Quote: Originally posted by apagogeas on May 18, 2004
--------------------------------------------------------------------------------
... no program can pick numbers by itself and win. This is because
prediction is a very subjective matter and requires some interaction
by the user. This is why lotto programs are considered tools and they
are not totally real prediction programs. As I said in another post,
a fully automated program is not a desired one because it will depend
solely on the programmer's intuition on prediction. ...
--------------------------------------------------------------------------------
Apagogeas
First, let me clear up a couple of points you made from the SUMS thread
which did not go unnoticed. You queried why anyone would want to use
Histories other than those generated by Lotto Operators. I would have
thought if there were no Histories or the Histories were for a short
period the answer would be obvious. The other reason is testing and it
is there you seemed to take great delight in joining with others to say
no PRNG would be suitable and there was talk about it not having enough
entropy.
Well let me quote from the Help file of Comscire, http://www.ComScire.com
the people that sell TRG's (True Random Number Generators ie the opposition
to PRNG's so we can't expect them to be too kind.
" ... Another generator that is believed to be random is the purely
mechanical generator, such as the type used for most lottery drawings.
Mechanical generators rely solely on complex initial conditions and
chaotic transformations to produce the appearance of randomness.
Precise quantification of bias in such a system is probably not possible
due to changes in the mechanical components over time, although there is
reason enough to believe that bias will exist.These drawing machines
produce almost no true entropy. ..."
I think that dismisses the entropy comments.
" ... The PCQNG has been tested extensively by well-known programs such
as Marsaglia's DIEHARD suite of tests and most of NIST's new tests.
ComScire's tests (ComScire RNGmeter), which are more stringent than
either of these, have also been applied to bit counts up to 10 billion.
The PCQNG has not failed any test for randomness. All other generators
we have tested fail ComScire's tests, with the exception of one type of
pseudorandom generator. ..."
Now I don't wish to suggest that my PRNG is the one they refer to but
merely to point out that they acknowledge that a PRNG can pass there
stringent tests. I would confidently say that my PRNG would pass there
tests as I apply many and varied tests myself. I would challenge you to
be able to distinguish between a History produced by my PRNG and that from
a mechanical generator.
This is a public Forum which means you can't expect to have your own
little private thread in which every other sentence extols some benefit
of your system and you luxuriate in a torrent of self-congratulation
for you own self gratification. I have left your thread alone but your
latest narcissistic posts have goaded me into thinking enough is enough.
I would be happy to pit my system against yours. In cricket it's referred
to as "putting runs on the board", in baseball I'm sure there's a similar expression and to tell you the truth I haven't a clue what the summer game is in Greece.
Have you ever tested your system to see whether it performs better than
Random Number Selections? Mine does - it's fully automated and there's not
a smidgen of intuition in it - creativity - yes, derived from number
crunching - yes. I think you would be well served to confine your remarks
to your own system and restrain yourself from making all encompassing remarks which imply you know everything about what everyone else has done or is doing.
Colin