We treat instructions as an experimental variable to study how their format affects the participants’ comprehension, speed of play, and experimental behavior. We find that, in a public good game, short on-screen instructions requiring forced inputs improve on subjects’ comprehension and...