Fraud Scoring

Composite risk scores on players, transactions or traffic sources — triage automation for finite investigation capacity.

Fraud scores aggregate risk signals — device and network fingerprints, behavioral patterns, payment markers, velocity checks — into ranked queues for review or automated action.

Affiliate-program scoring

Source-level scores watch portfolio risk: qualification-failure rates, chargeback incidence, duplicate-account density, behavioral uniformity across a partner’s players (the incent/farm signature). Governance mirrors player-side scoring: thresholds with human review at consequence boundaries (termination, clawback), documented evidence trails, and periodic false-positive audits. The metric to manage: investigation yield per flag — scoring that buries analysts in noise is expensive theater.