The generation and delivery of highly complex team and individual scenario-based training is bounded by a variety of requirements. This training method must accurately replicate conditions that include a realistic environment, rapidly changing events, multiple information sources, rules and procedures, and time and command-induced performance pressure. Past research indicates that training scenarios should be tailored to incorporate events to create valid learning opportunities, stimulate desired performance, and provide team stress management insights. However, the complexity of tailoring this "curriculum" presents a daunting task for novice and expert shipboard trainers. Therefore, a strategy to intelligently manage the process of event selection was determined to be important. Following our analysis of the state-of-the-art in scenario-based training systems, it was evident that, during scenario development, the event selection or creation process was left to the user. This places a nearly impossible demand on novice trainers to ensure that scenarios and supporting products are related to the mission, training objectives, and past performance. Indeed, while this task is within the capabilities of expert trainers, it is very time consuming, and not often done. This paper outlines efforts to create a formalized, user-centered architecture for assisting trainers in the selection of scenario events using performance history data, mission criteria, trainee identification, and other factors. To enable this, an event library had to be created that would "understand" training objectives, complexity, and inter-event relationships. Success will provide the capability for novice and expert trainers to harness the power of scenario-based training.