Gentleman ...
I have been testing the WNH example in the documentation by using my lottery for the test. It works. Most of the time, I am left with about 200 tickets. In some cases, I have about 1500. I still think this is great from a set of 3 million.
I have tried a new test that used the predictions for the WNH summary table for the SUM filter. I noticed something interesting. I thought that I would see more tickets getting generated. However, I did not find any matches. Keep in mind, I have only done one test so far.
I do realize that you are given a range to use for the SUMS +-10. What I was really wondering is what is the expected 2nd and 3rd prize distribution likely to be like if you are not within your margin of error on a few different SUMS.
Also when I added secondary filters to my perfect set of 300 to 1500, I was able to reduce my set down to a reasonable play size. BUT getting those 11 Sum values are critical.
Any thoughts ...