Social Network Analysis is the study and analysis of groups as social entities to attempt to mathematically understand the interactions of people and groups in society. Computers have forms of social interactions; they connect to each other and have client/server and peer to peer relationships. Computers also have trust relationships. Grouping computers into social cliques or block structures is difficult because of the different roles of computers. This presentation examines a new method for testing blockmodel strength. The new method is called allegiance. Allegiance is a measure of how much an actor is helping his block. Our intent is to measure blockmodel strength change by examining actor allegiance as the partition size changes. Ultimately we would like to use our developed approach as a way to automatically calculate the partition size of the proper blockmodel. This serves as a way to cluster computers to make network analysis more feasible.