How We Verify Every Number

Every RTP, odds and max-win figure on PlayStories is checked against at least two independent sources before it is published, and every probability is computed exactly from the game’s structure, not estimated. This page explains the process, the data, and what happens when we get something wrong.

The two-source rule

A number appears on this site only when it is confirmed by at least two independent sources. Provider documentation comes first (Spribe, SmartSoft, Pragmatic Play, BGaming, Stake’s published game math), followed by certification and payout databases and the operator’s own in-game info panel. When sources disagree because operators run different certified builds of the same game, we publish the full range and say which version applies where. When we cannot verify a number, we say so instead of printing one.

Exact math, not estimates

Probabilities are derived from how each game actually works. Plinko outcomes follow the binomial distribution of the board, so we compute every slot’s exact probability and cross-check with simulation; our Plinko study runs a full 2,000,000-drop simulation against the closed-form math. Crash games follow a published curve, so the chance of reaching a multiplier m is RTP / m, which anyone can reproduce. Where a provider publishes the formula, we replay it; where they publish only the RTP, we state that and mark the difference.

Open datasets (CC BY 4.0)

Our verification work is released as open data. You are free to reuse it with attribution to PlayStories.

• RTP builds dataset (CSV): 36 certified RTP builds across 12 popular slots, documenting how the same game ships at different returns (Book of Dead spans 96.21% down to 84.18%). Companion article: the RTP version problem.
• Max-win ceilings dataset (CSV): published win ceilings for 23 games and whether the odds of hitting them are disclosed. Companion article: max-win ceilings vs reality.

Corrections

Providers change their games. When Stake raised Plinko’s ceiling from 1,000x to 10,000x, we corrected every affected page and the figures in our charts the day we caught it. If you find an error, tell us via the contact page: verified corrections are published within 48 hours with the fix noted on the page.

Independence

No casino pays for coverage or influences a number. Links to operators exist only as citations for published game math and are marked nofollow. PlayStories is not a casino, takes no bets, and sells no gambling services. Content is informational analysis for readers 18 and over; if gambling stops being fun, free and confidential help is available at BeGambleAware and GamCare. See also our editorial standards.

Corrections log

26 July 2026, Spaceman RTP. We published Spaceman’s RTP as 96.50% on four pages and derived multiplier probabilities from it. Pragmatic Play’s own game page states 95%. All four pages were corrected, every derived figure recomputed, and a dated notice added to each. The 96.50% figure is widely repeated across the industry and may reflect an earlier or alternative certified build; where a studio’s current stated figure differs from the consensus, we publish the studio’s and say which is which.

26 July 2026, default RTP versus reported builds. A full audit compared the headline RTP on all 40 game pages against the figure each studio publishes on its own page. 37 matched exactly. Three did not: Sugar Rush 1000 (we said 96.53%, Pragmatic Play publishes 96.00%), Fruit Party (we said 96.47%, Pragmatic Play publishes 96.50%), and Spaceman, logged above. On all three the studio figure now leads, and multi-build ladders sourced from certification databases are labelled as reported rather than verified.

26 July 2026, sourcing claim on the RTP version study. Our RTP version study stated that every figure was verified against provider and PAR-sheet data. It carried no source for that verification. The wording now states the actual provenance: figures compiled from certification databases and published provider documentation, not independently tested by us. Where a studio publishes a figure on its own current page, that figure takes precedence and is what appears on the individual game page.