For multi-stage random experiments, probability trees are used for their description. Using the path rules, probabilities for the results of such multi-level experiments can be calculated from the probabilities that characterize the transitions from one step to the next. A natural next step is to look for sequences of random experiments, which can be represented by directed graphs. Analogous to the path rules, mean value rules serve to determine probabilities for an end of these random processes and their expected durations. Directed graphs and mean rules represent domain-specific strategies that take up and develop concepts of stochastic teaching at secondary level. Classification: D54, K64, N74, D29. Keywords: stochastics education, stochastically thinking, domain-specific strategy, MARKOV chains, directed graph, absorption probability, mean hitting time, transition probabilities, “Mittelwertsregeln”.