Sangoma-number divination

[Springbok]

Hello Barge
This is how I do it. Load the file, click on line 30. Look at the criteria window. Type in the code for the relevant skipsitter. I then uncheck all the columns I don't want to see for when I get the result all I want to see is the median and the relevant skip cycles asked for, no more no less. Then save the particular file with a relevant name such as Skipsitter2. The 2 will tell you that it only pulls out pairs which have appeared at least twice above the median. Below are some of the codes I use for each skipsitter file. It may be crude but it works brilliantly.

When you run tests note the range of draws you use. I use only around 100 previous draws. Test with the full bore pairs, then test with the alpha beta files. Note the difference in accuracy.


Skipsitter2(Draadsitter2)
#20 = #1
and
#21 > #1
and
#22 > #1

Skipsitter3
#20 = #1
and
#21 > #1
and
#22 > #1
and
#23 > #1

Skippsitter4
#20 = #1
and
#21 > #1
and
#22 > #1
and
#23 > #1
and
#24 > #1

Skippsitter5
#20 = #1
and
#21 > #1
and
#22 > #1
and
#23 > #1
and
#24 > #1
and
#51 > #1

 

[barge]

Hi Springbok

Many thanks,

I will give it a try. It is a different approach to mine, and probably easier(!). I was trying to put each skip up to 11 (or more) into it's own column, save it, then apply whichever filter I wanted to it.
I notice in skipsitter5 you have a column 51, is this column a skip? High Skip?

 

[Springbok]

Hi Barge
Col 51 to 59 is reserved for skips greater than 4. Let me explain how I look at skip cycles. The first skip(current referred as CS) is recorded as current as it is not a completed cycle. So I regard the first skip cycle as the second slot and so on. On the window it reports as the following;

CS......S1......S2......S3......S4......S5 and so on

To incorporate the fifth cycle I add columns until I get 51. I rename it as S5 and I delete all the extra columns that I created to get to 51. It is crude but it works well. The opening batsmen,skipsitters and inbetweeners all go up to 8 skip cycles. I have even dropped them down to 2 skip cycles if I am checking trips.

So create all these filter files and you too can have fun 'n games in conducting experiments with pairs and trips. . What you will find is interesting.There is absolutely no substitute for first hand experience working with pairs and trips at full bore or with the alpha beta etc files. May the force be with you.

 

[Springbok]

Coin tossing and medians

Toss a coin you got a 50-50 chance of heads or tails. So it roughly equals the median. Here is a typical result of a coin toss experiment. This is taken from 200 tosses. The number at the top shows the possible streaks and below the occurences(ie the runs of heads or tails before the streak reversed).

9.........8..........7.........6.........5.........4........3.........2..........> streaks
(1).....(0)........(1).......(2).......(4)......(2)......(16).....(20) > number of streaks

In a 200 toss run the coins were in some type of streak mode in 46 occurances. Therefore in 154 occurances they were yo-yoing as a tennis ball from one side of the median net to the other. It makes you think, doesn't it? Coins yo-yo more than streak.

 

[barge]

Springbok

Hi Springbok,

Thanks again, your method is certainly a lot simpler than mine, I probably over complicated it. I will give your's a try, as I can alter .clm files I already have according to the filter I want to use. It looks as though we may be spoiled for choice for the Tball, but we shall see....with the force, who knows??!

 

[jokerh1977]

Howdy ho my fellow dreamers. I've been absent off the board for a week or two and what progress has been made. Patrick, the program is totally different from the one I orginally downloaded! Kudos! It looks great!
I've been playing with my own version of the draadsitter column/filter on full bore pairs. The best I'm doing is 1-3 numbers out of a pool of 5-8 numbers (this is a 5/39 game). Not bad. What other results or methods are others incorporating? Perhaps we should begin pooling ideas/filters? Springbok, could you enlighten us with some or your methods of narrowing down? Perhaps step by step? I have not had ANY success using alpha-beta files, only the full bore method. Good luck guys!

 

[Springbok]

Mornin' Barge and Patrick

I added another criteria to the opening batsmen, inbetweeners and draadsitters which you might find useful. At the end of each specific filter I add another column and write in < #1. This is now what will happen. Supposing I load draadsitter4, the filter will only pull out pairs or trips which have exceeded the median 4 times only. In the past the filter would have also pulled out pairs which had exceeded the median more than 5 times. It also excludes the pairs and trips which have exceeded the median three times or less. So now you can check the efficiency of the skip cycles precisely in prediction. Another example is draadsitter2. It will pull out only pairs and trips which have exceeded the median twice, excluding all the other skip levels.

 

[barge]

Springbok

Thanks for that, narrows things down a little. I have noticed that results from sorts vary quite widely, ie whether a-z or z-a, logically z-a would seem to be the obvious one....but is there a case for using a particular one on each column?

 

[Springbok]

Barge

I don't know what kind of tests you are running but they seem to be nothing like the tests I run. For example I examine the performance of the Inbetweeners over 2 skipcycles, 3 skipycles etc over the median. I record the results. A success is a hit within 3 draws. If not it is a failure. This is done with trips and pairs. The results range from absolutely brilliant to very poor. I am using the full bore that is 1176 pairings and 18000+ trips.

You can replicate this test with inbetweener3 and inbetweener4. The data file is up to date yesterday teatime. Leave out the last 15 draws and load the data file. In the columns load intbetweener3 and hit the pairs-brilliant result. Now load Inbetweener 4 and hit the pairs- totally brilliant result. Both results bang on target. Remember intbetweener3 only pulls out pairs with precisely 3 skip cycles above the median- nothing else. Inbetweener4 pulls out pairs with precisely 4 skip cycles above the median-nothing else.

These results are totally brilliant. But the same test run on another area in the data file can produce a very poor result.

By the way I was using the 49s as a 6 number game with the bonus ball incorporated. The combinations are in sorted order. To get the same result your data file will have to be like that. I usually test as a 7 number game but I had forgotten to move the control to 7 so it was the 6 number game.

 

[Springbok]

Let's play yo-yo

Gee whiz!!! Thinking about most pairs and trips are in a non-trending mode like in the coin tossing experiment but are yo-yoing around the median I made two files. Yo -yo up is when the last yo was above the median so I called the file up, geddit! The same applies to the down file as the last yo was below the median. Incorporate the following criteria in you filters and be prepared for the shock of your life.

Each file scans for 7 yo- yos. If nothing comes up then delete the last yo and try again. Keep on deleting until results come up. Use pairs and trips and check results.

yo-yo up
#20 > #1
and
#21 < #1
and
#22 > #1
and
#23 < #1
and
#24 > #1
and
#51 < #1
and
#52 > #1
and
#53 < #1

yo-yo down
#20 < #1
and
#21 > #1
and
#22 < #1
and
#23 > #1
and
#24 < #1
and
#51 > #1
and
#52 < #1
and
#53 > #1

 

[barge]

Great stuff! I will try the Yo-yo later.
I too use a sorted data base for the 79s, but the seven number version. I have been trying to get a result on the next draw with results reasonably close to the grids we were using a while ago, although I never managed to produce Alpha files as you described them.
I use the full bore data, and usually Draadsitter4. Since you published the other variations, I am applying these across the board, as their speed of use helps a lot in trials, well done for that. I used the D4 earlier today(after the lunchtime result) and got a good result with the pairs for the teatime draw. I didn't bet as I was Shanghai'd into providing transport for some retail therapy for she who must be obeyed
I noticed one number came up in the Inbetweener3.
I mentioned sorting on the columns before as I have noticed that occasionally a column (Max, N+X etc) will come up with a result out of the blue, but interpretation of course is needed to decide if these are good results or not.
Anway, let's yo-yo!

 

[struxo]

to Springbok

Hi,
i just wanna check i doing right with your idea with <#1

Skipsitter2(Draadsitter2)
#20 = #1
and
#21 > #1
and
#22 > #1
and
#23 < #1


Skipsitter3
#20 = #1
and
#21 > #1
and
#22 > #1
and
#23 > #1
and
#24 < #1


Skippsitter4
#20 = #1
and
#21 > #1
and
#22 > #1
and
#23 > #1
and
#24 > #1
and
#25 < #1


Skippsitter5
#20 = #1
and
#21 > #1
and
#22 > #1
and
#23 > #1
and
#24 > #1
and
#51 > #1
and
#52 < #1

 

[Excellence]

Wonderful software Patrick! Many, many thanks. Plus an excellent documentation! What more can somebody wish? Well done!

A tone of ideas Springbok since I last visited and all very interesting. I also developed my skills a little in theory of probabilities lately and I reached the conclusion that it is best to compare the results of any system whit the random play. I once played a system for 3 weeks and after I compared my results whit the theoretical values I realized that if I had chosen numbers randomly I would have probably won the same amount… or lose the same.

I calculated some probabilities for each strategy that you presented here. They are all calculated for pairs but if they beat the odds for pairs they will surely beat the odds for trips to. For 7 / 49:
- opening batsmen: 57,81% for the next 3 draws; 68,35% for 4 draws;
- in-betweeners: 32,51% for the next 3 draws; 38,45% for 4 draws;
- median sitters: 14,06%.

For 6 / 49:
- opening batsmen: 52,41% for the next 3 draws; 62.85% for 4 draws;
- in-betweeners: 31,94% for the next 3 draws; 38,30% for 4 draws;
- median sitters: 13,36%.

This will help you evaluate your systems I hope. If the results are better, then you are definitely on the right track. If they are better but close to this values it might be still random.

Regarding the Skips of Skips. Here are some medians for skips of interest. This medians refer to the chance that the respective skip will be drawn.
-skip 4: 0 or 1 draws, for 0 47,17%, for 1 72,08%;
-skip 5: 0 or 1 draws, for 0 the probability is 37,68%, for 1 is 61,16%;
-skip 6: 0 or 1 draws, for 0 the probability is 37,68%, for 1 is 61,16%;
-skip 7: 1 or 2 draws, for 1 46%, for 2 60%;
-skip 8: the same as skip 7;
-skip 9: 3 draws;
-skip 10: 4 draws;
-skip 11: 5 draws;
-skip 12: the same as 11.

About the skips above 8 I am not so sure about those figures, harder calculations where made but I don’t think that it can be miles far. Also I must say that this probabilities are calculated for 7 / 49 game.

Regarding the last system presented. You said that you tossed the coin 200 times right? If is true, and if I know what a yo-yo is, then you should have 66 occurrences, not 154. If you refer to the fact that the numbers in lotto game oscillate up and down their median then the median for pairs is 2 not 1. I must experiment more whit Patrick’s software.

Hope this helps a little. Good luck.

 

[Patrick123]

Hi Everyone, been a bit busy lately, so have not had much chance to do anything. I'll be going through the posts and adding my 2 cents worth

Struxo, your one filter, Skippsitter4:

#20 = #1
and
#21 > #1
and
#22 > #1
and
#23 > #1
and
#24 > #1
and
#25 < #1

needs to be changed to

#20 = #1
and
#21 > #1
and
#22 > #1
and
#23 > #1
and
#24 > #1
and
#51 < #1

Column #25 is the tag column, used by the number filter.

Regards
Patrick

 

[Springbok]

Hi Struxo

The coding looks OK to me but delete the #25 and substitute #51. Remember you can port the coding to the other filters and be aware the opening batsmen require the first line with zero. Remember the parameters of the filters- that is a hit within 3 draws(with the 6/49 game). Test on your own lottery game. The results will sometimes delight you and sometimes infuriate you. Which is the best sequence. It varies. Draadsitter4 this morning gave me 3 pairs which hit in the very next draw which I was testing in the UK 6/49 but this will not happen every time(if only )

Hi Patrick, I noted your correction.

 

[Springbok]

Dumela Patrick

I am trying to find out what "hello" is in tsotsi taal , trying to expand my linguistic horizons. What makes me the moer in is when the filters don't behave within their parameters. As you said previously if they only phoned us we would tell them what the result should be. This would ensure a fair result.

 

[Springbok]

Hi Excellence

I will check your findings on the yo yos. The yo yo filters over time do not conform to the parameters so far. So more coin tossing exercises today If you are right then they become very useful in anti- prediction.

 

[struxo]

to Springbok and Patrick

Thanx guys for help.
I testing with swedish lotto 7/35 and keno.
I have good results with keno:
i check
100% single
100% trips with filters
skipsitters,inbetweeners and opening batesman
i also checking numbers who are 1 or 2 skips to max(works great in 2-3 draws)

For lotto i try to find out number to repeat from previous draw and play it in every line. Success about 60-70%

 

[Patrick123]

Hi All,

I've just updated WinnaLotto. I've added in a Filter Evaluation button.

To use it. Load some historical draws.

Load the required filters, goto Draw Analysis, then click the 'Filter Evaluation' button.

It will go through all the historical draws, apply the filters, and at the end give you a summary.

'x Rows' - The total rows evaluated.
'x Filt' - The number of filtered rows.
'x 1' - The count of hits where one ball was in the draw.
'x 2' - The count of hits where two ball were in the draw.
'x 3' - The count of hits where three ball were in the draw.
'x Hit' - The count of hits where at least one ball was in the draw.
'x Perc' - ('x hit' * 100) / 'xFilt'

You can now load another filter and evaluate it, no need to reload the data. But remember to recalculate if there are formulas in your columns.

Regards
Patrick

 

[Patrick123]

The 'Filter Evaluation' is ideal to use for spotting the filter trends on a macro level. This will help decide whether it's worth using that filter or not for the next draw.

Regards
Patrick