By strategically revealing their preferences, people can manipulate the outcomes of institutions, which may range from formal voting mechanisms to informal social processes. But the consequences of one's actions depend on what others do, and, thus, the preferences people reveal should depend on what preferences they think others will reveal. We now know that this interdependence is fundamental to politics, and one way to model it is to interpret the states of nature that a decision maker confronts so that those states include estimates of other people's probable choices. But if a person conditions his own choices on what he believes others will do, then he should assume that others do the same. Thus, we cannot assume generally that states of nature link alternatives to outcomes in some mechanical way. A science of politics thereby requires a theory about how people condition their decisions on the decisions of others when they believe that others are doing the same. To complicate matters further, people's motives may not be identical. Hence, whether we are talking about voting, candidates in election campaigns, coalitions, or bureaucratic choice, people's fates are both interdependent and conflictual, and any assumption about a benign environment of other persons, who assume the role of nature, is untenable. The hallmark of the reasoning we are about to expound takes both interdependence and conflict into account and assumes that people choose to maximize utility, recognizing their joint dependence. Hence, they take the calculations and dissimilar goals of others into account, and expect everyone else to do the same.