This article examines how to estimate the probability of faults causing inconsistent state in distributed systems using a variation of the Drake equation. The author demonstrates that even with low individual fault probabilities, systems operating at scale will likely experience state inconsistencies over time, illustrated through a microservices workflow example where a crash during coordinated requests leaves services in conflicting states.