Background. Player-generated reviews of video games represent a large, rich, and under-explored source of data for exploring what makes for an effective game. Aim. We explore whether existing theory, in the form of a comprehensive gaming taxonomy, suitably captures the issues that players raise when they review games. Method. User-submitted game reviews were coded along the dimensions of the comprehensive gaming taxonomy to test the frequency of usage of each dimension. Results. We found some support for the use of the taxonomy, as four of nine taxonomy features were frequently present in game reviews. We also found support for other features of reviews not encapsulated by the Bedwell et al. (2012) taxonomy. Specifically, we found that players often reviewed video games: a) holistically; b) by comparing them to other games, game franchises, and other reviews; and c) by judging the value of games in terms of time, money, and effort. These results have implications for using game reviews for future research.