In the video, it says that as n increases, the probability distribution becomes closer to a gaussian distrution and mean can be computed as np and variance can be computed as np(1-p). But as the graph shows, why this mean and variance computation also works when n = 2 which is not large enough?
Remember here n is representing a part of sample to the population distribution, so as n increase the graph distribution tends to looks concise because of it variance distribution to population distribution, leading the distribution to be gaussian(normal) with each increase in the sample size of a population.