Many of the suggestions you mention I have used in the past, Having these options did not improve my
play and often ended in confusion. You have to move beyond this type of analysis. After the draw it is so
easy to make connections to the history no matter which set was drawn.
To prove this update DM with any set you can think of and then see how many bits of information that leads
to that set being drawn even though it is just a made up set. If you add fake draws to the database don't
forget to remove them afterward using the remove button.
I use to include a master list which contained the stats for every filter based on the entire universe of sets.
From this list I determined how often each value was expected given the drawing is random. I don't know how
many took advantage of these list but I learnt them inside and out. This is the foundation for all my selections
as it gives me an idea of how many past draws I need to consider for each value. If a certain value shows in
30 percent of the sets in the matrix then I can expect 30 percent to show in the drawings. If the value is running
at 20% for the last 20 draws then I would expect it to show very soon as a random sample should. This is not an
exact science but given that you can range the values you can set more that one value to hit. When I make my
decision I also consider a few other factors.
Lets say that I am setting the odd number filter. 5-39 matrix
I know that 11628 sets have (0) odd numbers and 575757 / 11628 =49.5
I know that 77520 sets have (1) odd numbers and 575757 / 77520 = 7.4
I know that 184110 sets have (2) odd numbers and 575757 / 184110 =3.1
I know that 194940 sets have (3) odd numbers and 575757 / 194940 =3.1
I know that 92055 sets have (4) odd numbers and 575757 / 92055 =6.3
I know that 15504 sets have (5) odd numbers and 575757 / 15504 =37.1
Notice that sets with 2 and 3 odd account for 66% of all sets in the matrix and they account for 67% of the draws
for my 5-39. This is similar to the Law of large numbers but random sampling can be used with a much smaller
sample providing the sample is close to random which fits the lottery very well.
If I am looking at the odd number filter I look at the last 20 draws, I expect to see around 13 draws that had
2 or 3 odd numbers. If the total is greater then I might look at sets that had 3 and 4 odd numbers and then look
at the number of sets that had 1 abd 2 odd numbers. I play the range that best describes what needs to hit to
maintain the stats for the overall matrix. This does not always work out but it does very well much of the time.
Remember that you don't need to be 100 percent correct to hit some nice prizes.
Let's say you set the odd filter to 2/3 for a number of reasons and I would like you to consider this. Lets say that
the next draw ends with 1 odd number, since you played 2/3 you are still in the game for a 4of5 and the same
could be said if the draw has 4 odd numbers. It's a little like playing chess and your strategy should include the
what if question. Some choices are pawns and we use these to reach our play limit. Don't get bogged down in
having to make the correct choice every time. Have some fun, use your witts and see what you can do. Think
of DM as a strategy game where you are trying to catch your opponent in a small group and take his crown. It
can make playing much more fun, you don't have to buy any tickets to play.