The five-criteria rubric
Every operator we review is graded on five weighted criteria. The criteria are listed below with their weights.
| Criterion | Weight | What we measure |
|---|---|---|
| Onboarding and KYC | 25% | Time from sign-up to first hand, document rejection rate, re-KYC loops |
| In-session UX | 40% | Hand speed, table stability, lobby clarity, declare-to-payout time |
| RNG and game-state integrity | 15% | Independent RNG certification, deck shuffle transparency, audit trails |
| Payment rails | 10% | UPI / IMPS / cards availability, deposit and withdrawal friction, dispute resolution |
| Responsible-play controls | 10% | Deposit caps, session timeouts, self-exclusion, helpline visibility |
Why these weights
The weights reflect what an adult rummy reader actually experiences. A platform with a clean lobby but a one-hour KYC loop is a worse experience than a platform with a busier lobby and a five-minute KYC. The score follows the experience.
The in-session UX weight (40%) is the largest because that is where the reader spends the most time. The onboarding weight (25%) is the second largest because first impressions shape platform loyalty. The remaining three weights (15% + 10% + 10%) cover the trust and safety layer.
Data sources
Each criterion is scored on a combination of:
- First-party observation — the desk's own accounts at licensed operators.
- Operator-published data — RNG certificates, payout schedules, responsible-play directories.
- Reader-submitted data — via the corrections page, with timestamps and account IDs.
- Independent lab data — RNG certification registries from iTech Labs, GLI, and BMM Testlabs.
Every published score includes the data sources. Reader-submitted data is anonymised but the timestamp and the rubric impact are published.
Correction handling
Corrections are logged on the corrections page with the date, the affected page, and the rubric impact. The desk commits to a 7-day correction turnaround for verified disputes and a 30-day re-review for material operator changes.
The corrections page is searchable by operator and by date. The desk does not delete corrections; corrections are part of the publication record.
The scoring rubric
Each criterion is scored on a 1–5 scale. The weighted total is the platform's overall score.
| Score | Meaning |
|---|---|
| 5 | Best-in-class — passes the rubric bar with margin |
| 4 | Passes the rubric bar |
| 3 | Meets the bar with caveats |
| 2 | Fails the bar in one sub-criterion |
| 1 | Fails the bar in multiple sub-criteria |
Re-evaluation cadence
The desk re-evaluates every operator quarterly or sooner when a material product change is published. The review date is published in the platform review; older reviews are clearly marked.
The desk does not delete old reviews. Reviews are part of the publication record and the corrections page tracks every change.
Methodology questions, answered
Why is the in-session UX weight so high?
Because that is where the reader spends the most time. The score follows the experience, and the experience is dominated by the in-session UX.
How do you handle reader-submitted data?
Reader-submitted data is anonymised but the timestamp and rubric impact are published. Submissions with verifiable account IDs are weighted higher than anonymous submissions.
Can operators see the rubric before publication?
No. The rubric is published; the operator's score is published; the operator's pre-publication review is not part of the process.
What happens when an operator improves?
The desk re-evaluates and updates the score. The corrections page logs the improvement with the new rubric data and the old score.
What happens when an operator gets worse?
The desk re-evaluates and updates the score. The corrections page logs the regression with the rubric data and the affected criterion.