OC Score
OC Score is the planned assessment of a Casino using approved evidence and an explainable methodology. It would help readers see the reasons behind an assessment, rather than ask them to trust an unexplained number.
Scoring is deferred. The TrustScore methodology is a draft for future product, data and legal review. Its scale, formula, metrics, thresholds and public labels are not approved policy, and a manual score is not required for current platform work.
What the draft would assess
The July 2026 proposal groups 20 candidate measures into five areas:
| Area | Candidate measures |
|---|---|
| Legitimacy and transparency | Licence status; operating entity; availability and freshness of terms; restricted-country clarity; responsible-gambling tools |
| Payments and payouts | Method coverage; identity-check requirements; stated processing time; verified actual withdrawals; fees; minimum and maximum limits |
| Terms and bonus conditions | Wagering basis; maximum bet; cashout restrictions; eligibility and exclusions; inactivity, administrative or confiscation clauses |
| Community experience | Verified player ratings; complaint incidence; complaint resolution rate and time |
| Crypto information | Applicable on-chain risk and transparency information from a selected reliable source |
The draft proposes category weights of 25, 25, 20, 20 and 10 respectively and a 0 to 100 scale. These are discussion inputs, not final weighting or a score for any Casino. A fiat-only Casino would not be penalized for an inapplicable crypto category under the proposed normalization.
Unknown does not mean bad
The proposal distinguishes a known observation, information confirmed as not disclosed, an unknown value and a measure that does not apply. It also treats stale evidence, conflicting statements and an insufficient community sample separately.
For example, failing to retrieve a licence page does not establish an adverse licence finding. Likewise, a high score based on only a small share of the required evidence must not look equivalent to a high score with broad support. The draft shows coverage and evidence quality separately rather than multiplying confidence into the score.
It proposes "Insufficient data", "Provisional", published, stale and suspended states. An assessment would need the selected evidence and approval gates before showing a number. Exact gate thresholds and the wording visible to readers need agreement. A profile could remain public while its score is unavailable or suspended, with an explanation.
Evidence, approval and history
Published catalogue facts can supply candidate observations. Scoring separately admits and approves the input and its evidence. A parser result, a draft field or an editorial publication alone cannot change an approved score.
The proposed assessment retains the exact inputs and formula version used. Corrections supersede old observations instead of erasing them. The same retained inputs and formula should reproduce the same result, and an override must keep its reason and previous value. A material change requires human review under the applicable scoring workflow; the draft's numeric change thresholds still need approval.
Community evidence needs its own method
The draft counts distinct verified person-and-Casino experiences rather than every report from the same person. It proposes methods to account for small samples in ratings and complaints, with sample sizes, priors and any capped reputation effect still to be agreed. A partner notification can verify an event without proving that the person's opinion should be positive.
Complaint resolution, rejection as invalid, withdrawal by the user and no response are different outcomes. Reversing a contribution may require a new aggregate and assessment. Helpful votes do not directly enter the draft formula.
What readers should be able to inspect
The proposal would show the total or missing-data explanation, category breakdown, coverage, update date, formula version and short reasons. Source types and checked dates would explain support without publishing private proof, fraud flags or moderator identities. Material changes would have a readable explanation.
Commercial revenue and sponsorship do not enter OC Score or organic ranking. A user's reward balance is also separate from their reputation and from the Casino's assessment. Read commercial disclosure and community rewards for those distinctions.
The scoring review brief asks which evidence, measures and public explanations should be developed when this work is selected.