First Lottobuddy, Thanx for the link. It's a good one and I've bookmarked it.
Your suggestions are well taken. I generally do not attempt a level of refinement you indicate in your calculations, but, as I've said, rely on moving averages of sales figures from long experience with the matter as described recently in this thread. I know that (since the week contains an odd number of days) that it makes a difference whether a drawing is on a Saturday or a Wednesday. Other factors that effect sales are probably things like the size of other Jackpots in Powerball or MM states, weather conditions in states selling tickets, holidays, general economic conditions, etc, etc.
All of these factors invariably combine to result in some small percentage error in my numbers, but this is mathematic modelling, not firm high precision prediction. One assumption that is made in Poisson calculations is that the selection of numbers is truly random; i.e. everyone buys quick-picks. As we know, this is actually NOT the case: People play lucky numbers (often numbers in their birthdates) and people engage in wheeling. Both of these factors, wheeling and playing "lucky numbers," probably combine to increase the probability of a rollover by widening the number of unplayed possibilities with respect to the universe of possible combinations.
My own postings are "down and dirty" and are generated by a semi-permanent spreadsheet and posted quickly. Your's are undoubtably finer and more careful in the present case than mine. Remember though, that projections of sales, though often accurate, are subject to some error, no matter how much precision one attempts.
If I remember and have time, it is easy for me to post my sales assumptions inherent in my values. If at anytime you have time to offer refinements such as those you've posted here, please feel free to do so. Indeed, I welcome constructive input such as you have offered here. You have clearly understood how this works, and I have no doubt that on many occassions numbers you post will be closer to the mark than mine.
We are generally in the same area, certainly close enough to give a rough idea of what the overall likelihood for numbers of winners and/or rollovers are actually likely. That's my intent.
Good luck.