This chapter ventures into the f ield of network algorithms, using the prehistory of Google’s PageRank algorithm to discuss yet another way to think about information ordering. The chapter shows how algorithmic ordering techniques exploit and integrate knowledge from areas other than information retrieval – in particular the social sciences and citation analysis – and demonstrates how the ‘politics’ of an algorithm can depend on small variations that lead to radically different outcomes. The context of web search means that the various techniques covered in the second part of the book are brought together into a shared application space, allowing for a more concrete return to earlier discussions of variation and combination in software.