CasinoLens methodology: scoring, evidence and publishing
How we compare platforms, review promotional terms and assess game information. Current brand model: brand-evidence-4.1; ranking policy: single-public-score-1.
01 · Scope of the assessment
CasinoLens assesses traceable public information: disclosed rules, service conditions and completeness. Platform ratings, game profiles, offers and news are separate products. A game supplier’s credentials do not establish an operator’s licence, and a large bonus is not a quality score.
We have not completed real-money deposit or withdrawal testing, systematic customer-service testing, or a verified user-rating programme. Scores are editorial information assessments, not safety certification, protection of funds or hands-on service ratings.
02 · Seven dimensions and 25 core criteria
Payments and withdrawals carry 25%, covering methods, limits, timing, fees, KYC and records. Identity, service, games and offers each carry 15%; account protection carries 10%, and technical access and localisation 5%. The table follows the live model weights.
Domain history, external feedback, cryptocurrency deposits and withdrawals, bonus cancellation, game failures and marketing opt-outs remain supplementary information without separate core weight. Deposit support does not establish withdrawal support; the same fact must not earn duplicate points.
| Dimension | Weight | Core criteria and weights |
|---|---|---|
| Operator and history | 15% | Operator 11; history 4 |
| Payments and withdrawals | 25% | Methods 5; limits 6; timing 4; fees 3; KYC 5; records 2 |
| Service and complaints | 15% | Human support 4; hours 3; languages 2; complaints 6 |
| Games and suppliers | 15% | Categories 7; suppliers 8 |
| Offers and terms | 15% | Clarity 6; consistency 3; restrictions 4; changes 2 |
| Account protection | 10% | Freezing rules 3; privacy 4; limits and exclusion 3 |
| Technical access and localisation | 5% | Access 2; installation 1; localisation 1; maintenance 1 |
03 · Anchors and missing information
Each criterion has a value of 100, 50, 0 or unknown. A value of 100 means the relevant information and scope are clear; 50 means an explicit statement lacks important restrictions, steps or conditions. Zero requires evidence of failure against the criterion or a relevant adverse condition. Missing or inaccessible information remains unknown.
A withdrawal timeframe without its approval conditions may be partial disclosure. No fee statement means unknown fees, not a zero or free service. Conflicting specific and general offer rules can receive zero for consistency while both versions remain visible. One zero does not establish fraud.
04 · Calculating the composite score
Let K be the total weight of known core criteria and E the sum of each weight multiplied by its anchor percentage. Known-information score = 100 × E ÷ K. Evidence score = min(100, known-information score + licence-information bonus). Composite score = min(100, evidence score + brand awareness bonus). With K = 0 there is no meaningful score.
Illustration: K = 60 and E = 36 produce a known-information score of 60. A licence bonus of 2 and awareness bonus of 8 produce a composite score of 70 with 60% coverage. This is a hypothetical example, not a brand result.
Unknown criteria are excluded from the score denominator but reduce coverage. A score of 70 at 30% coverage rests on less information than 70 at 90%. New evidence can raise or lower the result. Displayed stars are the composite score divided by 20, not user votes.
05 · Bonuses and ranking eligibility
Licence-information tiers award 0.5 for a regulator logo or generic licence image, 1 for identifiable holder and licence fields in an image, 2 for a disclosed number, holder and issuer, 3 for a matching regulator record without confirmed domain or scope, and 5 for an active record explicitly covering the target domain. Applicable archived evidence must pass validation. A licensing slogan, supplier logo or test-laboratory logo does not automatically earn an operator bonus.
The editorial brand awareness bonus is 0–10 points, with at most one decimal place. It is added once and does not alter criterion weights, coverage, confidence or eligibility. It is not an independently measured traffic or audience statistic.
Provisional ranking requires at least 20% core-weight coverage across at least three dimensions. Eligible brands are ordered by the same displayed composite score, without a second ranking reference score. Equal scores share a rank and subsequent positions skip accordingly. Thin evidence makes rankings more sensitive to updates.
06 · Source strength and unresolved conflicts
Source-strength values are 60 for platform or related official statements, 70 for identified independent sources with reviewed scope, 80 for documented experience, 90 for regulator records and 35 for unclassified sources. A criterion without archived references receives 25. These are editorial parameters, not probabilities of truth.
Each criterion uses its strongest applicable source, without adding points for duplicates or syndication. An unresolved conflict multiplies its strength by 0.75. Overall confidence is the weighted average across known core criteria. Unreviewed non-platform source classifications are treated as unclassified.
Original links, capture times and available text or screenshots support review, with saved-file integrity checks. Operator claims, reporting, registry records and actual experience remain distinct. Different domains, regions, languages and product versions are not interchangeable without checking scope.
07 · Reviewing offers, games and articles
Each offer retains its own amount, deposit threshold, wagering multiplier and calculation basis, eligible games, withdrawal cap, validity, eligibility and sources. Specific rules are compared with general terms; conflicting numbers and missing dates remain explicit. Capture dates are not expiry dates, and no deposit does not mean unrestricted withdrawal.
Game review covers supplier, mechanics, RTP configuration, layout and bonus features. A demo link does not establish that a real-money configuration is identical. RTP is a long-run theoretical measure, not a prediction for one session. Game profiles currently require manual review and publication.
News separates event, announcement and capture dates. The upgraded workflow uses multiple sources, separate language generation and publication checks. Guides use practical steps, labelled hypothetical examples and FAQs. Available source imagery is preferred, with demo screens distinguished from illustrations.
08 · Review, multilingual publication and history
Collection, drafting, review and publication are separate states. Pending drafts must not appear as published entries or rankings. Records link to existing brands, offers and games to avoid duplicates. Publication uses database release records and file validation, rather than treating browser visibility alone as completion.
Chinese, English and Malay retain the same amounts, deadlines, wagering conditions, sources and unknowns while allowing natural local wording. Saving the guide directory updates matching homepage titles, summaries, covers and links.
Source changes, conflicts or association errors call for a record revision, recalculation where relevant and checks of affected pages. Scores are not continuous monitoring; historical records describe their captured scope. Weights and bonuses are editorial choices, not regulatory standards, and methodology copy must change with the rules.