How We Verify In-Game RTP

Why a single RTP number isn't enough

RTP (Return To Player) is the percentage of all wagered money a slot pays back to players over time. A 96% RTP means that, over a long enough horizon, $96 of every $100 wagered comes back as wins — and $4 is the house edge. It's a statistical long-run figure, not a guarantee for any one session, but it's the most important math fact about any slot.

The catch is that "the RTP of slot X" is not always a single number. Three independent factors can change what the actual RTP is on a given spin:

1. Provider tier selection. Several major providers ship the same slot in multiple configurations — typically 96.x%, 94.x%, and 88.x%. The operator picks one at integration time. Pragmatic Play is the best-known example, but Play'n GO, Habanero, and parts of the Yggdrasil catalogue do this too.
2. Bonus-round dynamics. Some slots have separate RTP figures for base game vs. bonus rounds, or for jackpot-pool versions of the same game. We document each variant separately when applicable.
3. Operator restrictions during bonus play. Even if the slot itself is configured at the headline RTP, a bonus's terms may exclude it or weight it at a lower contribution — which is a different problem, but one that compounds with low-tier configuration.

Multi-tier RTP: the Wolf Gold case study

Wolf Gold by Pragmatic Play is one of the most widely distributed slots in the industry. Its official RTP is published as 96.01%. That number appears on Pragmatic's website, in their B2B sales decks, and in nearly every third-party slot database.

What's almost never mentioned: Pragmatic Play also offers Wolf Gold to operators at 94.06% and 88.13%. Those alternative configurations exist in the same game certification and are picked at the operator's discretion. The difference between 96.01% and 88.13% is nearly 8 percentage points of house edge — meaning a player at the low-tier configuration loses, on average, roughly four times faster than at the headline tier.

For every multi-tier slot in our database we record:

• The highest tier the provider documents (rtp_theoretical_max).
• The lowest tier the provider documents (rtp_theoretical_min).
• The full list of variants, when available (rtp_variants_json).
• An archived snapshot of the provider page on the date we recorded the values.

As of 19 June 2026, 10 slots in our database carry multiple documented RTP tiers.

What counts as a primary source

We rank sources in three tiers of evidential weight, in descending order of trust:

1. Regulator certificates from accredited testing labs (eCOGRA, GLI, iTechLabs, BMM). These are the highest-trust source where available — they document the exact RTP a game was tested at and the date of certification. Most providers don't publish them; when they do, we cite them.
2. Provider game pages — the official URL where the developer of the slot publishes the game's data sheet. For NetEnt slots this is now games.evolution.com/slots/<slug>/ (NetEnt was acquired by Evolution in 2020). For Play'n GO, playngo.com/games/<slug>. For Pragmatic Play, pragmaticplay.com/en/games/<slug>/.
3. In-game paytable — the info panel inside the slot itself, as rendered by a specific operator. This is the only source we have for providers that don't publish public game pages (RTG, SpinLogic, Rival Gaming, IGT). We document it via timestamped screenshots tied to a specific casino and a specific date.

Aggregator sites (SlotCatalog, AskGamblers, slotsmate.com) are not considered primary sources. They're useful for cross-referencing, but we don't cite them as the basis for any value in our database. Their numbers are copied from somewhere; we want to be that "somewhere."

Why we archive every primary source on Wayback

Provider web pages change. A slot's RTP value can be updated, removed entirely, or relocated to a different URL when a developer is acquired (as happened to NetEnt). If we cite a URL today and that URL serves different content tomorrow, our claim becomes unverifiable.

For every slot we document, we trigger an archive on the Internet Archive's Wayback Machine at the moment of verification. The archived URL is stored alongside the live URL in our database. As of 19 June 2026, we have archived snapshots for 19 of our 69 documented slots.

A reader who wants to verify any single claim on a FreeExtraChips slot page can follow the archived URL and see the exact provider content we based the claim on — even if the live URL has since changed.

When providers don't publish RTP: the opaque tier

Several major providers do not maintain public game pages with RTP values. RTG (Real Time Gaming) and its successor SpinLogic Gaming, Rival Gaming, and the legacy IGT online catalogue are the most common examples in the US-facing market our database covers. Their slots collectively account for a large share of bonus offers — Cash Bandits 3 alone appears in 14 active bonuses on FreeExtraChips at the time of writing.

For these providers we cannot cite a primary source URL. Instead we use a two-step process:

1. Open the slot at an operator that carries it (e.g., SlotoCash for SpinLogic titles, Vegas Crest for Rival). Access the in-game paytable / info panel. Take a screenshot timestamped with the date and operator name. Record the RTP value the paytable shows.
2. Repeat at one or two additional operators. If the same slot shows different RTP values across operators, document each — this confirms the provider ships multiple configurations even though they aren't publicly disclosed.

We mark these slots with a transparency flag in our reviews. The absence of a public provider source is itself useful information for players evaluating where to play.

Per-casino observations: closing the loop

The theoretical RTP a provider declares is one fact. The RTP a specific casino actually serves is a second fact, and the two don't always agree. To document the second, we maintain a separate per-casino observation log: each entry pairs a slot with a casino and records the RTP observed in the in-game paytable on a specific date, with a screenshot.

As of 19 June 2026 we have 2 observations covering 2 slots. This is a coverage we're building up over time; we don't claim to have every slot–casino pair documented. When a bonus is offered specifically on a slot we've observed and the observed RTP is below the headline tier, we surface that fact in the bonus listing.

Update and correction process

RTP data is not static. Providers release new versions of games, change configurations, and occasionally update their certifications. Operators can switch RTP tiers without notice. We treat the database as living data:

• Scheduled refresh. A weekly job re-fetches the provider primary URLs for the slots with public game pages, parses the listed RTP, and flags any record where the current value differs from what we have stored. Flagged records are reviewed by an editor — never auto-updated. We've seen scrapers replace correct historical data with wrong new data when a CDN serves a stale page; a human always confirms.
• Reader corrections. Any slot data page on FreeExtraChips carries a "Spot a mistake?" link that opens a contact form pre-filled with the slot's identifier. Submissions are routed to [email protected] and acknowledged within five business days.
• Version log. Every change to a slot's record is timestamped with the editor who made the change and a brief reason (e.g., "Provider updated 2024-12 from 96.21% to 96.10%"). Older values are kept in the version log, not overwritten. Readers can see how the RTP we report has evolved over time.
• Methodology versioning. This page itself is versioned. The methodology described here is version v1, last updated 19 June 2026. Material changes to how we verify RTP will be released as a new version with a changelog entry, not a silent edit.

What this methodology does not promise

We are explicit about the limits of what RTP verification can tell a player:

• RTP is a long-run statistical figure. It does not predict the outcome of any single spin or session. A slot at 96% RTP can pay out 50% over an hour or 200% over an hour. The figure converges over millions of spins, not over an evening's play.
• Verifying RTP does not endorse a slot. A high RTP doesn't make a slot a good bet — volatility, hit frequency, and bonus terms all matter. We document RTP because it's verifiable, not because it's the only thing that matters.
• We cannot audit RNG fairness. RTP figures presuppose that the random number generator in the slot is correctly implemented. That's the role of accredited testing labs, not ours. Our job is to make sure players know what RTP the operator is configured to serve.

Frequently asked questions