How can you account for the fact that different infected people might transmit the disease to the same susceptible individual within the same time-step? Early on in an outbreak this would not be that much of a factor, but if the sizes of the susceptible and infected groups are similar later on, would there not be some overlap between transmissions? I don't think that Beta*I*S accounts for this, and I don't want to overestimate the number of infected individuals in my model.