6000 per grade level, so 78,000 K-12?
The Idea was 6000 per either grade level or roughly the same age, but yes you have the idea. I was a very simple program. I created 6000 random numbers, and counted how many passed trough a very small window. I only iterated 50 times so that it was easy to view all of the individual samples.
Usually with Monte Carlo you will be trying to solve problems that are not as easy to solve deterministically as this one, and you will run a very large number of iteration then apply traditional statistics to the results.
In this case I thought it would be interesting to approximate the variation you might see in 50 equal size cities, or in one city on 2 randomly picked years. Of course, in 2 successive years random would not make total sense because you would be resampling many of the same children, which should reduce the amount of variation observed.
