Traditionally, robots have been programmed to do precisely what their human operators instruct them to do, but more recently, they have become more sophisticated, intelligent, and autonomous. Once they reach a sufficiently high level of intelligent autonomy, they can support more collaborative interactions with each other and with people. As robots become more and more intelligent, we will begin designing systems where robots interact with humans, rather than designing robots that are commanded by people with continual oversight. One approach to assessing how humans and robots will interact in the future is to frame the problem as a collection of intelligent nodes. Multiple, collaborating, and interacting manned and robotic systems can be represented as a collection of dynamic, interacting nodes. This paper develops preliminary metrics to support understanding the extent of preferential attachment that would arise in a system of cooperating manned and unmanned systems (MUMS). The metrics seek to help explain if attachments are localized to specific situations or if they are more pervasive throughout a MUMS society.