The sample-size receipt behind every score
Every numeric claim on CryptoCasinoi (the 2,127 timed withdrawals, the 4,800 hours of gameplay, the $50,000 of bankroll committed, the $847,420 settled to wallet) traces to the editorial audit spreadsheet maintained by the CryptoCasinoi Editorial Team. The spreadsheet itself stays internal for player-privacy reasons; the aggregate breakdown is published here. This page is the transparency receipt for every figure quoted across the site.
The cross-program audit window
Six numbers anchor every claim made on the site. They cover the rolling 90-day cycle inside which every score and every ranking is calculated.
Filtered from 187 crypto-accepting brands at protocol step 1.
Sample N ≥ 96 per operator; per-network breakdown below.
Editorial bankroll deposited across the window. No press accounts, no comped balances.
USD-equivalent value that arrived in our wallets across the audit window.
4,200 audited sessions across slots, live, and provably-fair.
Listed at /blacklist with rejection reason; the receipt for every rejection.
The 8 ranked operators, transaction by transaction
Sample sizes vary by operator because every brand sets a different floor on the test cadence (Lightning-first operators yield large N quickly; mainnet-only operators take longer to reach the protocol minimum). All rows are taken from the 90-day audit window.
| Operator | Timed tx | Fast-rail median | First-pass success | Primary rails tested | Disclosed KYC trigger |
|---|---|---|---|---|---|
| Cinoslots | 318 | 4 s | 99.6% | Lightning (primary) + Solana SPL (n=84) | $20k equivalent cumulative |
| Bitstarz | 262 | 11 s | 99.4% | Lightning + BTC mainnet + BEP-20 USDT | 2 BTC cumulative |
| Cloudbet | 248 | 7 s | 99.5% | Lightning (primary) + BTC mainnet | 1 BTC equivalent |
| Gamdom | 264 | 8 s | 99.3% | Solana SPL (primary) + USDT | Risk-based (untriggered) |
| Duelbits | 284 | 7 s | 99.4% | Solana SPL + Arbitrum + BTC mainnet | $15k–25k cumulative |
| Wild.io | 184 | 42 s | 98.7% | USDT TRC-20 (primary) + BTC | Hidden threshold (flagged) |
| 22bet | 192 | 5 min 12 s | 98.2% | USDT TRC-20 (primary) + BTC mainnet | $2k rolling 30-day |
| 1xbet | 215 | 8 min | 97.6% | BTC mainnet (primary) + USDT mixed | $2k rolling 30-day (aggressive flagging) |
| Top-8 total | 1,967 | Plus the deposit-tested blacklisted-operator sub-samples (160 tx) = 2,127 total. | audit window total | ||
Per-operator transaction logs include the timestamp, source network, destination network, sample-side wallet hash (one-way salted), bet sizing during play, and time-to-balance delta. Player-side identifiers are not retained. Aggregate counts are stable; raw per-transaction rows are available on documented request (see access policy below).
Withdrawal medians by network across the full 2,127 sample
Each row is the network-level median across every operator that surfaced the network in their cashier during the 90-day window. Network overhead (confirmation latency) is logged separately from operator overhead so we can attribute slow medians to the right side of the stack.
| Network | Tx logged | Median time | 90th percentile | Median fee share | % of program sample |
|---|---|---|---|---|---|
| BTC Lightning | 795 | 11 s | 38 s | 0.01% | 37.4% |
| Solana SPL | 312 | 8 s | 22 s | 0.04% | 14.7% |
| BTC mainnet | 286 | 11 min 04 s | 47 min | 0.18% | 13.4% |
| USDT TRC-20 | 388 | 38 s | 3 min 12 s | 0.03% | 18.2% |
| USDT ERC-20 | 96 | 2 min 41 s | 11 min | 0.42% | 4.5% |
| ETH mainnet | 84 | 3 min 18 s | 14 min | 0.61% | 3.9% |
| Arbitrum (ETH/USDC) | 112 | 47 s | 2 min 18 s | 0.08% | 5.3% |
| LTC mainnet | 54 | 2 min 22 s | 6 min | 0.04% | 2.5% |
| All networks | 2,127 | Sample covers 2,127 of 2,127 transactions; the remainder were on tail-networks (BCH, DOGE, XMR off-ramp, etc.) with N too small per network to publish a median. | 100.0% | ||
The four no-KYC traps, incidence inside the 412-withdrawal sub-sample
These figures back the four-trap framework on the no-KYC blog post. The sub-sample isolates the 412 withdrawals tagged for no-KYC scrutiny inside the program's 2,127 total.
Where the 4,800 hours were spent
Gameplay is split deliberately across slots, live dealer, provably-fair native games, and table verticals so the Trust Index reflects every product line at the operator. Sessions are logged with timestamp, bet size, game ID, and outcome variance against the published RTP.
| Vertical | Hours | Audited sessions | Editorial note |
|---|---|---|---|
| Slots | 2,680 | 2,340 | 120+ studios across the audited operators; per-title RTP variance logged against published RTP. |
| Live dealer | 1,180 | 980 | Evolution + Pragmatic Live dominate the sample; per-table latency and dealer-rotation logged. |
| Provably fair | 640 | 720 | Dice, Crash, Mines, Plinko, Limbo across operators with native house games; seed-verifier integrity audited. |
| Table (non-live) | 300 | 160 | Sample exists primarily for game-weighting rule audit (slots vs table contribution to bonus clearance). |
When each figure is refreshed
Different inputs move at different speeds. Withdrawal-speed medians need the most frequent re-audit; methodology versioning is the slowest. The cadence below is the published schedule the editorial team holds itself to.
- DailyOperator status flag (live / suspended / pending re-audit)
- WeeklyBonus terms (welcome, reload, no-deposit)
- Every 30 daysWithdrawal-speed re-audit on every ranked operator
- Every 90 daysFull Trust Index re-score against the seven-criterion rubric
- Per releaseMethodology version bump and changelog entry
Why we do not publish the spreadsheet, and how to request a slice
The raw audit spreadsheet contains operator-side risk-flag triggers we do not want publicly indexed, and hashed sample-side wallet metadata which (although salted) is not safe to publish in bulk. The aggregate summary on this page is the published surface. Verifiable slices are available on documented request from journalists, regulators, and academic researchers.
Email [email protected] with the outlet and the specific claim being cross-checked. Slice is delivered inside 48 hours under a no-redistribution undertaking.
A regulator's formal request via United Kingdom channels receives the full operator-level transaction log, including timestamps and confirmation hashes. No undertaking required.
Academic and policy researchers receive anonymised, network-level slices for replication studies. Acknowledgement of source is the only requirement.
The methodology that generated this dataset is published in full at /methodology; the version stamp on every score is v3.3, last revised . Every figure on this page is the single source of truth for the rest of the site: when a number is revised here, the methodology page, every operator review, every category landing and every blog post that cites it update in the same release. No stale figure is left behind.