It is an error for lottery softwares not to have and make use of what I guess could be called combined filters or some other name(s).
When you have 2 choices it is either one or the other, few choices of filter patterns often lead to filter failure when using what I call "The Negative Prediction" technique.
Some examples of 2 choice Boxed or straight filters are: (Mostly already includes All)
High or Low (Mostly)
Even or Odd (Mostly)
In or Out (Mostly)
How can statistics tell you which of the 2 patterns will come out next and which one won't?
Your chances of being right and or wrong might be 50 percent either way, that is 50% chance of being wrong and or right, that is not good, but is bad.
If you have any 10 equal patterns and if you filter out only 1 of the 10, your chances of being right might be about 90%, but if you filter out 5 of them, you might lower your chances to maybe about 50%.
There should be a balance between amount of filtration and accuracy of the prediction, accuracy being the chance of the filtration not failing.
Too much filtration used on each filter lowers the accuracy of the prediction and makes for more filters failures.
But too little filtration leaves too many combinations that need to be bought and played and as we know, the lottery pays too little.
If the game is fair and honest, then we will only be dealing with what they might call "Random Factors", then more filters and patterns for each filter might be needed.
Having the wrong filters-patterns might be the same as having no filters-patterns as then they might be useless.
Winning by division (Filtering Out) and by numbers (Having many filters-Patterns) is what "Negative Prediction" is about.
We do see Hot Patterns here, but it is a very risky thing:
This listings format is uneven, the date should have 2 digits for the Day and also 2 for the Month, such as:
12/08/2009 instead of 12/8/2009.
The patterns themselves could just be represented by equal length symbols such as either digits or letters, but I know that that would confuse some or many people:
0 = Mostly Low
1 = Mostly High
2 = All Low
3 = All High
Then it would be easier to see and also count each of the 4 patterns there, but there could also be a choice of just:
Mostly Low and Mostly High, just 2 equal patterns, then:
0 = Mostly Low
1 = Mostly High
Each pattern then equals 1/2 of the total combinations.
But as I said, it might be best in some or many cases to instead use many patterns for each filter and if each of the patterns are more or less equal then even better yet.
Well, by combined filters what I have in mind is kind of like these examples:
Low-Even = LE
Low-Odd = LO
High-Even = HE
High-Odd = HO
Not exactly equal patterns, but close enough.
Those patterns in some way apply oe might apply to other numeric filters and or digits-numbers.
As to Hot-Cold filters-patterns, one should first setup the basic rules of their use, with them their being many more variables and also it might be best if the Hot-Cold counting is not done in the regular way, but in a weighted way, I also have posted about that before.
I might later today explain what I mean by that.