We open the free demo and note the starting balance. We spin the base game long enough to watch the tumble mechanic chain and to see how often nothing lands. We trigger the free-spins round the honest way, on scattered symbols, and where a game offers it we buy the feature to see what the fixed price actually returns. We read the return-to-player line, the bet range and the max-win rule straight off the game-info panel. We count things: how many tumbles a sequence ran, how high a multiplier climbed, how long a dry streak lasted. Then we cross-check every figure against the provider's own product page and any independent listing, and we flag every disagreement instead of papering over it.
How we test games: we play the free demo, log the mechanics spin by spin, and rely on published Ninja Gaming figures rather than affiliate marketing claims.
Demo first, verdict last: the order we work in
We play the game before we write a word about it. For every slot we cover, we load the free demo, deposit the play-money credits, spin through a real session, read the in-game rules and paytable panels line by line, and log what we find against the provider's published spec. Only then does the review get written. The score out of ten comes last, after the numbers are on the page, never before.
That order matters. It is easy to decide a game is an eight and then hunt for reasons. We do the reverse. Dede 5000, the high-volatility tumble slot from Ninja Gaming we are reviewing right now, is a good example of why: the marketing headline reads "5000", but the in-game panels showed us that figure is the top single tumble-multiplier value, not the win ceiling. We would have missed that reading press copy alone.
Our process, step by step
What we measure in a session
We record a handful of concrete observations from each demo run, and we are careful to label them for what they are. A demo session is a small, illustrative sample, not a certified simulation over millions of spins, so we never present it as proof of a hit-rate. What it does show is texture: how the game feels, how brutal the variance runs, whether the bonus is worth chasing.
On the current Dede 5000 build we deposited the demo credits, counted tumble sequences that chained past a dozen cascades, bought the free-spins feature for its fixed price to see the round from the inside, and watched the persistent multiplier climb from ten to sixteen across a single bonus. We report figures like that as measurements from one session, with the caveat attached, rather than dressing them up as a statistical claim.
Which sources we trust, and in what order
Not all sources are equal, and when they disagree we follow a strict order of trust. The in-game panels sit at the top. The rules, paytable and return-to-player screens that render inside the live demo are the primary source, because they are the operator's own binding statement of how the game pays, read at version level. When an affiliate page and the in-game panel disagree, the panel wins every time, no exceptions.
Below that come the provider's first-party materials: the official product page, the developer's published game spec and API record. These carry weight because the studio is stating its own figures on the record, though we treat marketing claims about operator counts or vague "certified" language with a raised eyebrow until something independent backs them.
Next are the primary registries and independent aggregators. A regulator's licence database, a company register or a recognised RNG-certification body settles a licensing or fairness question far better than any review site can. General slot-aggregator listings we treat as useful corroboration, one voice among several. Where a provider publishes 95.75% and an aggregator lists a different figure, we lead with the primary panel and name the discrepancy openly.
At the bottom, and effectively excluded, is affiliate marketing copy. We do not take a max-win number, a release date or a fairness claim from a page that earns a commission on it, our own included. If a figure exists only in affiliate write-ups and nowhere primary, we treat it as unverified and say so on the page rather than laundering it into a fact.
What our score out of ten means
The score is one reviewer's considered verdict, weighted across the things that decide whether a slot is worth playing. Return-to-player carries real weight: a base figure under 96% is a mark against a game, however good the bonus looks. We weigh the volatility profile against who the game is for, the max-win ceiling against how rare it truly is, and the licensing picture on how much protection a player actually gets.
Originality matters too. A slot that clones a better-known title beat-for-beat loses points against the original it copies, though a clean, playable clone is still a clean, playable game. Dede 5000 lands at 7.0 on our scale for exactly this mix of reasons: a working free-spins engine and a genuine 15,000x ceiling pull it up, while a sub-96% base return, a derivative Gates-of-Olympus formula and soft-touch Curacao licensing pull it back down. We would rather publish a defensible 7.0 with the reasoning shown than an inflated 8.5 nobody can check.
What a score is not
- It is not a prediction. No score tells you whether your next session wins or loses, because a certified random number generator makes every spin independent.
- It is not for sale. The number is unaffected by any affiliate relationship, and a poorly rated game keeps its poor rating.
- It is not final forever. When a provider ships a new return-to-player variant or a game changes, we re-test and re-date the review rather than leaving a stale figure standing.
If you want to see the process applied end to end, the full Dede 5000 review shows the demo findings, the paytable maths and the verdict together in one place.