A complete understanding of male reproductive success, and thus sexual selection, often requires an insight into male success in sperm competition. Genuine conclusions on male sperm competitiveness can only be made in real competitive situations. However, statistical analyses of sperm competitiveness from fertilization success data have been shown to be problematic. Here, I first outline a comprehensive general description of the different additive and nonadditive elements relevant for the outcome of sperm competition staged between two males. Based on this description, I will highlight two main problems that are frequently encountered in experiments aiming at estimating sperm competitiveness. First, I focus on potential problems when using standardized competitors versus random mating trials, because trials with standardized competitors do not allow generalization if male–male interactions are important. Second, I illustrate the necessity to analyze data on the logit scale rather than on raw proportions, because only the logit scale allows a clean separation of additive and nonadditive effects (i.e., male × male and female × male interactions).