How on-chain crypto casino tracking works

Every crypto-casino deposit and withdrawal is a transaction on a public blockchain. Anyone can read those transactions and the balances of the wallets involved — no permission from the operator needed. That single fact is what makes independent, verifiable casino data possible, unlike a traditional casino's private banking.

The hard part is knowing which wallets belong to which casino. We map them only from evidence we can defend: public block-explorer name-tags, operator disclosures, and on-chain behaviour patterns. Wallets we detect as casino-like but cannot tie to a named brand are labelled "unattributed" and kept out of verified figures — never guessed into a brand. See our attribution methodology.

Raw volume is misleading: it double-counts casino-to-casino transfers and includes internal treasury churn. We count only external-facing flow (real deposits and withdrawals with players/exchanges) and exclude wash/treasury-churn operators, so a headline number reflects player activity rather than money cycling internally. Reserves are read across every chain we track and shown with a coverage level, never as a claimed figure. See the live proof-of-reserves hub.

The core technique for finding an operator's full wallet set is common-input-ownership. When a transaction spends from several addresses at once, whoever signed it must control all of them — so those addresses almost certainly share one owner. Starting from a single known casino address (a public name-tag, or a deposit you made yourself), we expand outward through these co-spending links to map the cluster of hot and cold wallets the operator uses. It's a heuristic, not a proof — mixers, shared custodians and exchange wallets can blur it — which is exactly why we corroborate before attributing a cluster to a named brand and label anything uncertain as unattributed.

Most trackers hand you a single number and imply certainty. We do the opposite: every figure carries a coverage level (how much of an operator's footprint we've mapped) and a confidence grade, because honest on-chain data has to admit what it doesn't know. A reserve figure at low coverage is a floor, not a total. A volume figure excludes flows we can't attribute. This is less tidy than a confident leaderboard, but it's the difference between data you can rely on and data that merely looks authoritative — and it's why we never feature an operator we can't stand behind. See how we score confidence.

On-chain tracking is powerful but partial: wallet mapping is never 100% complete, an operator can use wallets we haven't found, and balances are a snapshot that can be funded temporarily. That's why we label coverage and confidence on everything and pair on-chain signals with third-party reputation data rather than treating any single number as the whole truth. See how we score confidence.