In this chapter, I suggest and find evidence for the contention that what is regulated in marriages that are stable is some ratio of positive to negative problem solving or positive to negative purely affective behaviors. Furthermore, it appears that there are only three distinct possible adaptations to regulating this ratio: one by using a lot of negative and positive affect, another by using a moderate amount of positive and negative affect, and the other by using a small amount of positive and negative affect. The ratio of positivity to negativity is a constant of about 5.0 in the three types of stable marriages, and less than 1.0 in the unstable marriages. This is true across conversations and coding systems. I argue that these three stable adaptations are distinct types rather than dimensions, because there are so many concomitants of each style or type. I speculate about the benefits and risks of each adaptation.