For the last two decades there has been an increasing debate in both the research and the professional literatures of physical education pedagogy focusing on how to teach games in physical education. A number of curriculums have been proposed including the Teaching Games for Understanding model or TGfU (Bunker &Thorpe, 1982), variations such as game sense (den Duyn, 1997), the Tactical Games Approach (Griffin, Mitchell, & Oslin, 1997), and most recently Play Practice (Launder, 2001). However, the research supporting these models has been plagued with findings that are often statistically non-significant, making interpretation of the effects of these models at best equivocal. There are two questions that have not been addressed in the extant research. The first is "How well do these tactical approaches teach students to perform tactics?" This is a different question than comparing tactics instruction to another form of instruction. The second question is "Can students apply these tactics in games?" The first question is related to instructional effectiveness in teaching competent performance of tactics, while the latter question is a matter of generalization. However, both questions are related. In this presentation the issues surrounding these two questions were be explored. In particular, I will argue that curriculums such as TGfU and Play practice have fundamentally different curriculum outcomes, and use different pedagogical strategies in achieving them, which researchers must attend to in the conceptualization of their studies. Second, with regard to the measurement of actual performance by students the selection of units of measurement in these studies must move from the general to the specific. Generic coding systems are neither contextually specific nor easily instructionally aligned with the pedagogies present in the models.Keyword(s): high school issues, middle school issues