One slot can ship at five different return percentages. The studio publishes the set; nobody publishes which one your casino received. So we store the range, name the source, and refuse to print a single number.
Published 2026-08-14Not changed since publicationNo review date set yetSources not re-read since publication
Habanero publishes, on its own game page for Fa Cai Shen, five return-to-player configurations: 86.00% to 98.00%◦. We read that page on 14 August 2026 and confirmed it in a second locale the same day. The studio does not say which of the five an operator receives as standard, and we have not marked one as the default, because marking one would invent the single most consequential fact on the page.
The spread between the lowest and highest build is twelve percentage points. A player on the lowest build and a player on the highest are playing games with the same name, the same artwork and the same paytable, and materially different economics. Any site that answers "what is the RTP of Fa Cai Shen?" with one number is answering for at most one of those two players and does not say which.
Our source-quality policy opens with a finding stated in bold: no provider publishes its configuration set, not the Western studios and not the Asia-first ones. It names the seven studios it checked: Pragmatic Play, Play’n GO, Nolimit City, JILI, Fa Chai, CQ9 and JDB.
Habanero is not among the seven, and Habanero publishes it. The finding was wrong as a universal, drawn from a sample that never contained the counter-example, and we have corrected it in place rather than quietly deleting it. What we have is one counter-example, not a new universal in the other direction: the other studios we have checked still do not publish their sets.
JILI’s page for Money Coming publishes a maximum win of 10,000x, a volatility label in words (LOW-MED), a payline count of one and a release month of December 2023, all server-rendered in a specification table, and no return figure of any kind.
PG Soft’s page for Fortune Tiger publishes a return of 96.81%◦, which we also found in a testing-laboratory report hosted on the studio’s own domain, and a "Session Volatility" of 2 on a scale the studio never states: not its bounds, not its direction. We store the return and we do not store the 2, because mapping an unscaled integer onto a word would be our estimate wearing the studio’s name.
So one studio publishes the volatility and not the return, and the other publishes the return and not a usable volatility. That is a difference in which fields are disclosed, not in how many, and no single letter grade for a studio can express it.
Spribe’s Aviator is quoted at 97% almost everywhere. We could not find that figure, or any figure, on any Spribe host. The game page renders 34 characters of text to a plain client; the promotional site’s how-to-play page renders about three words; the provably-fair page renders 35 characters. A search restricted to all three Spribe domains returned indexed pages and no percentage in any snippet from any of them.
So our record for Aviator carries no return figure and no maths row. It is the emptiest legitimate record in our database, and it is empty on purpose. Two of the five results that describe themselves as the official Aviator site are neither Spribe nor licensed operators; one of them titles itself with the 97% the actual provider withholds. A domain containing the word "official" is not an identity check.
Ezugi publishes three returns for Teen Patti: 96.63% on the ante, 95.51% on the pair-or-better side bet and 91.44% on the 3+3 bonus. It states all three as optimal, meaning conditional on the player making the right decisions, and Teen Patti’s ante carries a play-or-fold decision.
A figure that assumes correct play is not comparable with a slot’s return, which assumes nothing of the player because a slot offers no decisions. We therefore keep these three out of the table that sorts and print them in prose, attributed and labelled as skill-conditional. Prose does not sort. A number in a column invites a comparison, and this one would be false.
NetEnt’s own page for Starburst says 96.08%◦. Aggregators repeat 96.09%. We publish the studio’s figure and record where it came from, which is the entire reason we track the type of every source: a figure restated by an affiliate is evidence about the affiliate.
That 2012 title also publishes six figures on one page, including a hit frequency of 23.00%. The two most-quoted Pragmatic Play titles in our Asian corpus publish one number each, in a sentence. Age and reputation do not predict disclosure.
Every figure on this site carries the address it came from, what kind of source that was, and the date we read it. Returns are stored as a set, never as a single column on the game, so that a game with five configurations cannot be flattened into one by the shape of our database. Where a studio publishes nothing, we say the studio publishes nothing rather than borrowing a number from a site that also borrowed it.
We do not accept a figure whose only source is another affiliate. That rule costs us numbers on most pages, and it is what the rest of this page is an argument for.
The number that would actually answer the reader’s question is the one the casino in front of them has loaded: the per-operator, per-title reading. It is the join nobody in this industry builds, we have not built it either, and until we have, we will say so on the pages where it is missing rather than implying that the studio’s figure is the served figure.
Until then, the in-game information screen inside the casino client is the only place a player can read the configuration they are actually playing.
No. A return-to-player figure is a long-run theoretical average over a very large number of spins, and it says nothing about a session. It does not make a game due, warm, cold or ready, and no sequence of past results changes the next one.
Usually unpublished. The studio builds several and the operator receives one; neither side normally states which. The in-game information screen inside the casino client is the only place most players can read it, and it is the reading we treat as strongest.
Because for Fa Cai Shen five builds exist, they span twelve percentage points and the studio marks none of them as standard. A single number would be wrong for at least four readers in five, and we would not be able to say which four.
Not from Spribe, as far as we can read. Three Spribe hosts render almost no text to a client and no search snippet from any of them contains a percentage. We store no return figure for Aviator rather than restating one we cannot trace to the provider.
No. Ezugi states it as optimal, which means it assumes the player folds and plays correctly every hand. A slot return assumes nothing of the player. We print skill-conditional figures in prose and keep them out of any table that sorts.
Not a bibliography — these are the addresses behind the numbers this site stores about the 6 records this piece is about, with the date we read each one. Anything the article argues from that we have not stored as a figure is not in this list, and we would rather say that than pad it.
Nothing on this page has been corrected since it was published on 2026-08-14. When something is, it is listed here with the date and what changed — not quietly edited. How we handle a correction.
A wagering requirement times a house edge gives a real cost in pesos or dollars. The build that sets that edge is stored on this site for dozens of slots, and most bonus terms never name it.
PG Soft confirms 96.95 percent on Mahjong Ways 2 and 96.92 percent on the original. Neither page states a ways count, a payline total, or a volatility label.
PAGCOR-licensed casinos rarely lock only the bonus a player claims. The wagering requirement usually covers the whole balance, deposit included, until the combined total clears the multiple.
Two of PG Soft's own game-information sheets name a selectable range of RTP builds. Four others state flatly that no such range exists for that title. TopSlotsAsia checked all thirty-two titles it has published from this studio.