Applications that use games to harness human intelligence are increasing in popularity and such games are also known as Games With A Purpose (GWAP). Most GWAPs are collaborative in nature, requiring pairs of players to cooperate in a game to receive points. Competitive GWAPs, where players work against each other, are a more recent entrant, and have been argued to address shortcomings of collaborative GWAPs. In this paper, we focus on image tagging GWAPs where users play games to generate tags for images. Three versions were created: collaborative GWAP, competitive GWAP and a control application for manual tagging. The applications were evaluated to uncover users' preferences for these genres as well as the usefulness of the tags generated. Results suggest that the competitive GWAP outperformed the other two applications. Implications of the work are discussed.