Last year I set a piece of coursework for my students that involved modelling fictional (but plausible) numbers of jugglers arriving each hour at an EJC. The report had to be written up in a form of explanations and recommendations to the organising team regarding appropriate allocation of registration desks (and preregistration desks) including modelling the best opening hours (minimising the number of hours open while still minimising the number of people who arrive outside of those opening hours). Of course, the new registration system meant the previously plausible values of numbers processed per hour on the first afternoon were completely irrelevant.
That's the nearest I've come recently with combining my work as an academic applied statistician and juggling I'm afraid.