Fraud and Bot Detection in Gaming Marketplaces

Abstract

Cloud-scale game marketplaces face increas-ing threats from automated activity and coordinated abuse, including repeated redemption of gift coupons and single- use discount codes, as well as the creation of free or secondary accounts to inherit starter currencies or in- game assets. This paper surveys practical detection ap- proaches suitable for real-time production environments, emphasizing observable system signals rather than model- heavy or AI-driven methods. We describe techniques for monitoring API call sequences and rates, simple behav- ioral signals such as keystroke latency, and graph-based linkage analysis to detect clusters of accounts exploit- ing shared vouchers or resource transfers. Operational tradeoffs including latency, privacy, and computational cost are discussed, along with strategies for mitigating abuse through tiered interventions. This work provides a framework for deploying effective fraud detection while preserving legitimate user experience in rapidly evolving game marketplaces.