Constant re-evaluation of social affiliation is known to cause populations of individuals with different predetermined affiliation preferences to diverge into different network structures. In this study, rather than assigning to each individual a fixed affiliation preference, held throughout the duration of the dynamic network evolution, individuals were allowed an initial “learning period” during which they compared their own relative success, using each of three strategies, at maximizing their social status under three different metrics. Based on the outcomes from this learning period, individuals then chose one particular strategy. The organizational success and stability of the resulting populations was seen to be higher than those of the populations of individuals whose behaviors were predetermined. This indicates that individual-level evaluation and strategy choice in social affiliation preferences can yield strong benefits to the organizational success of the population as a whole.