Ok, now that the little 113 set side-tract is dust I will get back to what I was working on before.
The hybrid-3 is a spin-off of the of the daily game version of my rogue program. Maybe others
will be interested and work with the idea. Since the rogue values are ranged 0 to 3 they can be
converted into a 2-bit binary code.
Example.
digit 0 = 00 binary
digit 1 = 01 binary
digit 2 = 10 binary
digit 3 = 11 binary.
The idea is to analyze the binary using two separate analysis indicated by the arrows. Why do
it this way, lets consider that the analysis is limited to 0 or 1, 50/50. In the daily game version
of the rogue it takes 5 rogue values to reduce to a single line. If I insert 2 values for each rogue
position then it will return 32 lines which if correct produces a straight hit.
The advantage of doing it this way is that while it's not always possible to correctly choose both binary bits,
setting either to a 1 or a 0 will reduce the field of rogue value choices to two. If the top string is set to (0)
then the rogue value will be 0 or 1. This is useful when one but not both analysis can be set with any certainty.
Think of it as making a half choice. Selecting a value for either string reduces the field of choices to 2 vs 4 which
is what's needed to reduce to 32 lines.
The below picture shows the game history, first the actual set then the 5-digit rogue values. next is the 2- bit
binary for each rogue position and finally the number of 1's within the binaries for all 5 positions.
It's this value that was first thought would make a good filter as setting it to a single value reduced
the 32 lines down to 8. The below pic shows the total hits for each value.
In test, some values reduced the 32 sets down to 2 but the question remains as to how predictable
these values are. In my earlier post of the 113 lines I generated the list using the entire 1000 lines in
the matrix setting the 1's range to a single value. I started checking and found 2 or 3 straight hits within
the last 10 draws which lead to my earlier posted blunder.
What I am looking to do is create a better filtering method for the hybrid-3 program which would allow
the user to play more than one h-code. I normally can pin the h-code to 3 choices but without a better
filtering method this generates 48 lines.
Anyway, I thought I would post this in case anyone with the means wants to explore it while it's being
developed. My time is limited that I have to work on it but wanted to let everyone know that it's still
being developed. Not sure when it will be finished but wanted to touch base with hybrid-3 users.
RL