Counterfeit Detector: Risk-Score & ROC Plane

Brand-protection teams don't check every online listing by hand — a crawler scores each one against a handful of measurable signals and only escalates the riskiest for human review. This simulator recreates that pipeline as a pair of 2D analytics panels instead of a 3D scene: a synthetic marketplace of listings is plotted by seller-trust score against a logistic risk score computed from image/logo similarity, price deviation and trust together, while a continuously running crawler scans them at a chosen rate and routes anything past your threshold into quarantine. A second panel sweeps that threshold across every scanned listing in real time and draws the resulting ROC curve, the standard graph a real detection team uses to grade a classifier's true/false-positive trade-off independent of any single operating point.