Dynamic network visualization is a rapidly evolving field and computing a node-link layout for these graphs is one of the most studied problems in this discipline. Benchmark datasets and quality criteria are lacking in order to enable comparative evaluations of the layouts produced by different dynamic network visualization techniques. This paper proposes a collection of both discrete time and event-based dynamic quality criteria for evaluating dynamic graph layouts. Furthermore, we present and discuss datasets, generated from real data, which we include in a benchmark set for future researchers to use in their evaluations.