Approaching warfare in pre-modern states from the perspective of risk reduction, we see that royal marriage was one strategy rulers used to reduce the probability that they would lose a war. Judicious marriage exchanges intensified and prolonged patron-client relations between rulers or between rulers and societal elites. Clientelism could affect the size and composition of their armies. The more warriors and troops one could field, the greater the chance of not losing a war (Otterbein 2004; LeBlanc 2006). Examination of eight pre-modern states suggests that their rulers used the same patterns of wife exchange even though most states developed independently. Marriage secured long-term patron-client relationships, which they used to support their military efforts. When rulers married their kin or married them to rulers outside the system ("foreigners"), they did not gain military support. Analysis of these marriage-military patterns reveals several characteristics of pre-modern states. First, marriage alliances helped rulers form networks of support that helped them win wars. Therefore, marriage-and by extension, royal women-is a key component to the study of warfare and a critical mechanism of network formation, as Blanton et al. (1996) write. Second, alliances were based on a different organizing principle from Levi-Strauss' tribal societies, for rulers selected main wives (for themselves or their kin) based on relative rank rather than particular kinship ties. Third, marriage alliance reveals an important difference between alliance and patron-client relationships, a distinction that is often blurred in the archaeological literature.