Methodology
Where every number on rtpspy.com comes from, and how much weight to put on it.
Live numbers (hit rate, big wins, current RTP)
These are ours. We ingest real bets from crypto casinos as they are placed, so the live figures on a slot page (spins in the last 30 minutes, hit rate, recent big wins, biggest 24-hour multiplier, all-time records) are counted from real play. We measure them ourselves. Nobody hands them to us, and you will not find them anywhere else.
The feeds come from 21 crypto casinos: Stake, Gamdom, Shuffle, BC.Game, Razed, Thrill, JackBit, Degen, Roobet, Rainbet, Duelbits, Cloudbet, Wild.io, Winz.io, Bets.io, Playamo, 7Bit, BitStarz, Woo, King Billy, Oshi. Bets are normalised into USD using live exchange rates, capped at each slot's published maximum-win multiplier (so casino rounding artefacts can't fake a record-breaking spin), and deduplicated before they hit the database. Every bot writes a per-minute health snapshot we audit daily.
What a live hit rate is not: it is what players are getting right now, over a limited number of spins. It is not the game's long-run maths, and it does not predict your next spin.
Published specs (RTP, max win, volatility, reels, bet range)
These are not ours to measure. A slot's RTP is set by the studio, and confirming it by play would take millions of spins. So we take each figure from the studio's own material and from independent industry sources, and we publish it only when they agree. When they disagree, the field stays blank and the disagreement is logged. We'd rather show nothing than something wrong.
Hit frequency is the one spec sources routinely differ on. Where we show it, it is the middle value of the sources we hold, and it is an estimate rather than a confirmed figure.
For newly announced slots that aren't playable yet, we use provider-disclosed values and clearly flag the page as pre-release. Once the slot goes live, the same cross-check runs again on the released figures.
Reviews
Reviews carry the Mumbles byline and follow one structure: provider, headline stats, gameplay, bonus features, honest pros and cons, and a short verdict. Every number quoted in a review is pulled from the cross-checked fields above, never written freehand, so a review cannot cite a figure the page itself doesn't hold.
Every draft must pass an automated quality score before it can go live. Drafts that fail stay hidden and the page shows stats only. There is no manual sign-off on each individual review, so if you find something wrong, tell us and we will fix it: hello@rtpspy.com.
A review is rewritten when a slot moves from upcoming to released, and when a fact check finds a published figure has changed. There is no fixed refresh cycle beyond that, so an older review can be months old. Every slot page prints the date its review was last written, directly under the title, so you can judge the age for yourself.
What we don't do
- We do not accept payment for favourable reviews.
- We do not promise wins, payouts, or returns. Slots are random.
- We do not publish a spec our sources disagree on. The field stays blank rather than guessed.
- We are not a gambling site. We track and analyse public bet data.
Corrections
Spot a number that looks wrong? Email hello@rtpspy.com and we'll investigate within 24 hours. Any stat change goes through the same cross-check before it goes live.