Stochasticity is a measure of the randomness of a process. A highly stochastic process (often shortened to a ‘stochastic process’) – is one that features variables or parameters with different probability distributions. Because of this multiple iterations of the process can return different results.
Within the field of simulation, It’s important to differentiate stochastic methods for problem solving and stochasticity for problem representation. Stochastic methods use probabilistic techniques to solve problems – for instance genetic algorithms evolve programs by randomly iterating across the space of potential programs, and selecting and combining those that best solve a problem. Stochasticity can also be used to represent a problem or environment, for instance a self-driving car simulation where, with some probability, a pedestrian might cross the street and the car may or may not see them.
In HASH stochastic simulations can be created by using the HASH standard library statistical distributions. Conversely, you can create deterministic simulations by ensuring the stochastic elements of a simulation use the same starting seed, thus returning the same results every run.