Hi everybody,
With precious help from my friends I have developed a method that improves the odds of winning the Lotto 6/49 Canada by up to 20 times (for a 6-numbers set played).
The idea is to choose the 6-set played from a pool that is about 20 times smaller than the total pool of ~14 million possible 6-set. So instead of playing randomly with 1 in 14 million odds you can play not-so-randomly for 1 in 700 000 from my prediction list. If you play 10 sets from our list you get 1 in 70 000 odds of winning.
I call our list a prediction list because it really predicts a lot of winning sets of all types. So far these are the results of our prediction lists since June 2008:
6: (one set), for one draw out of 6 ... 10 draws
5+Bonus: (one set...a few sets) for one draw out of 8 ... 10 draws
5: (a few sets ... a few tens of sets), for 5-7 draws out of 10 draws
4: (hundreds of sets ... ~2000 sets) every draw
3: (thousands of sets ... ~23000 sets) every draw
2+Bonus: (thousands of sets ... ~15000 sets) almost every draw
A list of about 700 000 sets is about 20 MBytes in a text file and even archived in a zip file it's still ~ 4 MB. I've put all the zip files on a website and immediately after every draw I calculate and I publish the results in a checklist, as well as the current prediction list for the next draw. Everyone who plays Lotto 6/49 Canada is welcome to use them so feel free to ask me privately about the address (as I see the admins do not allow us to write URLs in public messages here).
Now, the reason for presenting this kind of prediction list here is that I am looking for methods of refining or using the list more efficiently. So instead of picking randomly 100 plays out of those 700 000, I want to pick 100 that don't overlap each other too much. A fragment of a such list looks like this:
2,3,5,11,20,49
2,3,5,11,21,40
2,3,5,11,22,25
2,3,5,11,22,30
2,3,5,11,24,45
2,3,5,11,26,35
2,3,5,11,30,38
2,3,5,11,32,40
The problem is that each new list is different from the previous ones, and its sets are not always in a continuous sequence. Obviously I wouldn't pick 100 successive sets, but if I skip randomly hundreds of sets I always get a set that overlaps one of the already picked sets. Example: if I pick 2,3,5,11,32,40 then one of the next picks might be 5,11,27,32,35,40 so I don't notice easily that I got 4 numbers overlapped (5,11,32,40).
That's a real waste of chances, but on the other hand verifying even 20 sets to prevent it always gives me headaches. Any ideas/methods for making this automatically?
Norbert