A compound event involves more than one component, such as two coin tosses or two dice. A simulation must reproduce the rules of the actual chance process: possible outcomes, probabilities, replacement/dependence assumptions and the number of stages.
Experimental probability is measured by relative frequency = favourable trials ÷ total trials. It varies from run to run. With more trials, results often settle closer to the theoretical probability, but there is no requirement that a finite simulation equal the theoretical value exactly.
Digital tools—including random-number generators and, where appropriate, generative AI used to help construct or run a simulation—must be checked. The mathematics is not trustworthy merely because software produced it. Verify outcome coding, randomisation rules and what counts as a favourable event.
A useful Year 8 investigation is rolling two dice and studying the absolute difference between them. Some differences occur more often because different numbers of ordered pairs produce them. Another is testing complementary events in a long-run simulation and checking whether their relative frequencies add to approximately 1.