Sampling theory applies to the number of simulations nrep just as much as the sample size n within a simulation.
mu and sd from 24 sims N(0,1) with n=10
Now, let's do the same with a number of repeats nrep of 1000.
mu and sd from 1000 sims N(0,1) with n=10
The number of simulations needs to be a big enough number to obtain a good representation of the distribution of the simulation results, e.g. 1000.
YOUR TURN:
You can play with these graphs by running and modifying the examples provided in your local exercise script (e.g. changing parameters in turn)