In telecommunication research, the assessment of strategic and operational performance has gained a lot of interest. So far, the variability within the operational research of telecommunication derives from the stochastic nature of call arrivals, call duration and abandonments. This study extends the stochastic drivers to capacity by incorporating agent behavior. In this study, we model a Discrete-Event-Simulation model of the Swedish emergency call service provider to assist the strategic issue of skills-based routing. Because of the pull system, we designed skills-based routing using an event-dependent overflow setting with a fixed waiting time threshold, where calls become visible to extended agent classes after a threshold value, providing an advantage to primary idle agents to answer before idle secondary agents. Our model mimics the tail of waiting times better than conventional methods. Because of the accompanied calculation opportunities, the model has assisted both strategic and operational issues in the organization. In addition to practical implications, the study proves that the stochastic nature of agent behavior is crucial. From a queueing theory perspective, the study provides interesting routing effects for an event-dependent overflow setting using a fixed waiting time threshold in a combined “X-N-design”, previously unexplored in research.