August 15 Piks

Sheba

Member

17-19-20-25-26-27-30-31-33-34-40-43-45-47


Sheba__:bouncy: :dog:~~~ ™
---------------------------------------------------------------
 

daleks

Member
got in late...

but it's still only 7 p.m. here on the west coast, and i didn't bring my numbers with me, but, played 30 lines, concentrating on the 42 and 47, and picked the rest of the numbers randomly from those numbers which had been drawn 2,3 and 4 times during previous 18 draws................:beer: .......corona..............
 
Nick Koutras said:
LottoSelector v3.2
picks
Using new method
called Bayesian Selection

Tech Papers here:
http://www.stat.washington.edu/raftery/Research/Bayes/bayes_papers.html


7,8,11,12,13,15,16,23,27,29,32,40,44,46
I am aware about that method or strategy in this case....And it worked well tonite....This ''Bayesian method'' should be good in about 1 out of every 5 draws....But when its good it could be dam good...
4+bonus tonite...
:agree2: :agree:
 
Dennis Bassboss said:
I am aware about that method or strategy in this case....And it worked well tonite....This ''Bayesian method'' should be good in about 1 out of every 5 draws....But when its good it could be dam good...
4+bonus tonite...
:agree2: :agree:

Hi Dennis,

I noticed the success of the method. So I added one more function to it.
"Feedback"
What this function does is:

After a selection of a number, it looks
at the history with the assumption that
this Number Will NOT come.

From the distribution, with that assumption, verifies that the Other selected numbers do belong in the Tails
of that distribution.
If so this number is given a lesser weight.
The process continues until the selected number quantity is satisfied.

I will use the same process for tomorrows 649.

The main difficulty I have with the selection methods I use is that they do not coincide on their selections.
If you have some ideas for prediction I'll be glad to offer my services to code them.
By doing that we can examine the validity of the method.

Nick
 
Top