Are communication failures common? We revisit a classic example of experimental coordination failure and explore, in a 2x2 design, the effects of deviation costs and loss avoidance. Our results suggest how to engineer coordination successes in the laboratory, and possibly in the wild.