A transparent, explainable system for finding slots that feel similar — not just look similar
Most "similar slots" lists are based on editorial opinion — someone picks games that seem related and writes them up. There's nothing wrong with that, but it doesn't scale, and you can never tell why a slot was recommended.
SlotMatcher takes a different approach. Every slot in our database is tagged across seven measurable axes. When you ask "what's similar to Sweet Bonanza?", we calculate a weighted similarity score against every other slot and rank the results. The formula is public. The breakdown is shown on every match card. You always know exactly why a slot was matched.
The 7 axes of similarity
Each slot is profiled across these dimensions. The score reflects how much DNA two slots share.
How wins are formed in the base game: paylines, ways-to-win, cluster pays, megaways, scatter pays, or xWays. This is the most important axis because it defines the fundamental feel of every spin. Two slots can share the same theme and volatility but play completely differently if one uses paylines and the other uses cluster pays.
Method: Jaccard index — measures overlap between two sets of mechanic tags. 1.0 = identical mechanics, 0.0 = nothing in common.
The free spins and bonus round structure: expanding symbols, sticky wilds, progressive multipliers, hold-and-win, collect symbols, respins, retriggers, xSplit, xBomb, and free-spins-pick modes. Two slots can use the same base mechanic but deliver a completely different bonus experience.
Method: Jaccard index on bonus mechanic tags.
Risk level: low, medium, high, or very high. This directly affects session feel — how often you win, how big the droughts are, and how explosive the bonus rounds can be. A player who loves the rollercoaster of a very-high volatility slot won't get the same feeling from a low-volatility game, even if the mechanics are identical.
Method: Ordinal distance — same bucket = 1.0, adjacent = 0.5, two or more apart = 0.0.
The maximum win in multiples of bet (e.g., 5,000× or 100,000×). This reflects the mathematical ceiling of the slot's payout structure. Slots with similar max wins tend to have similar win distributions.
Method: Numeric proximity — 1 minus the absolute difference divided by range, clamped to 0–1.
Return to Player percentage — the theoretical long-term payback rate. We store the default/top-tier RTP from provider specifications. A 2% difference in RTP (e.g., 96.5% vs 94.5%) is meaningful over extended play, so the formula penalizes larger gaps accordingly.
Method: Numeric proximity with divisor of 3 — a 3% RTP difference = 0% similarity on this axis.
Visual and narrative theme: mythology, candy, western, horror, fishing, etc. Theme is weighted lower because two slots with the same theme can play very differently. But when everything else is close, theme can be the tie-breaker that makes a recommendation feel right.
Method: Jaccard index on theme tags.
Same game studio (Pragmatic Play, Play'n GO, NetEnt, etc.). Provider gets the lowest weight because most players looking for "slots like X" want the same feel from a different provider, not just another game from the same studio. But same-provider slots often share design language and math patterns, so a small boost is appropriate.
Method: Binary — 1.0 if same provider, 0.0 if different.
The formula
Score = 0.30 × core_mechanics
+ 0.20 × bonus_loop
+ 0.18 × volatility
+ 0.10 × max_win
+ 0.10 × rtp
+ 0.07 × theme
+ 0.05 × provider
Each component produces a value between 0 and 1. The weighted sum is multiplied by 100 to produce a percentage score. Matches below 40% are not shown.
Score interpretation
Example: Book of Dead vs Legacy of Dead
Both are Play'n GO payline slots with expanding symbols, retriggers, gamble feature, Egyptian mythology + adventure theme, and high volatility. The only difference is RTP (96.21% vs 96.58%) and max win (both 5,000×).
Result: 98.8% similarity — the highest match in our database. The 1.2% gap comes entirely from the RTP difference (0.37% apart, producing ~88% on the RTP axis at our divisor of 3).
Compare this with Book of Dead vs Sweet Bonanza: different core mechanics (paylines vs scatter-pays + tumble), different bonus structure, different theme, different provider. Score: under 40% — correctly filtered out as not similar despite both being popular high-volatility slots.
Data quality
Every slot in our database is manually tagged using a standardized synonym dictionary that maps provider-specific terminology to normalized internal tags. For example, "Avalanche" (NetEnt), "Cascading Reels" (generic), "Tumbling Reels" (generic), and "Reactions" (BTG) all map to the same internal tag: tumble.
All data has been cross-reviewed against multiple independent sources including SlotCatalog, BigWinBoard, Casino.guru, and official provider documentation. RTP values reflect the default/top-tier version from provider specifications — note that some casino operators may offer lower RTP configurations.
The formula weights were calibrated through cross-validation: we verified that known near-matches (like Book of Dead and Legacy of Dead) score 90%+, while known non-matches (like Book of Dead and Sweet Bonanza) score under 40%.
Limitations
No similarity system is perfect. Here's what SlotMatcher doesn't capture:
Session feel beyond volatility. Two high-volatility slots can have very different hit frequencies, bonus trigger rates, and dead-spin patterns. Our volatility bucket (low/medium/high/very high) is a useful approximation but doesn't capture every nuance.
Visual and audio experience. Graphics quality, animation style, soundtrack, and overall production value are subjective and not scored. A beautifully animated slot and a basic-looking one with identical mechanics will score the same.
Operator-specific RTP variations. We store the default/top-tier RTP, but the actual RTP you experience depends on which version your casino has configured. Always check the paytable at your chosen casino.
We're continuously improving the formula and expanding the database. If you think a match is wrong or a slot is missing, we want to know.