Computational fluency is composed of three components: efficiency, mathematical flexibility, and conceptual understanding. Although research highlights the importance of all three of these components in mathematical understanding, a majority of assessments primarily focus on efficiency. In this paper, we propose a novel way of measuring and exploring student's mathematical flexibility using dynamic technology. We introduce Graspable Math, a dynamic mathematics notation tool, and present data from a goal-based puzzle task given to experts in mathematics. We then use the GM tools to present evidence demonstrating significant variability in problem solving approaches, efficiency, flexibility, and thinking among participants considered to be experts of mathematics.