Paging and List Update under Bijective Analysis It has long been known that Least-Recently-Used (LRU) and some of its variants are the strategies for online paging that are favoured by practitioners. However, under competitive analysis, and indeed under several alternative performance measures, LRU is no better than many strategies that are clearly inferior in practice. In a similar vein, Move-To-Front (MTF) performs very well in list-update applications with high locality of reference, yet it cannot be separated from other algorithms under the cost model proposed independently by Martinez and Roura, and Munro. In this presentation I will describe a natural measure for comparing online algorithms termed bijective analysis, and summarize some recent separation results. This line of research aims to bridge (or at least narrow) the disconnect which appears occasionally between competitive analysis and the performance of online algorithms in "the real world". Joint work with Reza Dorrigiv, Alejandro Lopez-Ortiz and Pascal Schweitzer.