People who make programs such as filter systems insist on making the filters of set values and that is why they fail more often than they should, they should let the past draws determine what values work best at a given time.
Filters programs should scan the past draws first and then based upon their finding, build-make their filters.
Filters should not go by names such as:
Low
High
Even
Odd
In
Out
Instead there should be kinds of filters:
The Above are 2 condition filters: 0 or 1 or A or B and 2 equal groups of 5 digits each.
They are 3 filters with 2 patterns for each filter.
There are 3 above, but they could be fewer or more depending upon the stats of the past draws, they could be:
Filter #1, 5 = 0 1 2 3 4 , 5 = 5 6 7 8 9
Filter #2, 5 = 0 2 4 6 8, 5 = 1 3 5 7 9
Etc.
------
What I am saying is that at a particular time for example Filter # 1 might or could work best as:
Filter #001. 5A = 05789, 5B = 12346 or any other combination and the same for other filters of this same kind, is all a matter of which combinations of digits might work best or seem to work best at a particular time or period of time.
-----------
The 50-50 % Mostly Filters same way, they would be based upon the above filters, for example:
Any "Full" pick 3 number that has 2 or 3 digits from #0015A(05789), would be "Mostly#0015A", if not then it would be "Mostly0015B"
-----------
Lowest-Middle-Highest kind of filter goes by the lowest to highest order of the digits, but it might happen that at a given time a different order might work best either for prediction or for filtration:
This order could change at any time according to what is working best such as:
Lowest > 3946518027 > Highest, Could be a new order if it works best at a given time or some other Lowest to Highest order of the digits.
------------
The same for all the other filters such as Low-Middle-High.
Low could be any 3 digits, Middle any 4 digits and High any other 3 digits. and or:
------------
Width kind of filters:
Width or Lenght in digits of the Lowest to Highest Digits.
This filter depends in the order of the digits from lowest to highest and that order could be any order of the digits that seems to be working best for this filter at any particular time, that order of the digits also apply to the Lowest to Middle and Middle to Highest Width filters.
--------------
Width Position filters those 3 filters would remain unchanged and as they are right now in the softwares:
Lth to Middle, Lth to Right and Middle to Rt Widths or:
1 Position to 2 Position, 1 Position to 3 Position and 2 Position to 3 Position Widhts.
-----------------
Well maybe you get what I mean (And maybe not).
About the same for any other possible filters including the VTracks.
--------
Filters such as Sums, LDRs and Roots stay the same as they are now, except for any added charateristics to them such as ""Even-Odd" Sums, LDRs and Roots", ""Low-High" Sums, LDRs and Roots"" Etc. Which depend on the digits that are assigned to such filters and which assignments should change as need as they should be dynamic.
------------
Anyhow just something to think about.
All of that is in accord to my study of the filters, their patterns and my attempts at prediction.
---------
I repeat, "Random is a dynamic process and so prediction also be"