How We Rate
Criteria
Each operator earns a score across six axes: welcome offer quality, ongoing promotions, cashier speed and limits, game-library breadth, support responsiveness and licensing transparency. Each axis has a sub-rubric with concrete thresholds — a cashier axis above 4.5 requires sub-24-hour average withdrawals verified on our own test accounts. The welcome axis weighs the match percentage, the wagering number, the minimum deposit, the max-bet cap and the expiry window; a flashy 300% match with 70x wagering and a 7-day clock scores lower than a 100% match with 25x wagering and 30 days to clear. The game-library axis counts not just title count but studio variety and the share of the shelf that is actually available to the geo in question — a 5 000-title library with 60% blocked for US players is a 3 000-title library for Kaleo's audience, and the score reflects that.
Scoring
Scores sit on a 1-to-5 scale with one decimal of precision. We don't publish a score below 2.0 (we delist operators that would score that low instead) and we don't publish a flat 5.0 — nothing is perfect, and the top of the scale is reserved for operators that would survive scrutiny we haven't thought of yet. Weighted average rounds to the nearest 0.1. The six axes aren't weighted equally: cashier speed and licensing transparency carry about a quarter of the final score each, because a slow payout or an opaque licence invalidates almost every other positive. The welcome offer carries the least weight — a loud bonus is a one-time event, the cashier flow is a lifetime of sessions.
Limitations
What we can't measure directly: long-term RTP on individual sessions, operator practices around dormant accounts and anything happening inside the operator's bonus-abuse detection layer. We flag those explicitly in each review rather than guess. If a reader has evidence on one of those axes, we're glad to update — email the desk with the receipts. We also can't measure the quality of a support interaction we didn't have ourselves, so our support axis is bounded by what two editors can observe in a single quarter; reader-reported experiences feed into the next re-review but don't move the current score mid-cycle.