How Detect The Bet classifies merchants
What we classify
Detect The Bet maintains a registry of merchants involved in gambling and gambling-adjacent activity that standard merchant category code (MCC) controls cannot see. Each entry is an entity (the operator behind the charge), not just a descriptor string: operators rebrand, rotate billing names, and change processors; the registry follows the entity through those changes, and every classification carries its history.
Categories
| Category | Description |
|---|---|
| Sportsbook | Licensed sports betting operators |
| Fantasy sportsbook | Daily fantasy and pick'em style operators |
| Sweepstakes casino | Casino-style games using sweepstakes currency mechanics |
| Online casino | Casino operators, including offshore and crypto-rail |
| Prediction market | Event-contract trading platforms |
| Lottery | Lottery and iLottery services |
| Skin gambling | Wagering with game-item currencies |
| Mystery box | Chance-based product boxes billed as retail |
| Loot box funding | Platform currencies spent on chance mechanics |
| Crypto onramp | Funding rails observed feeding gambling activity |
| Prop trading | Fee-at-risk "challenge" trading services |
Categories are descriptive, not legal conclusions. We do not label legally ambiguous products as illegal; we characterize what the merchant is and how it bills.
How evidence works
- Every classification carries a confidence level tied to how it was verified. The highest tier means we directly observed the merchant's billing behavior. Lower tiers mean reported by credible sources, or inferred and not yet verified.
- A block recommendation is reserved for our highest-confidence classifications. Entities we have not yet verified to that bar are surfaced for review rather than declined, so a thin or unverified match never hard-blocks a customer.
- Every classification records its supporting evidence with lineage: when the entity was first seen, every classification change since, and why. Past decisions are reconstructable.
- Classifications recommend; institutions decide. Detect The Bet identifies and characterizes activity. The financial institution sets the policy response (block, review, allow) according to its own risk tolerance and its customers' choices.
Ambiguity and behavioral signals
Some activity cannot be attributed to a nameable merchant. These cases are flagged for human review only and are never automatically declined. The same is true of all behavioral signals: they queue a case for the institution's review with a plain-language reason; they never block a transaction.
Corrections and appeals
Misclassification claims from merchants or institutions are investigated against the evidence record, answered within a stated window, and resolved with a lineage entry either way, so the history shows the claim and its outcome. False positives are treated as seriously as misses: the registry is curated conservatively, and a bias against false flags is a design principle, not a tuning accident. To report a suspected misclassification: alex@detecthebet.com.
What institutions control
Which categories apply, what each recommendation maps to, which confidence tiers act automatically versus queue for review, and every final authorization decision. No customer transaction data or personal information flows to Detect The Bet in the feed's operation: the feed is delivered signed to the institution and evaluated entirely inside its environment.