The VIP Index editorial clarity score measures how much programme information is recorded in our profiles and perk tables. Model 1.0 uses five components totalling 100 points. The weights are our editorial choices; the build applies the same rules to all 39 profiles. Read the components alongside the underlying terms.
Entry and tier clarity: 25 points
Entry earns 8 points when the entry field is anything other than “Programme not publicly confirmed”; a non-empty tier-name list adds 5. An entry note longer than 50 characters adds 4. These checks establish recorded detail, not whether admission is easy.
The remaining 8 points require a numeric cue in the threshold field: a number followed by tier, status, point, credit, crown, wager, XP, USD, EUR or %, or a $, € or £ symbol followed by a number. This pattern detects recorded wording; it does not validate the threshold or distinguish every possible unit.
Reward mechanics: 25 points
The reward check reads the profile’s cashback field and benefit descriptions. Naming points, credits, cashback, rakeback, a rebate, bonus, reload, FanCash or Crown Cash earns 8. A numeric percentage, points or credits followed by “per” or “=”, or a $, € or £ amount earns another 7.
Wording about wagering, house edge, rake, losses, earning, redemption or conversion adds 5. Claim, expiry, redeem, convert, bonus, credit or cash wording adds the final 5. These case-insensitive text cues can overlap. They identify recorded language, including qualified or negative statements; they do not prove a complete calculation or claim process.
Published perk detail: 20 points
Each distinct mapped perk category contributes 2 points, capped at 12. Categories are deduplicated; repeating the same perk cannot raise this part. Each tier row containing at least one perk description adds 2 points, capped at 8. Base rewards outside tier rows also contribute category keys.
A programme with six categories and four described tier rows reaches the cap. More categories do not establish greater cash value, and a described benefit may remain discretionary or restricted.
Conditions and retention: 20 points
Each recorded restriction adds 4 points, up to 8. The combined cashback, threshold, benefit and restriction text earns 4 more for expiry, reset, annual, monthly, retention, inactivity or a qualification-period cue. A market description longer than 20 characters adds 4.
The final 4 come from exclusion, eligibility, dependence, variation, restriction, invitation or “not guaranteed” wording in that combined text. These are text-presence and length checks, not judgments that the restrictions are comprehensive or the programme is available to you.
Evidence confidence: 10 points
A profile marked “indexed-official” receives 4 points because direct source access is limited. Every other evidence value receives 10 under the current function; the build defaults an absent evidence label to “live-official”. The editorial convention is to use that label for reviewed live official sources. Source count adds nothing, and the function does not independently open or authenticate a source.
Worked example: BetMGM Rewards
The current dataset gives BetMGM Rewards 93/100: entry 25, rewards 18, perk detail 20, conditions 20 and evidence 10. Its recorded admission rule, five tier names, numeric Tier Credit thresholds and explanatory note satisfy all four entry checks.
Reward terminology, calculation/redemption wording and conditions cues contribute 8 + 5 + 5. The stored reward text lacks the numeric rate or conversion pattern, so that 7-point check contributes zero. This says nothing about rates outside the recorded fields.
Eleven mapped categories exceed the six-category cap; five described tier rows exceed the four-row cap, giving 12 + 8. Its restrictions, annual reset, market description and eligibility/exclusion wording satisfy the conditions checks. Its live-official evidence classification supplies 10.
Contexts, dates and reproducibility
Market-specific online, international/offshore-facing online and resort programmes have separate lists. Within each context, totals descend; tied scores use programme-name order through JavaScript localeCompare. There is no extra commercial or subjective tie-breaker.
This edition’s profile source-check snapshot is 12 November 2026; the ratings page modification date is 13 November 2026. A page date does not mean every linked term was checked again that day. The downloadable /data/editorial-ratings.json records model version, criteria, totals, components and reasons. Recalculation also requires the same profile/perk inputs and scripts/rating.mjs; the JSON alone is an output, not the full source snapshot.
Questions about the index
Does a higher score mean a safer or more valuable casino?
No. This is neither a licence or safety assessment nor a cash-value estimate. It contains no player-review average, personal-testing claim or promise of winnings. Eligibility and suitability require separate checks.
What can the automated checks miss?
They detect terms in recorded profiles, not audit full live terms. Negation, qualifiers and incomplete descriptions can still trigger points; differently worded facts can be missed. Unpublished or negotiated rates cannot be assumed. Missing points mean the check was not satisfied in this snapshot, not that a benefit never exists.
When will a score change?
When recorded fields, mapped perks or evidence labels change, the next build recalculates the components. Changing the rubric itself requires a model-version update so readers can distinguish new rules from new programme information.
See every editorial score or download the component output. Read the editorial team profile for responsibility and the source method for evidence limits.