Interactive visualizations are key in communicating process mining results. For process discovery, the most important visualizations are directly-follows graphs (DFGs), which can quickly become complicated. To handle this, tools provide activity and path sliders for users to interact with the DFG and filter displayed information. However, these sliders currently make the DFGs instable: Their layout changes for each slider setting, making it difficult to identify relevant information. Dynamic graph visualization research has proposed to address this problem by stabilizing graphs, preserving the so-called ‘mental map’. In this paper, we apply this concept to DFGs and conduct an empirical study to analyze whether the mental map improves process analysis performance at different expertise levels. In line with results from other domains, we find that the mental map has no positive effect on process analysis performance; we rather find a significant negative impact. We also find that experts have lower response times than the other groups, which suggests that they identify relevant information more quickly, although not always accurately. These results provide empirical insights into the visualization of DFGs and offer potentials for future research and practice.