mobile play is on 4G.
The deposit-limit enforcement test counts. A site that lets you set a A$100 daily cap and then quietly takes a A$150 deposit fails this criterion. The self-exclusion flow counts too — one click and a confirmation beats five screens and a "cooling-off before reactivation" that flips itself back on after a week.
Each subscore is marked 1–10. The headline figure is the weighted average, rounded to the nearest tenth and then mapped onto a 5-star display where one point is worth 0.5 stars. KingBet9's present 4.3/5 traces to a weighted 8.6/10 across the eight criteria.
I don't publish the worksheet with each review — no one's asked, and it would clutter the page — but it lives in the editorial files and the breakdown is yours on request via the editorial contact. If the headline score looks out of step with the narrative below it, ask and the subscores are on their way.
Some findings trigger a fixed downgrade independent of the weighted score. Each of these is a question of trust, and the framework refuses to average them away:
KingBet9 did not trigger any of these during the test cycle. If they do in a future re-test, the score will move, the cons list will lengthen, and the update will be documented by date at the top of the review in line with the editorial policy.
No eight-criterion framework can pin down every nuance of every operator. The weights presume an average AU player: mid-stakes pokies, the odd live-dealer round, cards and PayID ahead of crypto, mostly on a phone. A high-roller chasing seven-figure progressives weights Game Library differently than I do, and a crypto-only player leans harder on Payments than PayID. The framework is tuned to the median reader, not to everyone.
The framework is also revised when the market changes. The addition of "AU-specific rails" to the Payments criterion came after PayID availability at offshore sites became meaningful. When the Curaçao LOK reform finishes its transition, the Safety & Licensing criterion will likely add a new sub-input. Changes are versioned and announced; the last revision date is at the top of this page.
