Gwoof
I don't do test like you might expect, what I do is crunch data using many different tools and
record the results. I know that if the predicted values are correct then the winning set will be
in the lines generated.
As far as doing a user setup as in the pics, I only do that when I want to post a picture. My test
methods are very boring but I can run thousands of test in the few seconds it takes to do a actual
setup. All the stuff in the program is first coded as a test program. Options are first tested for
accuracy then tested to see the reduction power. Next it's tested to see how well it works when
combined with stuff that has already gone through the process. Some stuff works very well as a
single option but has little or even a negative effect when combined with other options. Once all
these test are complete I then have to test it's predictability. Once this stage is finished I have a
very close estimate as to how often it will win in the hands of the general public given they follow
the rules.
So to answer your question, prediction methods are being tested. Some have lost more than they
should, some break even and some win way more than expected. The problem is that none are
consistent enough to keep the average user interested. I have a number of tools that would make
a few thousand profit over six months but most players don't have patience enough to stick with
them. As I have said several times, if the MS and FS slides can't be predicted then the program is
for naught. On the up side I can test new ideas as they pop up so there is always hope. I think I
can make a nice profit now by timing my play and waiting for certain key events to fall into place.
RL