BushHappy number elimination method

W

winhunter

Member
Downloaded

Hi,

I downloaded your method and am going to have a look at it. Your explanation didn't make much sense, so maybe by analyzing how it calculated a single digit (ball #1), I can determine how to setup a scoring method in WINHunter.

Based on what you said though, I am very confident that this method is a good candiate for inclusion into WINHunter.





Andrew
 
D

Dennis Bassboss

Member
Re: Downloaded

winhunter said:
Hi,

I downloaded your method and am going to have a look at it. Your explanation didn't make much sense, so maybe by analyzing how it calculated a single digit (ball #1), I can determine how to setup a scoring method in WINHunter.

Based on what you said though, I am very confident that this method is a good candiate for inclusion into WINHunter.





Andrew
Click to expand...
If you put it up ..I will look at it for sure...Your device is too good to overlook any new toys in it... :agree: :agree2: :agree2:
 
W

winhunter

Member
385 28/08/2004 Sat 2 7 16 17 37 48
386 01/09/2004 Wed 1 8 26 28 35 46
387 04/09/2004 Sat 2 18 27 32 40 45
388 08/09/2004 Wed 9 12 14 15 43 46
389 11/09/2004 Sat 4 7 13 17 36 46
390 15/09/2004 Wed 7 24 30 32 38 46
391 18/09/2004 Sat 4 24 29 38 45 47
392 22/09/2004 Wed 28 30 33 35 39 44
393 25/09/2004 Sat 9 14 27 41 45 46
394 29/09/2004 Wed 5 7 9 11 14 27
395 02/10/2004 Sat 15 16 29 37 42 48
396 06/10/2004 Wed 2 5 16 19 35 42
397 09/10/2004 Sat 5 6 21 27 32 42
398 13/10/2004 Wed 8 14 19 38 40 42
399 16/10/2004 Sat 7 11 15 20 21 23
400 20/10/2004 Wed 3 11 22 24 41 47
401 23/10/2004 Sat 15 25 29 30 43 48
402 27/10/2004 Wed 3 7 20 35 39 44
403 30/10/2004 Sat 9 13 20 26 38 43
404 03/11/2004 Wed 3 7 8 15 40 49
405 06/11/2004 Sat 9 11 15 21 36 47

For the prediction of the numbers that will not appear in in the next Draw - ie 406, each of the 49
numbers will be tested as if it appears in Draw 406 for each of the Draw categories.
From these tests, the number of occurrences of a particular number in NN Draws for Draw category
NN will be determined. If this value is equal or higher to the pre-determined criteria for Draw
category NN, then the prediction is that that particular number will not appear in Draw 406
Click to expand...


Using from draw 405 to draw 385, let's test ball number 9.
# of occurences of ball-9 in 10 previous draws (405 to 396)
would yield 3. right?

I need to know how to score the balls, and what level to set the "Criteria for Draw Category NN".

If you are only counting occurences of balls in the last NN draws, then you can use the SingleHit Processor and set the weight for the :Hit Count Weight" at the bottom of the Processor settings list for that processor. This weight simply adds the weight value to the score for that ball number for each occurence of that ball within the draw range set by the parent filter. In the filter settings, set Drawings to Bypass = 0 and Drawings to Use to the number of previous draws used to calculate that level of criteria.

Then, in the selector method, set the default selection type to Incusion and the minimum to select to the Criteria level.

If I am understanding what yo uare calculating, this will generate the same results for your tests.
 
D

Dennis Bassboss

Member
winhunter said:
Using from draw 405 to draw 385, let's test ball number 9.
# of occurences of ball-9 in 10 previous draws (405 to 396)
would yield 3. right?

I need to know how to score the balls, and what level to set the "Criteria for Draw Category NN".

If you are only counting occurences of balls in the last NN draws, then you can use the SingleHit Processor and set the weight for the :Hit Count Weight" at the bottom of the Processor settings list for that processor. This weight simply adds the weight value to the score for that ball number for each occurence of that ball within the draw range set by the parent filter. In the filter settings, set Drawings to Bypass = 0 and Drawings to Use to the number of previous draws used to calculate that level of criteria.

Then, in the selector method, set the default selection type to Incusion and the minimum to select to the Criteria level.

If I am understanding what yo uare calculating, this will generate the same results for your tests.
Click to expand...
Don't know if I follow here..does the parent filter becomes the default filter....giving the weight of the ball.. What would happen if we set the Bypass at more or less than 0 here...??
 
W

winhunter

Member
Bypass

Bypass sets the starting point in the past history as to where to start couting hits. Bypass of 0 starts at the last (most recent) draw. Bypass of 1 skips the lat draw and starts on the next one after that, etc.
 
D

Dennis Bassboss

Member
Re: Bypass

winhunter said:
Bypass sets the starting point in the past history as to where to start couting hits. Bypass of 0 starts at the last (most recent) draw. Bypass of 1 skips the lat draw and starts on the next one after that, etc.
Click to expand...
Than the Bypass is used for shrinking the history to time frame isnt'it?
I'm already doing this by cutting the history file into whatever numbers of draws i want to use...but it seems now that it can be done within a single history file ...Great news for me! :agree: :agree2:
 
T

tomtom

Member
Re: Downloaded

winhunter said:

.....
this method is a good candiate for inclusion into WINHunter.

....



Andrew
Click to expand...

I'm not sure if you know most of the people have difficulties inputing the data(draws) , and many gave up using the program because of that type of difficulty...Is it possible to fix that somehow?
 
W

winhunter

Member
If I understand correctly...

It seems that BushHappy will want to use multiple filters in a single group.

Each filter will have a single processor, and each filter will have an increasing number of "Draws to use" value.

Since WINHunter isnt dividing anything here, we can use a weight of 1 for each Single Hit processor as explained earlier.

Just be sure to set the minimum to select value in the selector accordingly based upon the number of balls you expect to eliminate.

His spreadsheets still use raw integers to count the balls, and thus some balls have higher hit counts. If we are looking to eliminate those high hit count balls, then we can use the Hit Count method of the Single Hit Processor.

But until BushHappy can explain exactly how his method works using the data, I'm still scratching my head looking at spread sheets....



Andrew
 
D

Dennis Bassboss

Member
Re: Re: Downloaded

tomtom said:
I'm not sure if you know most of the people have difficulties inputing the data(draws) I believe, and many gave up using the program because of that type of difficulty...Is it possible to fix that somehow?
Click to expand...
I don't think so...when they read the Winhunter threads they will find a lot of examples there... :agree:
 
T

tomtom

Member
Re: Re: Re: Downloaded

Dennis Bassboss said:
I don't think so...when they read the Winhunter threads they will find a lot of examples there... :agree:
Click to expand...

Sure... Andrew, can you fix the problems with loading the history? If you have some spare time of course..
 
W

winhunter

Member
Re: Re: Bypass

Dennis Bassboss said:
Than the Bypass is used for shrinking the history to time frame isnt'it?
I'm already doing this by cutting the history file into whatever numbers of draws i want to use...but it seems now that it can be done within a single history file ...Great news for me! :agree: :agree2:
Click to expand...

Yes, these two settings allow you to set the range of history that filter should target. Then, when you backtest the stack, you can see how well that setting worked against the entire history.

Think of it as a sliding slot. The bypass sets the starting position of the slot, and the draws to use sets the width of the slot.

There is also a cyclic history function within the filter, which allows you to set when the filter is on, and when the filter is off. You could create two filters, each working inverted of the other, and both of them with different processors running beneath them. But perhaps this discussion should move to another thread in the WINHunter section....


Andrew
 
W

winhunter

Member
Re: Re: Re: Re: Re: Re: Downloaded

tomtom said:
projected NUMBER of numbers...newer knew how many numbers I will get as a final result...
Click to expand...

Are you talking about the selector method here?

And as for the thread, I created it there as it isn't relevant to this thread.



Andrew
 
T

tomtom

Member
Re: Re: Re: Bypass

winhunter said:
But perhaps this discussion should move to another thread in the WINHunter section....


Andrew
Click to expand...

Sure.Hope you might find some time and continue with developing the program ...However, thanks for the answers..
 
D

Dennis Bassboss

Member
Re: Re: Re: Bypass

winhunter said:
Yes, these two settings allow you to set the range of history that filter should target. Then, when you backtest the stack, you can see how well that setting worked against the entire history.

Think of it as a sliding slot. The bypass sets the starting position of the slot, and the draws to use sets the width of the slot.

There is also a cyclic history function within the filter, which allows you to set when the filter is on, and when the filter is off. You could create two filters, each working inverted of the other, and both of them with different processors running beneath them. But perhaps this discussion should move to another thread in the WINHunter section....


Andrew
Click to expand...
Very interesting..Yes...We are under BUSHAPPY thread...lets move... :agree: :agree2:
 
bloubul

bloubul

Member
winhunter

Well I got the first (old) method as well as the latest one.
The only negative part is that it needs to be completly automate. It is very long process at present, but other than that I can't fault it in any way. I think BushHappy has something here, and I must compliment him on a massive task.

BTW where can I download your "Software"


BlouBul :cool:
 
B

BushHappy

Member
Andrew,

When I first started, I had Criteria for Occurences etc.
I admit I was very lucky at the beginning, but I the method has since failed quite a few times.

I have now decided to do away with criteria and provide tables showing milestones/records of occurences in each category and how often these records are broken, equalled or nearly hit.
See sheet '100 Drw Cat Predict for Drw 433' of file 'BH No Elimination Method - Predictions.xls' You will see that if #21 hits in the next draw, it will have occurred 21 times in 100 Draws. This has only occurred 1 in 15.86 Draws.
Now look at Columns M to Q of sheet '100 Draws Stats' of file ''BH No Elimination Method - Predictions.xls' You will see that this occurred 21 times in 333 draws - hence 1 in 15.86
Column P is for the occurrence interval The highest occurrence interval is 78. We could say that, it will be unlikely to ever have an interval of 79 - thus if an occurence of 21 in 100 Draws has not occurred for 78 draws, then any numbers falling in this catergory will probably hit.
Column Q is consecutive occurences - the highest is 2 (1 time). We could say that, it will be unlikely to ever have a consecutive of 3- thus if an occurence of 21 in 100 Draws has occurred 2 times in a row, then any numbers falling in this catergory will probably not hit.

The final decision on wether to predict a hit or a miss for a particular number is left up to the user, as I don't think a PC can think for you - maybe with AI?

I will be adding more columns to the 'stats' sheets, but this takes time and the files will get bigger. In the end, maybe set theory can be used - intersecting sets. I will have to get out my old books.

I have not really studied WinHunter, so when you use its terminology, I am not sure what you are talking about. Anyway it will be an honour for me if you can use some of my ideas in your program.

Cheers,
BushHappy
 
You must log in or register to reply here.

Sidebar

Top