This paper investigates the failure of blockchain smart contracts to deliver fair and transparent trading environments, focusing on decentralized exchanges (DEXes). It documents the widespread deployment of arbitrage bots that mimic high-frequency trading strategies, exploiting inefficiencies through frontrunning and transaction reordering. The research quantifies bot activity by analyzing transaction data to measure revenue and strategies, highlighting the competitive bidding in priority gas auctions (PGAs) where bots pay high fees for early execution. This leads to the formalization of PGAs as a continuous-time, partial-information game-theoretic model, providing a framework for understanding these dynamics.

Furthermore, the paper introduces the concept of miner extractable value (MEV) as a broader phenomenon beyond PGAs, representing systemic risks to consensus-layer security in blockchain systems. Through empirical analysis, it demonstrates that MEV poses a realistic and immediate threat to Ethereum, undermining the stability promised by decentralized architectures. The study concludes by emphasizing the adaptation of traditional financial exploitation methods into blockchain economies and releases an interactive web portal (http://frontrun.me/) to offer real-time data on PGAs, fostering community awareness and further research into mitigating these risks.