nick
It's true that the top 3 hitting values do not show together very often or almost never would describe it better.
The same thing applied to the pool member selections but I noticed that the members were doing better and
better and this caused a problem because the random-bias was not there. This left me with no real data to
work with. Using the random generated sets works better for this method. Think about it this way. Let's say
that I know that the top 3 won't all hit together and that several sets of groups will have these 3. I then look
at the remaining 3 groups in each set. Through careful analysis of these other groups I can gain a very good idea
of which will most likely hit. Most of the random sets of groups will only match 2 or 3 depending on if the draw is a
TG=4 or TG=5. If once I get lets say the best 4 I then do the process again using those 4 groups I can estimate
very well how it will do. If too many of the 15 groups contain 3 or 4 then I need to make another run using one or
two of the other groups. The process is not that hard until you start to code it. Sometimes what is simple to see with
our eyes is very hard convert into code. One of the lowest showing will also hit and is a product of the random
selections, The more a group shows the less likely it will hit. One of the groups will come from the top-3 and one
from very near the bottom. All we really need to do is look for the set of groups that has both of these and work
from there. As far as changing games to play the groups are what throws me, the rest is easy to adjust to. When
I get this program finished it will allow me to play about any game I want to play without interfering with my setups.
I might even be able to end with only 5 total groups which will be even better. I think you understand what I am
doing even though the earlier explanation was confusing at best. Thanks for taking the time to read and study it.
RL