The topic of graphs is a commonly used informatics concept that contributes to the development of computational thinking skills. As research shows, it can be used effectively, for example, in STEM education at various school levels. However, the form of incorporating this topic into teaching is still the subject of research. Tasks related to this topic commonly appear in the International Bebras Challenge task database. The contestants’ results in these tasks show that well-formulated tasks can be reasonably demanding for middle school students. Therefore, our research focused on investigating possible implementations of graph tasks in lessons. Three tasks from the Bebras challenge were selected, and a worksheet was created for each of them. The worksheets were then verified with a smaller group of students in class. These worksheets consisted of so-called microworlds, each focusing on a different graph representation. Our research problem was to determine how to appropriately grade the difficulty of the tasks in the worksheets so that the topics are motivating for the students and the tasks are not overly cognitively demanding. To address this, qualitative methods of data collection and analysis were used, working with over 100 students aged 10–14 at three different schools. The main goal was to better understand the cognitive demands of graph tasks for middle school students and to formulate specific recommendations for practice. One of our results indicates that the connection of motivation with real-life experiences has a significant impact on students’ results and understanding.