Using a combination of game theory, statistical modeling, and historical analysis, Walter argues that a credible third-party guarantor dramatically increases the likelihood of success for any civil war peace agreement. Walter notes that conflicts and tensions will necessarily arise between any factions who are enemies one day, and forced to coexist peacefully the next day. Without a third-party to monitor and/or enforce the agreement, those conflicts are much more likely to escalate into a full-blown resumption of hostilities. It's a simple theory, explained in a clear and convincing manner, and is an important step forward in our understanding of conflict.