Little is known about the impact of politeness in online communities. This project combines deductive and inductive approaches to automatically model linguistic politeness in online discussion groups and determine the impact of politeness on desired outcomes, such as getting people to reply to one another. We find that politeness triples reply rates in some technical groups, but rudeness is more effective in others. The model can be applied as a "politeness checker" to encourage people to write in ways likely to garner a response from specific communities. Keywords Linguistic politeness, computer-mediated communication, community responsiveness ACM Classification Keywords H.5.3 [Information Interfaces]: Group and Organization Interfaces - Collaborative computing, Web-based interaction, Computer-supported cooperative work