Sangoma-number divination

[Springbok]

Hi Patrick

I have not looked at the latest incarnation of Winnalotto. I have been checking the internal filter. What I find is that it remains static. I was checking its accuracy on the last 11 draws of the 49s. Top ranking was pair {3,14}. As I went down, reloading draw by draw, it still gave the same pair. I then loaded the Uk draws and lo and behold it gave the same pair. This appears to be rather strange. When I load the data and then hit the pairs and go to the internal filter , it has already sorted the internal scores from lowest(top of the column to highest bottom of the column0. In the past I would have to right click and activate the sorting. Very puzzling. Has the computer got the sangoma virus?

 

[Patrick123]

Hi Springbok,

Very strange indeed, I'd suggest, clear both internal & column filters, then make sure that the right numbers of rows is displayed for balls, pairs & trips. Thereafter load your filter again, maybe it is some hidden character that the edit box is not showing, but causing hassles with the display.

Regards
Patrick

 

[Patrick123]

Hi Everyone,

And thanks to everyone for their well wishes. I've modified the Filter Evaluation slightly.

When the filter does not display any rows, 100% is evaluation figure instead of the original 0%.

What this does, it means that the filter is working properly instead of displaying the 0%. If anyone has a suggestion what might be the best way to calculate this, I'd appreciate it.

Thanks
Patrick

 

[Excellence]

Hi Patrick,

Again congratulations for the program. I read the documentation and many pages in this thread but I still have a lot to catch up. I also tried the latest update. I noticed that, if a filter selects in a certain draw 24 pairs for example and from those 24 pairs 12 contain at least 1 number drawn then the probability is calculated as 50% right? My problem is the next: we do not play all 24 pairs. That’s why I think the probability might not be the right one, if we want to compare it to random play. Calculating the probability of choosing 24 pairs in a draw is different from the one calculated for one pair.

I also have a question related to the filters. Skip 1 is the current skip? At the beginning I did not thought so but I read your answer to Struxo related to the filter called Skippsitter4. As I understood a skippsitter is a ball that has the current skip equal to the median. In your columns the current skip is #4. In that code Skip1 (#20) is equal to the median (#1). The next 4 skips must be above it. I got it right or a skippsitter is a number that has the last ended skip equal to its median?

 

[Springbok]

Goie aand Patrick

Had a look at the latest incarnation of the program. Wow! some cool programming there(you are converting me to windows programming). The problem I have is understanding the results. I will run the checks manually, record the results and then look at the screen and try and make sense of the output. I understand the extreme left(past1. past2 etc) . At the moment I think it is easier to win the lottery than decipher the output-but I am a slow learner

Thinking about the 100% evaluation to show the filter is working properly how about a green dot or simply a symbol like OK.

Fooling around this morning with the UK 6/49 using draadsitter4 and 3 plus the historic high gave a predicted 3 pairs for the next draw. The high came up, and one of the other pairs contained two numbers that hit. That's 50% of the UK lottery solved on that occasion.

Ignore my earlier comment on the internal filter, I think I was having a blond moment.

 

[Springbok]

Hi Excellence

The next 4 skips must be above it. I got it right or a skippsitter is a number that has the last ended skip equal to its median?

Let me answer for Patrick. A skipsitter is sitting on the median in its current skip which is ongoing. The idea behind the skipsitter technique is to work out what it will do next. Will it hit or will it carry on. If it has in its previous skip cycles always hit above the median then we hope and pray it will rebalance and hit next, simple.

 

[Springbok]

Hi Excellence again

I would like to add that the higher the skipsitter(draadsitter) number the more precious it is. . It is like a diamond. I treat my draadsitter filters with the greatest of respect if they forecast one pair only. When you set up the filters make sure they give you the genuine result. For example skipsitter7(draadsitter7) must only give you the last 7 skip cycles above the median. No more no less, got it! If you do not set limitations to each filter then they will give you zillions of pairs. If you come across a forecast with skipsitter 7 or more, bow down and humble yourself before it.

 

[Excellence]

Thanks for the explanation Springbok. I now realize that Skip1 is the current skip. I lived under the impression that Skip1 was the last ended skip. Thanks again.

 

[Excellence]

Thanks for the tip. I will surely do that.

 

[Excellence]

One more question if I may. How can I analyze above Skip5? I read a previous post of Springbok and all that I had understood is that I have to add columns until the 51, which represents the Skip6. I got it right?

 

[barge]

Hi Patrick123

Well done regarding the new enhancements to the prog!
I am a bit confused over the test results. The results themselves are perfectly clear but how many draws are they evaluating? For example, I loaded 10 historical draws to check Opening Batsmen. My results show 10 rows, but the columns Pfilt and Tfilt indicate that far more than 10 draws were evaluated. How do the results relate to the 10 draws I loaded in the first place?
The filter, I have to say, has produced some remarkable results!

 

[Patrick123]

Hi Everyone,
Thanks for the thumbs up. Let me try and clarify some of the vague points.
Firstly, Columns 30 to 50 are reserved for user created columns, thus #51 to #60 will be skip cycle 6, 7, and upwards.

The percentages that are displayed, I take more as a confidence percentage than anything else.

The results of this filter I created on the last 12 draws (latest: teatime 5 Sept).

#20 <= 4
and
#21 <= 4
and
#21 > #1
and
#22 <= 4
and
#22 > #1
and
#20 <= #1
and
#20 = 2

	B Rows	B Filt	B Hit	B Perc	P Rows	P Filt	P 1	P 2	P Hit	P Perc	T Rows	T Filt	T 1	T 2	T 3	T Hit	T Perc
Past 1	49	0	0	100	1176	0	0	0	0	100	18424	8	2	0	0	2	25
Past 2	49	0	0	100	1176	4	1	0	1	25	18424	10	1	0	0	1	10
Past 3	49	0	0	100	1176	6	3	0	3	50	18424	30	13	1	0	14	46
Past 4	49	0	0	100	1176	8	1	0	1	12	18424	64	30	3	0	33	51
Past 5	49	0	0	100	1176	5	2	0	2	40	18424	35	22	2	0	24	68
Past 6	49	0	0	100	1176	2	0	0	0	0	18424	6	2	0	0	2	33
Past 7	49	1	0	0	1176	14	5	0	5	35	18424	10	4	3	0	7	70
Past 8	49	0	0	100	1176	3	0	0	0	0	18424	27	1	0	0	1	3
Past 9	49	0	0	100	1176	0	0	0	0	100	18424	8	1	0	0	1	12
Past 10	49	0	0	100	1176	6	2	0	2	33	18424	8	3	0	0	3	37
Past 11	49	0	0	100	1176	2	0	0	0	0	18424	25	0	0	0	0	0
Past 12	49	0	0	100	1176	4	3	1	4	100	18424	38	19	16	3	38	100

These values are the values that are displayed at the bottom of the Drawing Analysis screen.

It tells me:

• If a ball is selected, don't play it(Past 7).
• If no balls/pairs/trips are selected, then the filter is accurate as none fitted the criteria.
• Of any pair that is shown, basically only one number from the pair will be drawn.
• For trips, Usually only one of the 3 balls will be drawn.
• It also shows me that sometimes when it comes in, it hits hard (Past 12) very likely the same number or numbers being repeated. so If you use this filter, select the numbers that appear a number of times in the trips or doubles screen. In this case, 11 on the pairs screen and 11 and 32 on the trips screen.

In Summary, when it works, it works, otherwise it can be quite erratic, so don't use it on it's own, rather back the results up with another filter.

The Evaluation extends only to the number of historical draws that you have loaded. i.e. If you have 10 loaded you will get 10 rows labeled 'past 1' to 'past 10'. If you spot a trend, you can then simply select current, analyse the results and have some sort of confidence or skepticism as to how accurate the prediction is.

Regards
Patrick.

 

[Patrick123]

Hi Barge,

The PFilt and TFilt, show the number of rows that would have been selected for that draw out of the 1176 or 18424 pairs or trips respectively.

Regards
Patrick

 

[Patrick123]

Here is a filter, like the median sitter, it is basically for trips analysis and over the last 20 draws(Teatime 5 Sept), has only predicted trips 9 times of which it missed 3 times, or 6 out of 14 rows, now it's just to get rid of those false positives

#20 = #2
and
#1 = #2
and
#1 = 1
and
#15 > 25
and
#6 = 4

Regards
Patrick

 

[Excellence]

Hi Patrick,

Thanks for explanations. I tested the latest addition and I was wondering how I could save the results provided by Filter Evaluation button. It will give some interesting comparisons between filters and, if you want later to add a criterion to a filter you can always compare the past results whit the new ones.

 

[Patrick123]

Hi Excellence,

Thou wish is my command, watch this space

Regards
Patrick

 

[barge]

Hi Patrick123

Thanks for the clarfying that. I had input filters which produced a lot of hits which seemed to indicate that far more than just the 10 draws were being analysed. In fact these were real hits, but spread over a lot of draws......

 

[jokerh1977]

Attencion Patrick

Well... I have an idea, tell me what you guys think. Patrick, would it be possible to create a button (let's say "Number Frequency") that would count up the single numbers in the pairs or trips? Let's say you have a column/filter that produces 900 trips that meet your particular criteria. The Number Frequency would show us, number by number, how many existed in those 900 trips.

Example:
322- 38
264- 27
165- 16
etc.

This might help us find a "probable" number to appear. Whaddya think?

 

[Patrick123]

Hi Everyone,
I've updated WinnaLotto. There is now a save button, thus whatever is on the grid can be saved as a csv or text file.
I added a right click menu option to graph the 'Number Frequency' , so you simply right click anywhere on the grid and it will pull the count onto the graph as per jokerh1977's definition, this way, it does not destroy the data that you're viewing.
Regards
Patrick

 

[Patrick123]

You have the capability of:
Loading a filter and sending the Number Frequency to the graph.
You can then repeat the process with a bunch of similar filters before evaluating the graph, as each series will be labeled with the filter's name. Though I'd suggest comparing trips with trips & pairs with pairs so that you don't get a biased view.

Regards
Patrick