Stake Engine: what the platform is and how its slots really pay — a breakdown of 7 649 games
⚠️ 18+ | Affiliate Disclosure: This article contains affiliate links—we earn a commission on clicks, but there’s no cost to you. Gamble responsibly: BeGambleAware.org - Link
Last updated: 12 November 2026
Stake Engine (since September 2026, simply Engine) is the Stake casino platform where third-party studios release their own slots. Its catalogue currently holds 7 649 games from 2 082 publishers, and since autumn 2026 they have been appearing not only on Stake but at other crypto casinos too.
What sets Engine apart for a player: the maths of every game is public. We downloaded that data for the whole catalogue and worked out what stands behind the familiar “RTP 96.7%”. In short: the RTP is honest, but in a typical game 90% of spins end with nothing or with a payout smaller than the bet.
At a glance: key facts about Engine slots
| Parameter | Meaning |
|---|---|
| What it is | Stake’s platform for third-party studios, launched in April 2025 |
| Games in the catalogue | 7 649 (snapshot of 11 November 2026) |
| Publishers | 2 082 |
| Typical RTP | 96.00–96.70% — for 84% of games |
| Highest RTP in the catalogue | 98.95% (a plinko game) |
| Slots with 100% RTP | None |
| Typical max win | ×15 000 (catalogue median) |
| Dead spins in a typical game | 71% |
| Where to Play | Stake, Housebets, Rainbet and other casinos that have connected Engine |
What Stake Engine is
Stake launched Stake Engine in April 2025. It is a server and a toolkit with which an independent studio builds a game and releases it straight into the casino lobby. In return, the developer gets a share of the revenue from its game — at launch Stake announced 10% of gross gaming revenue.
On 15 September 2026 the platform was renamed Engine and opened to other operators. A casino connects once and gets the games of all studios at the same time. According to Stake at the time of the announcement, the platform’s games had generated more than $23 billion in turnover and 9.7 billion bets.
There are many studios, but they differ greatly in size. 45% of the publishers in the catalogue have exactly one game. The top of the Engine section in the Stake lobby is taken mostly by two studios — Paperclip Gaming and Valkyrie.
How an Engine slot works
The outcome is picked from a ready-made table
A conventional slot calculates the result at the moment of the spin. On Engine, the studio runs a simulation in advance and stores the finished outcomes — a typical game has 300 000 of them in the base mode. Each outcome has a weight (probability) and a payout.
When you press “spin”, the server picks one row from that table. That is why RTP here is not an estimate “based on test results” but an exact number: the sum of all payouts weighted by probability.
The slot has no memory
The platform’s rules require every bet to be independent. Jackpots, gamble features, round continuation and early cashout are prohibited.
For a player this means one simple thing: the slot does not “save up” anything and does not “owe” anyone. A run of two hundred dead spins does not raise the chances of the two hundred and first.
The tables are public
Engine runs a public fairness verification service. Using the server seed, the client seed and the spin number, anyone can reproduce a result and make sure it was taken from the published table.
The same service returns statistics for every game mode. The whole breakdown below is built on them. For 163 games out of 7 649 the service did not return full statistics — they are not included in the spin calculations.
What RTP Engine slots have
We took the base-mode RTP of every game — in the form the player sees it, with two decimal places.
| Base-mode RTP | Games | Share of catalogue |
|---|---|---|
| below 94.00% | 195 | 3% |
| from 94.00 to 94.99% | 133 | 2% |
| from 95.00 to 95.99% | 534 | 7% |
| from 96.00 to 96.49% | 4 068 | 53% |
| from 96.50 to 96.69% | 1 511 | 20% |
| exactly 96.70% | 869 | 11% |
| from 96.71 to 96.99% | 46 | 1% |
| from 97.00 to 97.99% | 234 | 3% |
| 98.00% and above | 59 | 1% |
The three most common values are 96.00% (33% of games), 96.50% (17%) and 96.70% (11%). Only 4% of the catalogue is above 96.70%.
The highest RTP in the catalogue is 98.95%, in a plinko game. Among slots, 98% does occur: for example, Waylanders Forge by Valkyrie. The lowest is 90.00%.
The RTP of one and the same game can differ between modes. A regular bet, a boosted bet and a bonus buy are separate tables, each with its own figure. Check the RTP of the exact mode you are playing.
How the metric itself works — how it is calculated and why it “converges” only over the long run — is explained in a separate piece: what RTP in slots is - Slot rtp.
Are there Engine slots with 100% RTP?
No. We checked the modes of every game in the catalogue — there are 117 521 of them — and none reaches even 99.9%.
An RTP of 100% means a zero house edge: over a long distance, players collectively get back everything they staked. Some crypto casinos announce such terms for their own in-house games — Dice, Mines, Plinko and the like — but not for slots from third-party studios. On the Duel site, for example, the in-house games are labelled “100% RTP” (checked on 12 November 2026). Gamdom and BiggerZ have announced similar terms. The list of games and the limits are different at every casino and change over time.
And even a 100% RTP does not make a game a winning one. It is an average over millions of rounds, while a single session can still end either in profit or with a zero balance.
Open Duel → - C 6be3de85b8869935
18+ | Play responsibly
Why an RTP of 96.7% does not save a short session
RTP shows how much is returned on average over a huge distance. It says nothing about how that money is distributed. In many Engine slots the distribution is very uneven: a large part of the return comes from rare big payouts.
What the whole catalogue showed
We took the base mode of 7 486 games and looked at three shares: spins with no payout, spins with any payout, and spins that return at least the bet.
| Metric (base mode) | A quarter of games have less | Median | A quarter of games have more |
|---|---|---|---|
| Dead spins (payout 0) | 62% | 71% | 77% |
| Spins with any payout | 23% | 29% | 38% |
| Spins paying at least the bet | 6% | 10% | 16% |
Put simply, per 100 spins of a typical game: about 71 are dead, about 19 pay less than the bet, and about 10 pay at least the bet.
In 24% of games, 94 spins out of 100 or more end “in the red”.
Max wins, meanwhile, are huge. The catalogue median is ×15 000, 39% of games have a cap of ×25 000 or higher, and the record is ×1 000 000. But the chance of hitting the maximum in a typical game is roughly 1 in 3.3 million spins.
12 games from the top rows of the lobby: what is left after 100 spins
We took 12 games from the upper part of the Engine section in the Stake lobby and reproduced 200 000 base-mode spins for each. Then we cut them into 2 000 sessions of 100 spins.
The conditions are simple: a budget of 100 bets, 100 spins in a row at one bet each, with no bonus buys and no boosted bets. The “remainder” is the percentage of the budget left at the end of the session.
| Slot (studio) | Max. Payout | Dead spins | Remainder after 100 spins (median) | Sessions losing 70% or more | Sessions in profit |
|---|---|---|---|---|---|
| Dragonspire Frostfall (Paperclip Gaming) | ×20 000 | 81% | 43% | 27% | 20% |
| Lazy Vampire (Paperclip Gaming) | ×50 000 | 81% | 41% | 34% | 24% |
| Midas Squares (Paperclip Gaming) | ×25 000 | 81% | 40% | 39% | 24% |
| Deal with the Devil (Paperclip Gaming) | ×66 666 | 81% | 37% | 40% | 22% |
| Private Jet (Paperclip Gaming) | ×77 777 | 73% | 37% | 39% | 21% |
| Minedrop 2 (Paperclip Gaming) | ×50 000 | 81% | 32% | 47% | 16% |
| Undead Samurai (Glass Cannon) | ×50 000 | 80% | 29% | 51% | 14% |
| Book of Death (Valkyrie) | ×10 000 | 92% | 22% | 59% | 22% |
| Lab Breach (Sunnie Gaming) | ×20 000 | 80% | 21% | 69% | 18% |
| Night of Nyx (Valkyrie) | ×24 000 | 85% | 20% | 59% | 25% |
| Waylanders Forge (Valkyrie) | ×80 085 | 94% | 15% | 62% | 20% |
| Odin’s Vault (Valkyrie) | ×500 000 | 95% | 10% | 59% | 19% |
The RTP of 10 games on the list is 96.7%, Dragonspire Frostfall has 96.0%, and Waylanders Forge has 98.0%.
In the softest game on the list (Dragonspire Frostfall) a typical session keeps 43% of the budget, in the harshest (Odin’s Vault) — 10%. Between 14 and 25% of sessions end in profit, while losing 70% of the budget or more is the outcome of 27–69% of sessions.
Note Waylanders Forge. It has the highest RTP on the list — 98% — yet its median remainder is one of the lowest: 15%. A high RTP does not make a game soft — it only means the rare big payouts are even bigger.
Something else stands out: five of the six Paperclip Gaming games have the same share of dead spins — 81%. Games from one studio are built on a shared mathematical base; what changes is the theme and the artwork.
The harshest scenario is losing 90% of the budget or more. In the four Valkyrie games on the list, 26–50% of 100-spin sessions end that way. Over 50 spins with a budget of 50 bets, it is 41–67%. For the other eight games the figures are 1–5% and 5–14% respectively.
One game — five volatility levels
For some Engine games, several variants of the same mode with different volatility are published at once. There are 97 such games out of 7 649, and they include Paperclip Gaming hits. The RTP of the variants is identical, but the behaviour is not.
Here are the five variants of a regular spin in Lazy Vampire — the same 200 000 spins and 2 000 sessions for each:
| Variant | Dead spins | Spins paying at least the bet | Remainder after 100 spins (median) | Sessions losing 70% or more | Chance of a maximum win |
|---|---|---|---|---|---|
| VLOW — the softest | 50% | 19% | 72% | 0% | 1 : 19 500 000 |
| LOW | 68% | 13% | 56% | 7% | 1 : 9 000 000 |
| Regular | 81% | 8% | 41% | 34% | 1 : 4 000 000 |
| HIGH | 91% | 4% | 24% | 56% | 1 : 4 000 000 |
| EXTREME — the harshest | 97% | 2% | 13% | 61% | 1 : 4 000 000 |
With the same RTP, a typical session keeps anywhere from 13 to 72% of the budget. The price of softness is the chance of the max win: in the softest variant it is one in 19.5 million spins, in the regular one — one in 4 million.
If a game offers a choice of volatility, it affects a short session more than the choice of the slot itself.
Bonus buy: what comes back in a typical case
A bonus buy on Engine is a separate mode with its own table. In the six games we checked, its RTP matches the RTP of the regular game. We reproduced 20 000 purchases for each.
| Slot (studio) | Purchase price, bets | Median return, bets | Median return, % of price | Purchases returning less than 30% of the price | Purchases that paid for themselves |
|---|---|---|---|---|---|
| Odin’s Vault (Valkyrie) | 200 | 120 | 60% | 12% | 19% |
| Minedrop 2 (Paperclip Gaming) | 100 | 47 | 47% | 33% | 23% |
| Book of Death (Valkyrie) | 100 | 41 | 41% | 40% | 27% |
| Dragonspire Frostfall (Paperclip Gaming) | 100 | 38 | 38% | 42% | 23% |
| Private Jet (Paperclip Gaming) | 100 | 36 | 36% | 44% | 23% |
| Waylanders Forge (Valkyrie) | 100 | 17 | 17% | 59% | 24% |
A typical purchase returns noticeably less than its price: the median return is from 17 to 60% of the price. 19–27% of purchases pay for themselves, that is, roughly every fourth or fifth.
The RTP is the same as in the regular game. What changes is the speed: ten purchases at 100 bets each equal the turnover of a thousand regular spins in a couple of minutes.
Why max wins happen so often on streams
The explanation lies in the game modes. The max win is calculated from the base bet. In the expensive modes a purchase costs hundreds or thousands of bets, while the cap stays the same. So the maximum lands far more often there — and looks far more modest relative to the price.
| Slot | Mode | Price, bets | Chance of a maximum win | Maximum relative to price | Bets of turnover per one maximum |
|---|---|---|---|---|---|
| Waylanders Forge | regular spin | 1 | 1 : 2 002 135 | ×80 085 | 2 002 135 |
| Waylanders Forge | most expensive purchase | 2 000 | 1 : 276 | ×40 | 552 000 |
| Lazy Vampire | regular spin | 1 | 1 : 4 000 000 | ×50 000 | 4 000 000 |
| Lazy Vampire | most expensive purchase | 1 000 | 1 : 2 999 | ×50 | 2 999 000 |
| Odin’s Vault | regular spin | 1 | 1 : 26 366 441 | ×500 000 | 26 366 441 |
| Odin’s Vault | most expensive purchase | 5 000 | 1 : 4 869 | ×100 | 24 345 000 |
| Private Jet | regular spin | 1 | 1 : 6 222 160 | ×77 777 | 6 222 160 |
| Private Jet | most expensive purchase | 1 000 | 1 : 4 657 | ×78 | 4 657 000 |
Take Waylanders Forge. On a regular spin the ×80 085 maximum lands on average once in 2 million spins. In the most expensive mode, which costs 2 000 bets, it lands once in 276 purchases. But relative to the purchase price that is no longer ×80 085 — it is ×40.
Per bet wagered, the chance of the maximum is of the same order — you can see it in the last column. The expensive modes simply turn over thousands of times more money in the same time.
We do not know on what terms and in which modes particular streamers play. But no special terms are needed to explain frequent maximums on air — the published tables are enough. A clip with a “max win” says nothing about what will happen at a regular bet.
What this means for a player
- Look beyond RTP. Two games with the same 96.7% can leave 10% or 41% of the budget after 100 spins. The share of dead spins and the chance of getting the bet back say more about a short session.
- Count your budget in spins. Over 100 spins, the typical result in these games is minus 57–90% of the session budget. Factor that in beforehand.
- Do not expect the slot to “pay back”. Every spin is independent; the game has no memory. Raising the bet after a dead run only means spending the budget faster.
- A bonus buy is no better value than regular play. The RTP is the same, but the money goes a hundred times faster.
- Do not judge a game by streams. In modes costing hundreds or thousands of bets the maximum lands thousands of times more often per round. At a regular bet the same event is one in millions of spins.
- Check the volatility level. If the game gives a choice, the soft variant stretches the same budget several times longer.
- Check the game before you bet. The RTP, the chance of the max win and the share of dead spins can be looked up in the public data before the first spin.
- Set your limits in advance. Decide before you start how much money and time you are ready to spend, and do not revise it as you play. This is entertainment, not a way to make money.
Where to play Engine slots
Stake — the platform’s home venue. It has the fullest catalogue, including games released as Stake exclusives.
Open Stake → - C 3c03f09b593aa0d9
18+ | Play responsibly
Housebets — the Engine.io section appeared here at the end of October 2026. As of 11 November 2026 it holds 96 games from at least 54 studios; 45% of them have an RTP of 96.7%. There are no Paperclip Gaming hits in the section, but there are Valkyrie, Uppercut Gaming, Xposed Studios and Butter. Do not confuse it with House Studios — that is the casino’s own programme with different games.
Rainbet — the casino’s site has an Engine section with games from Paperclip Gaming, Valkyrie and other studios. Other crypto casinos have also reported connecting the platform.
The set of games differs from casino to casino, while the maths of a particular game is, as a rule, the same: the outcome table is tied to the game, not to the venue. What depends on the casino is the bet limits and which studios made it into the lobby.
Open Housebets → - C 9b3df7dc19622c2d
18+ | Play responsibly
More about Housebets — in our casino review - Housebets Review. None of these casinos accepts players from every country: check the list of restrictions in the terms before you register.
If you would like to get the terms straight first, we have guides to slot volatility - Slots volatility and to how to choose slot machines - Slots.
Frequently Asked Questions
What is Stake Engine in simple terms?
It is the Stake casino’s platform on which any studio can release its own slot. Stake takes care of the server, hosting and payouts; the studio does the maths and the design. Since September 2026 the platform has been called Engine and works with more than just Stake.
What RTP do Stake Engine slots have?
84% of the games in the catalogue have a base-mode RTP between 96.00 and 96.70%. The most common values are 96.00%, 96.50% and 96.70%. The range across the whole catalogue is from 90.00 to 98.95%.
Are there Stake Engine slots with 100% RTP?
No. We checked every mode of every game (117 521 of them): not one has an RTP of 99.9% or higher. A zero house edge is found only in the in-house games of certain crypto casinos, not in Engine slots.
Can you verify that a spin was fair?
Yes. On Stake, the server seed is revealed after you rotate the seed pair. With it, the client seed and the spin number, the result is reproduced through Engine’s public verification service and checked against the published outcome table.
Why does a slot not pay out after a long dead run?
Because it has no memory. Engine’s rules prohibit mechanics that depend on past bets. The chance of every spin is the same, no matter how many dead ones came before it.
Why do streamers hit the max win so often?
Most often it comes down to the game mode. The win cap is calculated from the base bet, so in a mode costing hundreds or thousands of bets it lands once in hundreds or thousands of purchases, not once in millions of spins. Relative to the purchase price, such a “maximum” is ×40–×100.
Are Engine games at Housebets and on Stake the same thing?
The maths of a game is the same. What differs is the selection: some studios keep their games as Stake exclusives, so the catalogue at other casinos is smaller. Bet limits are also set by the casino itself.
How we calculated
- Source. Engine’s public fairness verification service: the game catalogue and mode statistics. No data on real player bets was used.
- Snapshot date. 11 November 2026. Games in the catalogue: 7 649; modes: 117 521; spin statistics were obtained for 7 486 games.
- Sessions. 12 games × 200 000 reproduced base-mode spins = 2 000 sessions of 100 spins each. The margin of error of the median is up to ±3 pp.
- Bonus buy. 6 games × 20 000 purchases.
- Volatility levels. 5 variants of the regular Lazy Vampire spin × 200 000 spins.
- Chance of the max win by mode. Taken from the service’s statistics for each mode.
- Rounding. Shares in the tables are rounded to whole percentages.
- Limitations. A flat bet and one mode per session. Boosted bets and mixed play will give different figures.
Sources
- Engine’s public fairness verification service (game catalogue and outcome tables), snapshot of 11 November 2026.
- Engine developer documentation: game requirements.
- Stake’s announcement of the renaming of Stake Engine to Engine, 15 September 2026.
- The Engine.io section in the Housebets lobby, checked on 11 November 2026.
🔞 Play responsibly. Gambling is entertainment, not a way to make money: in the long run the edge stays with the casino. Do not bet money you cannot afford to lose. Help is free and confidential: BeGambleAware.org - Link | GamCare - Link
