Recently, visualization of sport data in general, and soccer data in particular, has received much research attention. Visual sport data exploration helps to understand behavior and performance of athletes and teams, identify possible influence factors, and changes over time, among other important tasks. In soccer match data, much of the play is determined by direct interactions between players spatially close to each other, competing for influence. We introduce a novel visual analytics system for exploring pairwise player interactions using a trajectory-based data representation in a highly interactive multiple view approach. Our notion of player interaction is based on proximity of pairs of players, and respective motion patterns represented as trajectories. Our approach segments player interactions from soccer match data, as the basis for linked analytical views. A matrix view allows to explore interaction frequencies between players, group of player roles, and assess overall game dominance between teams. An appropriately defined interaction glyph allows to compare interactions based on player motion, ball possession, and pitch position. We further investigate the design of a descriptor encoding the geometric configuration of interaction trajectory pairs, enabling common analytical tasks like clustering or searching for similar interactions. We demonstrate the applicability of our approach by use cases on real soccer match data, detailing the analytical tasks supported by our system.