A pseudo-random seed is the initial state that fixes the sequence of numbers a random number generator will produce. Setting the seed makes simulation research reproducible. Changing the seed changes the sample path draws but not the underlying model. Forgetting to control seeds can make two “identical” Monte Carlo runs hard to compare.