This article explores the modelling of social problems from the point of view of experimental practices. As we observe an expansion of experimental spaces in the present, especially in the form of real-world labs (living labs), the question arises how knowledge for collective problem-solving is produced in the hybrid zone between different social fields and heterogeneous actors. The analytical lens applied here is knowledge modelling using indicators. The question is, how against the background of different qualities in political discourses (political ability, will and obligation) a balance in problem-solving can be found, how work by means of indicators can take hold, or whether the factual content of problem situations is distorted by indicator politics. These questions are pointedly developed using the example of real-world laboratories.