Using a pool a numbers from recent draws for the next one!

[thornc]

Hello all,
I've been looking at something lately, it all evolved from reading R.Perkis' documents.


I used WH and Saliu's MDIEditor to test it, it seems to work 1 in every 6 draws in the Can649, in PT loto it's a bit better but I won't get into it....

The idea is to use all the numbers from the past N draws, as a poll to play the next draw, with N=5 the pool of numbers will contain 5 winners once every six draws, approximately!
I checked the last 200 draws of 649 and 31 times this works!

So if one can find a good 20-30 wheel we're set!
Or we can find a way to get rid of some of the extra numbers and use a better wheel!

I the last PT loto I hit a 3 in 20 numbers, I played 21 because I needed an extra for the wheel!

Any comments...
Andrew I will post soon about using WH to reduce the selected NoN to a more playable size.

 

[Dennis Bassboss]

That makes sense to me...I just talked about that yesterday actually replying to Maggie in 6/49...

 

[Beaker]

I have used this polling of history for years - even before I read Perkis's paper - Ion Saliu also.

Lottoman uses this in its software along with other factors to maximize the prediction.

The trick is how many draws back.

This is very effective in Super7.

 

[Dennis Bassboss]

I have used this polling of history for years - even before I read Perkis's paper - Ion Saliu also.

Lottoman uses this in its software along with other factors to maximize the prediction.

The trick is how many draws back.

This is very effective in Super7.

Me too Beakers....way way back I was using it.... What is good about it is that it seems to follow a pattern at times...

 

[Beaker]

Me too Beakers....way way back I was using it.... What is good about it is that it seems to follow a pattern at times...

Set it up for S7 Buddy if you haven't already - you'll win. It seems that the lower the number of balls (47) the better this strategy becomes. thats why PT loto is so good.

 

[thornc]

Hehe,

I knew you guys would have noticed it before, but what I would like is to try to find a way to evolve the idea into some kind of strategy.....

The trick is how many draws back.

There is where WH comes in, using the optimer I noticed, that using 4/5 gets us the best poolsize/nohits...

What is good about it is that it seems to follow a pattern at times

Yap, on the Ca649 it seems to hit for a few draws and then skip a bit then hit again... on the PT it seems to be appearing with a trigger !!!

It seems that the lower the number of balls (47) the better this strategy becomes

?? The PT lotto is a 6/49 game also...equal nob as the Ca649

 

[Brad]

Strange but true ... I cut my teeth so to speak on a very similar strategy a number of years back, using BC49 of course.

Used N=4, with and without bonus, subbing for and without numbers that repeated within N=4, etc. Have results going back to draw one ... maybe it's time to dust off the old spread sheet and feed it into WH somehow

I like the direction of this thread, keep it coming thornc and all

 

[thornc]

with and without bonus, subbing for and without numbers that repeated within N=4, (...) feed it into WH somehow

I like the direction of this thread, keep it coming thornc and all

BTW, my results were without bonus...that lowers the poolsize!
In WH I used the singlehit processor and optimized on the filter!
I haven't got much else to say at the moment, I posted hopping that some of you folks would point something that would give me some new ideas....

 

[Brad]

I can provide more specifics if you're interested ... but later on as I'm behind on my chores now

 

[Beaker]

Sorry thornc, I thought PT loto was 6/45, my apologies

 

[Dennis Bassboss]

Sorry thornc, I thought PT loto was 6/45, my apologies

You are all excuse.... Beaker...

 

[Dennis Bassboss]

I can provide more specifics if you're interested ... but later on as I'm behind on my chores now

Same thing here but all my chores are in front of me...

 

[Brad]

Same thing here but all my chores are in front of me...

... I'd rather have a bottle in front of me
than a frontal lobotomy (an oldie but a good one)


Just a sample, looking at BC49 draws 867 to 870, predicting for draw 871:

867- 08 16 18 23 33 41 B09
868- 04 09 22 33 35 37 B05
869- 03 20 24 25 26 34 B21
870- 01 24 38 41 45 48 B31

871- 09 20 34 37 38 45 B33

In the above case all 6+B hits came from prev 3 draws, although at the time I was using prev 4 draws, removing reps (33,24,41) and not using bonus numbers ... so that meant 6+B/21.

The more 'successful' approach that year (2000) was selecting 24 numbers (prev 4 draws), then replacing reps with other numbers, mainly recent bonuses, to maintain the 24 count. This yielded a total of:
3/4/5/5+B/6 hits - 39/21/3/1/2=66/105 draws, ie: ~63% of the preds had at least a 3 hit.

Overall hit avg. from 1992 to 2001 (1036 draws) was:
3/4/5/5+B/6 -33.8/23.5/3.6/3.2/1.1 or 62.9% min 3/24 hits per year.

Not sure how that compares to random, never did work that out. Have the same conundrums as thornc ... reducing the count without losing good numbers, or finding a cost effective wheel ...

Other approaches I'd like to try, time permitting, is skipping the most recent draw then focusing on the next four or five, etc ... this is where WH comes in, wish I had it 4 yrs ago !!

 

[thornc]

Sorry thornc, I thought PT loto was 6/45, my apologies

No problem here beaker...

Brad, others
Interesting, I haven't looked at the BC649 yet, but you see to state that you have draws with the 6numbers from the previous 3 draws? Man in PT the best I could find for that was 6/7 draws....

My idea was to try reducing the poolsize with the most over repeated numbers in the draws and perhaps put in some cold
numbers instead... I've been trying to do this with WH without big results.
Getting rid of the over repeated numbers only reduces the poolsize by 2/3 at best, and it slightly lowers the good results....

Brad
Have you looked to see if you could find a trigger that predicts the results??

 

[Brad]

Thornc,

6 in 3 draws does not happen too often, that is an actual case from my data and I would say it is an exception. But 6 or even 6+B from 4 draws is not uncommon. For example, in 1999 it happened 4 times, but on avg once/year ... I have stats till ~mid 2002.

Re: Triggers ... no luck, but that doesn't mean there aren't any. Must not forget that I acquired the data 'the hard way', did not use software.

Maybe I should look at all this again with a pair of fresh eyes, possibly glean something I missed, but I have a feeling that in order to get better (more frequent) results, a few 'colder' numbers should be included in the preds.

 

[thornc]

Maybe I should look at all this again with a pair of fresh eyes, possibly glean something I missed, but I have a feeling that in order to get better (more frequent) results, a few 'colder' numbers should be included in the preds.

Yes, that's what I'm looking into now...trying to create a stack in WH to include the numbers from the previous draws, exclude 2/3 to hot numbers and include 1/2 cold numbers... haven't figured out a way to do the second include yet!

 

[thornc]

Just took a quick look at BC649, using MDIEditor... it seems to be a tough game!

Either way, the 5in6 is there around 30 times in the last 200 draws, 5 winners where in the previous 5/6 draws... no counting bonus!

Didn't test with WH to get a poolsize, but it should be around 25 numbers....


This last draw was amazing, just like this week's draws for PT lottos!

 

[Brad]

This last draw was amazing, just like this week's draws for PT lottos!

... ain't that the truth !

Your're right, BC49 can be tough but it's $0.50/ticket vs $1.0 for the CAN National 6/49 ... thus the attraction, ie: cheaper or bigger wheels can be used (on a small budget). One other diff is BC is computer generated whereas the National has balls.

Cheers

 

[Dennis Bassboss]

Originally posted by Dennis Bassboss in 6/49 discussion for april the first...
Nice observation Maggie....When they are hot ...They really are... Always keep in mind that most of the times many are coming up from the last 4 to 6 draws...When you added these you sometimes have less than (4x7 draws=28numbers for the last 4 draws) and less than (6x7=42 numbers for the last 6 draws) because of repeaters from draw to draw...Personnally I always look at that aspect when choosing my numbers... I once posted these for Super seven also somewhere on this board...I know a player that plays all the high from the previous draw and all the lows from previuos draw in 2 seperate wheels...to complete both wheels he looks at the last 4 draws and when this is done he also puts in two cold numbers to complete both sets...He often wind up winning up to 3 and 4 correct numbers...but he also nailed 5+bonus once 4 years ago... It is a nice way to start picking your numbers...

This was the post in question here...

 

[Dennis Bassboss]

Of course when you criss-cross these againts the double announcer grid and other stuff....You get a very nice look at things to come.... a funny one...

And this came right after...