National research evaluation systems that use metrics for the assessment of academic institutions are usually regarded as exemplifying the same neoliberal model of governance that, with minor variations, is implemented worldwide. This essay argues, however, that despite apparent similarities, metrics are used for different aims in different national cases. It compares the use of figures in the UK RAE/REF, a prototypical neoliberal framework, with various schemas of assessment that have been used to evaluate Russian universities in recent decades. It argues that in the RAE/REF, the principal role of statistics is to solve ‘the lazy agent’ problem by creating a prisoner's dilemma for academic institutions, while in the Russian case, statistics serve to solve ‘the corrupt knower’ problem, preventing collusion between the assessor and the assessed. The essay concludes by putting forward some hypotheses on the origins of different approaches to quantification.