SlotTrust Score methodology v1.0
The SlotWhizz Trust Score is not an opinion with a number attached. It is the reproducible output of an append-only evidence ledger, computed under the versioned methodology on this page. Everything below is published so it can be checked, argued with and held to account. Effective 2026-07-03.
First evidence-ledger methodology. Trust Score = sum of four component scores (Licensing 35, Payout 25, Editorial 25, Bonus fairness 15) minus an abuse-resistant, time-decayed verified-player-complaint penalty. Each component is resolved from the HIGHEST-PRIORITY current evidence available for that component; confidence reflects how much of the score rests on independent (P1–P2) evidence.
1 · Evidence priority hierarchy
Each score component is resolved from the highest-priority evidence available for it. As stronger evidence is added, it automatically outranks weaker evidence — so the score depends more on independent facts over time, and less on what an operator says about itself.
- P1 Independent measured observation
- P2 Official regulator information
- P3 Verified player evidence
- P4 Public operator disclosure
- P5 Editorial judgement
1b · Where evidence comes from — four channels
Evidence collection is a product with four channels, not one internal process. Every item in the ledger declares which channel it came from.
The Payout Testing Protocol defines exactly how Lab evidence is produced — the channel that turns a payout time into a measured, comparable fact.
2 · Component weights
3 · Verified complaint penalty
A time-decayed, abuse-resistant deduction driven by P3 (verified player) evidence.
- Saturating cap: −16 points maximum, so no campaign can sink a score.
- Independent-reporter requirement: a lone reporter applies only 40% weight; full weight needs ≥2 independent reporters.
- Time decay: report pressure halves roughly every 9 months.
- Per-reporter cap: one reporter contributes at most one unit of pressure.
4 · Confidence model
Every score carries a confidence level — how much the number itself can be trusted, based on the evidence behind it.
- Weighted from: independence (45%), verification (25%), freshness (15%), volume (15%).
- Full independence credit needs ≥2 distinct P1–P2 sources.
- High ≥ 0.66, Medium ≥ 0.4, else Low.
- A score with no independent (P1–P2) source can never read High — a wall of self-stated claims stays Medium at best.
5 · Anti-manipulation rules
- Complaint evidence counts only after human moderation (status=approved).
- Per-reporter cap: one reporter (keyed on ip_hash, the strongest identifier) contributes at most one unit of pressure — brigading cannot stack.
- Independent-reporter requirement: a lone reporter applies only 40% weight; full penalty weight needs ≥2 independent reporters.
- Time decay: report pressure halves roughly every 9 months, so stale grievances fade.
- Saturating cap: total complaint penalty can never exceed 16 points, bounding any coordinated campaign.
- Append-only ledger: no observation is deleted or edited, only superseded — tampering is detectable by replay.
6 · Limitations (the honest part)
- Most operators are currently backed mainly by P4 (operator-stated) and P5 (editorial) evidence. Where that is so, Confidence is Low/Medium by design — the number is an informed estimate, not a measured fact.
- Payout speeds are operator-stated (P4) until SlotWhizz runs its own withdrawal tests (P1). Confidence rises only as measured observations are added.
- Licence tier reflects the class of regulator, not a per-operator audit of ongoing compliance.
- Editorial component (P5) is human judgement and is the lowest-priority input by design.
- Absence of complaints is not proof of safety — it can also mean low volume or low reporting.
7 · Version history
Methodology versions are frozen. New logic ships as a new version; historical scores stay reproducible under the version that produced them.
Machine-readable: /data/methodology.json. Trust Scores are never paid for or negotiated. 18+. Play responsibly.
