The RTP printed on the tile, and what it does not promise

Where the number on a BK8 tile comes from

Open the Slots tab in our Cambodian lobby and the figure is already there. We print RTP directly on the game tiles, next to a New or Hot badge and the studio name — on the game and provider pages rather than on the homepage. Our Nextspin page reads exactly that in its Top Games row: New, Pac-Man's Glory, Nextspin, RTP 96.95. Drop into the main grid below and the tiles carry the name and the return only, with the provider line stripped out. The number we print is the number the studio supplies for the build we actually run — not a house average, not a category figure, and not something borrowed from a review site.

• RTP is theoretical and long-run. It is calculated across a simulated volume of spins far larger than any person will play in a lifetime.
• It is a property of the build, not of the casino. Two operators running different configurations of the same title will publish different numbers, and both can be correct.
• It is not a payout schedule. Nothing in 96.92% tells you how the 96.92% arrives — in steady small wins, in one enormous spin, or mostly inside a bonus round you may never trigger.

Only two of the studios behind our most-opened slots publish a return at all. PG Soft prints it, and prints considerably more besides. Pragmatic Play prints it. JILI publishes nothing — no percentage appears anywhere on its own site, not on the game pages and not in the site's own player JavaScript. The 97% figure attached to JILI slots across affiliate blogs came from affiliates, not from the company that wrote the maths. Below is what to read in its place.

The highest published RTP in our lobby, title by title

Six titles on our shelf carry a return published by the company that built them. Ranked, they run from 96.95% down to 96.48% — a spread of under half a percentage point across the whole group, which tells you something immediately useful. Anyone selling a highest RTP slot as though the distance between the top and the bottom of this table decides an evening is selling you a rounding error.

The column that separates these games is the last one, not the third. Mahjong Ways 2 and Lucky Neko both advertise 100,000x; Fortune Tiger advertises 2,500x. Those three return almost the same percentage over the long run and feel nothing alike over an hour. So if you arrived searching for the highest RTP slot, the honest answer is that four PG Soft titles share the top of this table within a quarter of a point of each other, and choosing between them is a choice about shape rather than about value. Our Mahjong Ways breakdown works through the two mahjong builds in detail.

A headline RTP hides where the money is paid

PG Soft does something most of the industry avoids: it publishes the split. Mahjong Ways returns 96.92% in total, and the studio breaks that into 68.89% paid by the main game and 28.03% paid inside the Free Spins Feature. Close to a third of everything the slot ever gives back is sitting behind three scatter symbols. Fortune Tiger is starker still — 96.81% overall, of which 64.96% comes from ordinary spins and 31.85% from the Fortune Tiger respin, a game with no free-spins round at all. Lucky Neko leans the other way, at 75.34% main game against 21.39% free spins.

That split is the most useful thing on this page. An evening that never triggers a bonus is simply not the game the headline number describes — it is the base-game slice, and on Fortune Tiger that slice is 64.96%. It also explains why two titles with near-identical returns produce completely different sessions. Lucky Neko hands back three quarters of its return while you are just spinning. Mahjong Ways holds nearly a third of its return in reserve until the scatters land, which is why the same 96.9-ish figure can feel generous on one screen and mean on another.

The same slot can ship at more than one RTP

Here is the part almost nobody writes down. A slot is not one fixed piece of mathematics. Pragmatic Play publishes 96.50% for Gates of Olympus, while slot databases report lower builds of the same game in circulation — 95.51%, plus a third value quoted inconsistently as 94.05% and 94.5%. Those are operator-selectable configurations. An unqualified 96.50% on somebody else's page may not describe the version their casino runs, and there is no way to tell from the outside. Which is precisely why we put the figure on the tile rather than in an article.

We also run six Pragmatic Play titles as BK8-branded reskins, and their published returns sit on our own Pragmatic Play page where you can read them before you open anything:

Four of those print 97.6 — above the 96.50% Pragmatic Play publishes for its standard Gates of Olympus, and above every provider-published figure in the earlier table. Note also that the two Gates builds on our own shelf do not agree with each other: the standard reskin prints 97.6 and the 1000 version prints 96.5. That is the whole argument for reading the tile in front of you instead of a number you found elsewhere. Compare both against the studio build on our Gates of Olympus page before you decide which one to load.

Five games in the lobby publish no RTP at all

JILI is one of the busiest studios in Southeast Asia and it publishes no return figures whatsoever. Its five most-played titles and the site's own player code contain zero percentage values. This is not an oversight on one page — the company is owned by Taiwan-listed International Games System, with a market capitalisation around US$7.2 billion, and it still declines to print RTP anywhere. So read what it does publish, which is more than most players realise.

Two things stand out. All five carry the same LOW-MED label, and that label is real rather than a placeholder — JILI's catalogue genuinely spans LOW-MED, MED and HIGH-MED, and the sequel Fortune Gems 3 is rated HIGH-MED with a 10,125x ceiling. Second, the max-win numbers argue with each other: JILI's own page gives Super Ace 10,000x while slot databases give 1,500x, which happens to be the figure JILI itself publishes for a different game, Mega Ace. Take the studio's number, note the disagreement, and treat the JILI shelf as somewhere you judge by feel rather than by percentage.

Volatility and hit frequency are the rest of the sentence

RTP alone is half a sentence. The other half is how the return arrives, and there the published data gets messy fast. Pragmatic Play's own social-casino site labels Gates of Olympus Medium while slot databases rate it High. The same site labels Sweet Bonanza Low while databases put it between Medium-High and High. We are not going to arbitrate somebody else's rating, but you should know the disagreement exists before you let a volatility badge make a decision for you.

• PG Soft runs a published scale — Entertaining (Low), through Medium, to High — and all four of its titles here sit at Medium. That rating separates PG Soft from other studios far more than it separates one PG Soft game from another.
• Hit frequency is the figure nobody advertises. Pragmatic Play's own social-casino site gives Sweet Bonanza 50% and Gates of Olympus 28.82% — two games with almost the same return, one of which pays something on half its spins.
• Advertised max win is a ceiling, not a forecast. PG Soft ran a one-billion-spin simulation on Mahjong Ways and recorded a maximum exposure of 2,893x against its 25,000x cap; Mahjong Ways 2 produced 1,708x against 100,000x.

Put those three beside a return figure and you have something you can act on. A 96.48% game that pays on half its spins is a different evening from a 96.50% game that pays on 28.82% of them, and both are different again from a 96.81% game with no free-spins round in it. Percentages rank slots against each other. Structure decides what an hour with one feels like — and the two questions have almost nothing to do with one another.

Testing it yourself, in dollars, before you deposit

None of this has to be taken on trust. Run the title in demo mode first, at the stake you would genuinely use, and count how long the feature takes to arrive. Ten minutes of that teaches you more about a 28.03% free-spins slice than any table can, including this one. Our demo slot guide covers which games run in play-for-fun mode and what changes the moment the same game switches over to cash.

When you do deposit, the balance runs in US dollars. That is structural rather than a preference: our Cambodia region is restricted to a single currency, USD, and the string KHR appears nowhere in the platform configuration. With Cambodia 83.4% dollarised by broad money at the end of 2025, most readers here already hold a dollar account at ABA or ACLEDA. Fund from that one, because the Bakong rail performs no conversion in transit — the sending currency has to match the receiving account, and there is no in-app FX along the way.

A first deposit on a Cambodian account carries +1500% and 250 free spins. Promotional spins interact with RTP in one specific way worth knowing: they attach to selected slots rather than to the whole lobby, so the game's own published return still applies but the shortlist is drawn up for you. Check the eligible list on our free spins page against the tables above and spend them where the published number is highest. Open an account when the arithmetic makes sense to you, not before.

What RTP will not do for you

Three claims circulate in Cambodian gaming groups and none of them survive contact with the arithmetic. There is no hour of the day when a certified random number generator pays better, because it has no clock. There is no stake that moves the published return — 96.50% is 96.50% at the bottom of Gates of Olympus' 0.2 to 240 bet range and at the top of it, and the only thing a larger bet changes is how quickly you find out. And no run of previous spins predicts the next one, which is the entire point of the certification these games carry.

Two JILI games look like exceptions and are not. Money Coming gates its features by bet level — higher stakes unlock the x10 multiplier and the richer wheel outcomes — and Fortune Gems sells an Extra Bet that raises your stake by half to land high multiplier symbols more often. Both change what the game can do at a given stake, and neither is a published-return exception, because JILI prints no return at all.

The certification is what makes any of these figures worth reading in the first place. Our games are audited by BMM Test Labs, GLI and iTech Labs, and those marks sit in our footer rather than buried in a paragraph. PG Soft's titles are validated by BMM and Gaming Associates. Pragmatic Play is RNG-tested by Gaming Laboratories International, Quinel and Gaming Associates, and licensed by the Great Britain Gambling Commission under account number 56015 and by the Gibraltar Licensing Authority under RGL No. 107. A published RTP from an untested generator is a marketing line. From a tested one it is a measurement — and that is the difference between the numbers here and the ones on a page that will not tell you who checked them. Play a tested build.

Questions Cambodian players actually ask

Keep reading