How the CCD Trust Score works
Every casino in our index is scored by a deterministic engine across six weighted dimensions. No casino can pay to change its score or its position in a ranking. This page documents exactly how the score is built: the signals we collect, how conflicting data is resolved, how often scores refresh, and what to do if you think something is wrong.
Six weighted dimensions
Each dimension is scored 0–100, then combined using the fixed weights below. Weights are set in code, not per casino, and sum to 100%.
Payout Reliability
Drawn from community withdrawal reports, review-aggregator complaint rates, and forum complaint density.
Dispute Rate
Raw value is dispute percentage (lower = better). Displayed score = 100 − dispute rate. Sourced from review-aggregator complaint rates and forum complaint density, normalised against community post volume.
Transparency
Whether a casino holds an active gambling license and how reputable that licensing jurisdiction is. Premium regulators (MGA, UKGC, Gibraltar, Isle of Man, Alderney) score highest; unlicensed operations score lowest.
Longevity
Domain age as a proxy for operating track record. 5+ years scores highest; new domains start conservative until they build history.
Sentiment
Aggregated review-aggregator star ratings and community sentiment signals, weighted toward recent activity over historical noise.
KYC Clarity
How clearly a casino documents its identity-verification requirements upfront. Both no-KYC and full-KYC casinos can score well here — the penalty is for surprises at withdrawal time, not for having KYC.
Final score formula
The six dimension scores are weighted and summed into a composite 0–100 Trust Score. When you use the AI Advisor, a separate Match Score (how closely a casino fits your stated preferences) is blended in: finalScore = trustScore × 0.6 + matchScore × 0.4. A trusted casino always ranks near the top of a matched result, but the one that best fits your specific needs can rise above a marginally higher-trust casino that doesn't match your requirements.
Where the data comes from
Five active sources feed the pipeline. Each one fails independently — if a source is unreachable, we don't penalize the casino; we fall back to the last verified value.
Community withdrawal reports
Direct, structured reports submitted by players on this site — the hardest signal to fake because it requires specific claim data.
Trustpilot
Review scores and complaint-keyword rates pulled from public review pages.
Sentiment and withdrawal-complaint mentions across relevant crypto-gambling communities.
Domain age (RDAP/WHOIS)
Public domain registration records, used for the longevity dimension.
License registries
Jurisdiction and active-license status checked against the licensing body on record.
Try the formula yourself
Drag the six dimensions and watch the composite score respond — this widget runs the same scoring function as the live pipeline, so it behaves exactly like the real thing.
Stored as a raw dispute percentage — lower is better; the pipeline inverts it internally.
This widget calls computeTrustScore() — the exact function the nightly pipeline runs — with the six weights shown (30/22/18/12/10/8). Real scores use measured signals, not sliders; low-signal casinos are additionally capped by the data-confidence tiers described above.
Data Confidence tiers
A score is only as good as the data behind it. Rather than presenting every score with false certainty, we label how much verified data supports it. You'll see these tags on casino cards and detail pages.
Data Confidence: Fully Verified
Multiple independent sources agree: verified license, community withdrawal reports, and payout data refreshed within the last 30 days.
Data Confidence: Verified
Core facts (license, domain age) are verified but community withdrawal data is limited, or the refresh is 30–90 days old.
Data Confidence: Limited
Needs more data — sparse verified signal, usually a newer or lower-volume casino. The score is a conservative estimate, not a confident rating, until more signal accumulates.
Update cadence
- Automated refresh: every casino's score is recomputed daily against all five sources.
- Community withdrawal reports: incorporated as soon as they're submitted and reviewed.
- Change alerts: subscribers are checked weekly for score, KYC, or payout-status changes worth notifying.
When sources disagree
No single source can dominate a score. Where a dimension has multiple data points — say, a Trustpilot complaint rate and a Reddit sentiment reading — they're averaged together, not cherry-picked.
A missing or failed source doesn't penalize a casino. If a fetcher can't reach a source, that dimension falls back to its last verified value rather than being marked down for our own data-collection gap.
Direct evidence outranks aggregated sentiment. A structured community withdrawal report carries more weight in the payout-reliability dimension than a general review-site star rating, because it requires specific, falsifiable claim data.
Found something wrong?
If a score, license status, or KYC detail looks outdated or incorrect, email trust@cryptocasinodirectory.com with the casino name and what you believe is wrong. We review corrections against source data and update scores when credible new information is available. Full details on affiliate relationships are in our Affiliate Disclosure.
See it in action
Run a query in the Match Engine, or browse the full index by trust rank.