Data sprints have emerged as a popular way to involve stakeholders in datawork. In this chapter we discuss what it takes to turn a sprint into a productive situation of inquiry (in the sense of Dewey, 1938). We argue that sprint organizers must work actively to counteract an otherwise docile setting where the preference for agreement between datasets and the expectations of participants is a constant risk. This produces what we call the "sprinter's dilemma": a situation where sprint participants must at once be excited enough about the potentials of new methods to meet and collaborate, and at the same time be purposely frustrated and willing to accept a reversal of the direction of inquiry. Whereas sprints are initially justified by a belief in the ability of data to solve certain pre-defined problems, they are eventually at their most productive when the nature of these problems becomes the subject of collective scrutiny through engagements with data. [For the complete volume, "Designs for Experimentation and Inquiry: Approaching Learning and Knowing in Digital Transformation. New Perspectives on Learning and Instruction," see ED661573.]