Yes, Stan,
This is my first post of any kind anywhere. I am a retired computer systems technician from the auto industry. First of all ..I just wanted to say your software is one the more promising thought through serious approaches to this type of analysis I have seen. And I am a bit older so ...I have seen many over the years. I have a lot of ideas on different strategies you may want to follow up on or not. Now some of the filters I don't quite follow and some even though they look like a mathematician thought them out...at least on the mega millions and power ball some have little effect or weight in the analysis area. That is again I maybe using them wrong.
The new approaches I was thinking about were to use a data mining concept. Now in all fairness you seem to come close, but after analysis you should have a computer sorting and weighting of analysis characteristics and these characteristics could automatically be loaded and sorted with variables adjusted automatically to maximize yield and the result could be part of your filter builder concept. Now the yield could be tested by saving a unanalyzed section of the lotto for testing. The testing could be a feed forward and feedback testing. You could also incrementally feed forward in your analysis techniques for trends in each analysis characteristic or filter. I would appreciate any comments or feedback, thanks for your time and consideration