Estimating discrete games of complete information is often computationally difficult due to partial identification and the absence of closed-form moment characterizations. This paper proposes computationally tractable approaches to estimation and inference that remove the computational burden associated with equilibria enumeration, numerical simulation, and grid search. Separately for unordered and ordered-actions games, I construct an identified set characterized by a finite set of generalized likelihood-based conditional moment inequalities that are convex in (a subvector of) structural model parameters under the standard logit assumption on unobservables. I use simulation and empirical examples to show that the proposed approaches generate informative identified sets and can be several orders of magnitude faster than existing estimation methods.