How do you know the approximate number of red and green intervals that will be graphed in the middle graphs?

We are using this website: http://www.rossmanchance.com/ISIapplets.html

In what way(s) does the confidence level affect (change) the width of a confidence interval?
In what way(s) does the sample size affect (change) the width of a confidence interval?
How do you know the approximate number of red and green intervals that will be graphed in the middle graphs? Is the percentage always correct? Explain.
Describe at least two ways to make a confidence interval wider.
Explain why “there a 95% chance that this interval (.46, .52) will contain the population proportion” isn’t a correct interpretation of a confidence interval.