How we calculate the WhoPays Score
A reproducible 0–100 score. No casino pays to change it.
The seven components
Fixed weights. They sum to 100. Withdrawal Speed and Player Complaints carry the most.
The Score is 0–100, not a percentage
A WhoPays Score of 81 does not mean "81% of withdrawals succeed." It is a single trust number on a 0–100 scale, weighted across all seven components. We read it in bands:
Where each datapoint comes from
Every component is normalized to 0–100 from sourced data. Each input carries a source and a date — nothing is scored "on a hunch."
| Component | What we measure | Primary source | Last refresh |
|---|---|---|---|
| Withdrawal Speed | Median + P95 time-to-cash, normalized so <15 min ≈ 100 | Our own test withdrawals + player reports | Jun 2026 |
| Player Complaints | Complaint rate per 1k players, weighted by severity | Trustpilot, Reddit, forums, our complaint records | Jun 2026 |
| KYC Complexity | Trigger threshold + documents required, scored low-friction = high | Test accounts + documented player cases | May 2026 |
| Payout Success Rate | Approved ÷ requested withdrawals | Player reports + operator data where disclosed | Jun 2026 |
| Payment Methods | Count + reliability of local cash-out routes | Cashier audit per country | May 2026 |
| Support Resolution | Resolved share + time-to-resolution on stuck payouts | Complaint records + support tests | Jun 2026 |
| Bonus Fairness | Wagering terms that obstruct real-money withdrawal | Editorial T&C review | May 2026 |
Community Proof of Payment (verified withdrawals) enriches and cross-checks these inputs — but the Score does not depend on it. Editorial collection and test withdrawals carry the load.
Any score can be rebuilt from its source data
Take the seven component scores for a brand, multiply each by its fixed weight, and sum. The result is the WhoPays Score — every time, for anyone. There are no manual adjustments, no editorial "feel," and no casino can buy a bump. If a number looks wrong, it traces back to a dated source you can challenge.
Worked example: Betway Nigeria → 88×.25 + 84×.20 + 80×.15 + 86×.15 + 75×.10 + 78×.10 + 70×.05 ≈ 81. See it on the brand page breakdown →
One brand. One baseline. A score for your country.
A casino can pay fast in Nigeria and stall in Kenya — different methods, different KYC, different processors. So each brand has a global baseline Score, then per-country overrides for availability, payment methods, speed and KYC. The country page shows the adapted score_<country>, not the baseline. The question we answer is "will it pay ME, in MY country."
overrides
See per-country scores on country hubs →
Methodology change log
Every change to the formula or weights is dated and public. Weight changes trigger a recompute of affected brands.
| Date | Change | Impact |
|---|---|---|
| 2026-06 | Locked the seven weights; removed all manual adjustments. Score is now formula-only. | All 50 brands recomputed |
| 2026-05 | Added per-country overrides (availability, methods, speed, KYC) → score_<country>. | Country scores split out |
| 2026-04 | Switched Withdrawal Speed from average to median + P95, so one slow tail no longer skews the mean. | Speed re-normalized |
| 2026-03 | Initial WhoPays Score™ published — 0–100 scale, seven components. | Launch |
We take money from no casino
No operator pays to raise its Score, appear higher, or remove a complaint. We earn nothing when you sign up. That's the only way a trust number means anything — if it can be bought, it can't be trusted.
Spot a number that looks wrong?
Every score traces to a dated source. If a datapoint is stale or off, tell us — we'll show our working or fix it.
flag Report a data error