Title: Economically Viable Randomness

Abstract: We study the problem of providing blockchain applications with \emph{economically viable randomness} (EVR), namely, randomness that has significant economic consequences. Applications of EVR include blockchain-based lotteries and gambling. An EVR source guarantees (i) secrecy, assuring that the random bits are kept secret until some predefined condition indicates that they are safe to reveal (e.g., the lottery's ticket sale closes), and (ii) robustness, guaranteeing that the random bits are published once the condition holds. We formalize the EVR problem and solve it on top of an Ethereum-like blockchain abstraction, which supports smart contracts and a transferable native coin. Randomness is generated via a distributed open commit-reveal scheme by game-theoretic agents who strive to maximize their coin holdings. Note that in an economic setting, such agents might profit from breaking secrecy or robustness, and may engage in side agreements (via smart contracts) to this end. Our solution creates an incentive structure that counters such attacks. We prove that following the protocol gives rise to a stable state, called Coalition-Proof Nash Equilibrium, from which no coalition comprised of a subset of the players can agree to deviate. In this stable state, robustness and secrecy are satisfied. Finally, we implement our EVR source over Ethereum.