Every slot ships with multiple RTP variants already built and certified. The casino picks one - usually not the highest. We break down exactly how that configuration system works, how each variant gets certified, and what actually changes between them.
By Marcus Reid·Published May 27, 2026·10 min read·Intermediate guide
Look, the term "RTP configuration" sounds more technical than it is. Here's the plain version: every slot game is built with a probability engine under the hood. That engine can be set to run at different return levels. The developer pre-builds several settings - each one is a separate, independently certified version of the same game. Casinos pick which setting to activate.
That's it. No code change on the casino's side. No rebuilding of the game. It's a back-end setting - the same kind of toggle a platform uses to enable or disable a payment method. The game looks identical. The symbols, the layout, the bonus feature all stay the same. The probability tables underneath shift.
Worth noting: this is not a grey area or a workaround. It's an intentional feature of the licensing model. Providers offer it because larger operators often negotiate access to higher-RTP variants as part of their deals. Smaller casinos may only get access to lower variants entirely.
In plain terms: Think of it like a petrol engine with different fuel maps. Same engine, same body, same seats. But the fuel map changes how efficiently it runs. RTP configuration is the slot equivalent - same game, different probability weighting.
How the RNG Connects to RTP Configuration
The RNG - Random Number Generator - produces an endless stream of numbers. Those numbers map to outcomes: symbol positions, win amounts, bonus triggers. The mapping is defined by a lookup table. That table is what changes between RTP variants.
Honestly, the RNG itself doesn't care about RTP. It just generates numbers. The configuration layer sits between the RNG output and the outcome display. A higher-RTP configuration maps more RNG values to winning outcomes - or to bigger ones. A lower-RTP config maps fewer of them there.
RNG Output
An integer - something like 4,827,391 - generated in milliseconds every spin. Truly random. The same at all casinos. The RNG has no concept of RTP.
Config Layer
A probability table maps that integer to a reel outcome. This is what the casino configures. Variant A maps more integers to high-paying positions. Variant C maps fewer. Same RNG, different table.
Outcome
The reels stop, the symbols land, the win (if any) is calculated. Players see this part. They don't see the table that determined it.
Paytable Display
The in-game paytable reads the active configuration and displays the corresponding RTP percentage. This is the only place the real configured RTP appears. Review sites typically show the maximum variant.
The RNG is typically certified once per game title by the testing lab. The configuration tables are certified separately - one certification per variant. So a slot with 3 RTP variants has 4 certifications: the RNG itself, plus one for each variant.
How RTP Variants Actually Differ
Use the tabs below to explore what each variant tier typically looks like. The specific percentages vary by provider and game, but the pattern holds fairly consistently across the industry.
92-94%
Typical low variant RTP range
Who runs itBudget operators, some offshore casinos
Casino margin6-8% of every spin
PaytableSometimes hidden at this tier
From what we saw at Pistolo, the Hacksaw Gaming slots ran at this level - Le Zeus at 92.37%, Le Fisherman at 92.34%. That's close to 4pp below the published max. In practice, you lose roughly twice as fast compared to the high variant of the same slot.
94-95.5%
Typical mid variant RTP range
Who runs itMost mid-tier operators
Casino margin4.5-6% of every spin
PaytableUsually visible
This is roughly where most of the slots at Wild Tokyo landed on our tests - Gates of Olympus at 94.5%, Big Bass Bonanza at 94.04%, Merge Up at 94.25%. The published max on those was 96.5%, 96.71%, and 97.25% respectively. Mid tier, in practice.
96-97.5%
Typical high variant RTP range
Who runs itPremium or regulated operators
Casino margin2.5-4% of every spin
PaytableAlways visible - they want you to see it
Wild Tokyo ran 6 of the 15 slots we tested at exactly the published max RTP - Coin Win Hold The Spin at 96.06%, Vikings Go Berzerk at 96.1%, Tidal Gold at 96.17%, Luck of Panda at 96.33%, 3 Witch Pots at 96.07%, Book of Dead at 94.25%. When a casino runs max variants, they typically make that visible.
Worth knowing: Not all providers offer 3 tiers. Some offer just 2 (low and high). A few offer 4. The spread between lowest and highest varies too - on some Pragmatic Play titles we tested, the spread was about 2pp. On Hacksaw Gaming slots at Pistolo, the gap reached nearly 4pp.
Which providers offer what tends to depend on the game's target market. High-volatility slots aimed at VIP players often have wider variant spreads - because casinos running those games want the flexibility to recoup losses from big wins by lowering the base RTP elsewhere.
Typical Variant Ranges by Provider
Pragmatic Play
~94%~95.5%~96.5%
Hacksaw Gaming
~92%~94%~96.3%
Play'n GO
~94.25%~96.51%
OnlyPlay
~94%~97%
NetGame
~96%~96.3%
Yggdrasil
~94%~96%~96.1%
Ranges based on published documentation and tested readings. Exact variants vary by game title.
How Each RTP Variant Gets Certified
Before a casino can run any RTP variant, that variant has to be certified by an independent testing laboratory. iTech Labs, BMM Testlabs, eCOGRA, and GLI are among the most common. The process is roughly the same across all of them.
That said, the certification doesn't guarantee fairness in the sense most players expect. It guarantees the game behaves as documented - not that the documented RTP is high.
1
Provider submits the game build
The full game code is submitted to the lab, along with documentation of every RTP variant - the probability tables, symbol weights, and expected return for each configuration.
2
Lab runs statistical simulations
The lab runs the game through billions of simulated spins for each variant - typically 10 million rounds minimum, often far more. They verify the actual return matches the documented RTP within an acceptable margin.
3
RNG integrity is verified separately
The RNG is tested independently - for statistical randomness, seed unpredictability, and tamper resistance. A certified RNG means outcomes can't be predicted or manipulated. It does not mean the RTP is favourable.
4
Certificate issued per variant
Each variant gets its own certificate. The certificate states the tested RTP and the conditions under which it holds - bet size ranges, active features, and so on. Casinos must use only certified variants in licensed jurisdictions.
5
Casino activates a variant
The casino selects one of the certified variants through their platform admin panel. The selection must match the RTP displayed in the in-game paytable. That link is what regulators audit - not whether the RTP is high.
The important bit: Regulators typically audit whether the displayed RTP matches the active configuration - not whether the RTP is good value for players. A casino running a certified 92% variant with "92%" in the paytable is fully compliant. That's legal. That's the system.
What Changes Between Variants - and What Stays the Same
This is where a lot of players get confused. Switching from a 94% to a 96.5% variant of the same slot does not make it a different game. The experience is nearly identical. What changes is subtle - and mostly invisible.
Stays the Same
Game title, theme, and all visual assets
Symbol names, positions, and artwork
Bonus feature rules and mechanics
Number of reels and paylines
Min and max bet limits
Maximum win cap
Wild and scatter symbol behaviour
Changes Between Variants
Symbol frequency in the probability table
Win frequency (how often any win lands)
Average win size relative to stake
Bonus trigger frequency (in some games)
The RTP figure in the paytable
Expected session loss rate
In practice, the difference between variants isn't obvious in a single session. Variance drowns it out. Strangely, you could play 200 spins on a 94% variant and walk away up - and play 200 on the 96.5% variant and lose more. Short-term, variance wins. Long-term, the configuration gap compounds.
The maths, simply: On a $1 stake, a 94% slot costs you $0.06 per spin on average. The 96.5% variant costs you $0.035 per spin. Across 1,000 spins that's $60 vs $35. Same game, same session length - $25 difference. Just from the configuration choice.
RTP Configuration in Practice - What We Recorded
We tested 30 slots across 2 casinos - Pistolo and Wild Tokyo - recording the RTP from the in-game paytable and comparing it to the published maximum. The bars below show the gap we found on each tested slot. A zero gap means the casino is running the highest available variant.
Le Fisherman (Pistolo)96.33% pub / 92.34% actual
Le Zeus (Pistolo)96.26% pub / 92.37% actual
Merge Up (Wild Tokyo)97.25% pub / 94.25% actual
Big Bass Bonanza (Wild Tokyo)96.71% pub / 94.04% actual
Gates of Olympus (Wild Tokyo)96.5% pub / 94.5% actual
The Dog House (Pistolo)96.51% pub / 95.51% actual
Coin Win Hold The Spin (Wild Tokyo)96.06% pub / 96.06% actual
Vikings Go Berzerk (Wild Tokyo)96.1% pub / 96.1% actual
Tidal Gold (Wild Tokyo)96.17% pub / 96.17% actual
The pattern we noticed: Wild Tokyo runs a mixed configuration - some slots at max, others at a mid variant. Pistolo ran everything we could read at a low or mid variant, and hid 4 RTPs entirely. Neither casino is consistent across all providers.
So in practice, knowing the configuration system matters. The same slot genuinely costs different amounts to play depending on where you play it. Check our full RTP database to see which variant is active at each casino we've tested.
Marcus Reid
iGaming Data Analyst
Marcus built the RTP testing methodology used across this site. He has read the in-game paytable on every slot in the database - at each listed casino - using a funded account. The bar chart above shows real readings, not estimates. His background is in probability modelling and statistical testing, not casino marketing.
RTP Database
