Very Nice:
When teaching confidence intervals, this gets your point across. Use Football in the Fall and basketball in the spring:
Friday: IBM projects that The University of Florida has a 20% chance of winning.
Monday: The result is in and Florida lost. There is nothing random now. But you can say my procedure had a 95% chance of trapping the true value.
After the experiment, our confidence interval t either contains the parameter or not. We do not know which but the truth now k=knows with certainty.
Simulations can also help.
Best,
Jon
------------------------------
Jonathan Shuster
------------------------------