With the size of most airlines, you could count on some pilots being unwell, and come up with a model for it. The worst case would be that the occasional pilot

With the size of most airlines, you could count on some pilots being unwell, and come up with a model for it. The worst case would be that the occasional pilot would be being paid to do nothing because everyone turned up well and on time. This overbooking of pilots would probably cost less and have less disruption than breaking other peoples schedules. The model is one where Game Theory is suited to modelling the outcomes. The senior pilots don't want to lose their priveleges, the airlines don't want to lose their senior pilots. Yet the airlines also want to attract the best up-and-coming pilots, so having a predictable schedule would probably work as a recruiting draw.