Sangoma-number divination

Patrick123

Member
Analyzing the NZ Lotto Draws, excluding the Bonus Ball, there is about a 90% chance that any 4 balls come from the last 10 draws, 3 from 7 Draws, 2 from 5 Draws & 1 from 3 Draws. So I was thinking of possibly using your method to get the high probability balls, then removing taking whats left only of the last 10 draws and take a cut of the best weighted ones for position 4, a cut from the last 7 draws for ball 3 & repeating the process for the ball 2 & 1 as well. I'll try and get this added in during the course of the weekend.

Regards
Patrick.
 
Good day gentlemen. First I must apologies for my poor english skills, that’s why I will try to be as short as possible. The ideas presented by Springbok, and not only, are surrounded by lot of interest. But I am sure that for some members they are surrounded by a lot of fog to. Unfortunately I could count myself among those. That’s why I would like to put a few questions related especially to this last system presented. I must say from the beginning that no answer is obligatory. So let’s start.
The first step is to create groups of 2 numbers from the total of 49. 23 groups consist of 2 numbers and one of 3 numbers. The alpha numbers are created randomly. All clear by now for everyone. Then Springbok says that he is using 12 groups, each one consisting of 24 pairs. At least that is what I understand. Why making more groups?
The main problem of every lotto strategy is: which numbers do I play? Let’s say we have 12 groups (alpha, bravo etc). According to the Skip Chart presented by Springbok, obtained whit the MDIED, all alpha pairs have a median skip of around 2. That means that in 50% of the cases the one of the numbers from the pair will hit within 2 or 3 draws. So if we play both numbers for 3 draws let’s say, we will have at least a 50% winning chance. Hope I got this correct. But don’t forget that the median is a variable. But then where did the 4 skips came up? And those 50% or 75%? How can we say whit a certain probability that a number will be drawn in the next 4 draws? Well we could use Ion Saliu’s programs. I say we could but in my opinion it is more difficult in this case. The problem: what is the probability that, from 24 pairs the pair 2 (e.g.) will be drawn? If we use Mr. Saliu’s SuperFormula.exe, we obtain p = 29,29% for DC = 50% and N = 2. For N =3, p = 20,63%. My math skills are also poor so I cannot find right now another way to calculate this. Anyway we calculated this probability for a reason: we can calculate using again SuperFormula.exe the DC for this p for different N (e.g.: p = 20,63%, N = 8, DC = 84,25%). This means that if we choose a pair that has a skip = 3 we have 84,25% chance that the pair will be drawn in the next 5 draws. At least this is how my mind interprets. I was wrong before a lot of times, I will be wrong many times.
So the question: how do we choose the numbers and from which group (alpha, beta etc)? The tendency I observed is to choose the numbers that have large skips, preferably the ones close to the maximum skip. But we must take into consideration that this maximum skip is from the last 200 draws. Quite few in order to have a maximum skip value representative for the number. (e.g.: the current skip = 8, maximum skip in the last 200 draws = 12, this does not guarantees us that the pair will be drawn in the next 4 draws, never 100%).
Patrick123 made an interesting chart whit numbers and weights. I must say I do not know what a weight is so that chart is totally unclear to me. If Patrick wishes to explain it would be great, if not then it was not to be.
Already a long post so I finish here. Thank you for reading this and for possible replies.
 

Springbok

Member
Hello Excellence
No need to apologise for your English as what you have written is very clear. Let me answer your questions as far as I can.
*Why am I using 12 groups? I am looking for certain patterns in the skip history. The more groups I have then the better chance I have of finding what I am looking for. If alpha has not got what I want then I go to bravo and then to charlie, then to delta etc. Because the patterns in each number group are different I increase the chances of finding the pattern I want.Usually the first 4 groups give me what I want.
* Why are four skips important? We are betting on a pair of numbers that will hit in 4 draws so we can make money. If a number in the pair hits we are in profit if we increase our betting up to 4 draws.How do I know it will hit. When I look for a suitable pair of numbers(alpha1, bravo12, etc) they must show a certain type of behaviour over say 200 draws(you could have 300 or whatever you want).I look at the current skip(n). I add 4(represents the next 4 draws) to n. I have a total. When I look at the history of the skips of this number pair. no skip must be greater than n+4. Logically then, if the behaviour of this number pair has not exceeded n+ 4 in its history then it should be a safe number to bet on to hit within the next 4 draws and guarantees us a profit. When you find these characteristics in a number pair then you bet on that number pair.
*Ignore Saliu's SuperFormula in this case(it is theoretical). We are dealing with real history.Your interpretation of DC is correct.
I am sure Parick123 will explain his weightings for you.
 

Spot

Member
Optimun number of Draws

Patrick123 said:
Well the NZ Lotto predictions were a bit of a disappointment though #16 came in,
Hi Patrick,
Do you use all 1000+ draws for your calculations?
I did some test with another program and got best results not using the bonus ball, and only using the last 400 draws.
Spot
 

Patrick123

Member
Hi Excellence,
With regards to the weighting, I start off with all balls having a weight of zero.
I have about 50 of these Alpha sets with 2 pairs. If the last skip value is above the median, I add the the difference between the median and last skip value to each number's weight that makes up that pair. I also allow you to define 2 more groups (beta and delta) with x amount of numbers in each tuple, and go through the same weighting process. Sorting the numbers by their weight, I then select the ones with the highest value as theoretically they are the most likely to be due. In the program, under Ball Analysis, Draw Probability, you can see thse ratings.
Selecting, Either Alpha, Beta or Delta in the Combination Box and clicking on 1 -50 will show you that sets pairings, with the median value and skip history.

Hi Spot,
Yes, I used all 1049 draws for my calculations as well as the bonus ball, very likely that may be one of my mistakes as the trends do tend to change over the years. Thanks for the tip, I will add in an option where we can decide how many draws to analyse.

Regards
Patrick
 
Thank you gentlemen for your prompt responses. Unfortunately I find myself in the situation to ask two more questions.
The first one is for Springbok. Now I understand very clearly why we should make more sets of pair numbers but the only thing that remains unclear is what we are searching for, what kind of pattern. I mean we could choose any number whit a current skip equal to 0, 1, 2, 3 or more as long as this current skip + 4 is less then the maximum skip in the last 200 draws or more. The current patterns that I am ware of are: the last 2 or 3 skips before the current under or above the median, spia numbers. And this patterns only offer an slightly improved chance, not guarantee success. So what are we looking for? I will try to make some analysis of mine based on 49s draws, see if I can discover something but the odds are small. We are not the first people who try to find patterns.
Many congratulations Patrick for the program. I wish I would know to make things like that to but it is a problem of time right now. The second question of mine is related to this program and it may be transformed in 2 questions. I downloaded and unzip the program. I have imported the csv file. After I done that I went to Ball Analysis, click on Draw Probability but it gaved my an error. I am sure that I must do something else before but I don’t know what (I strongly think it is related to the fact that when I imported the database file the program said that the date is not valid). Also I would like to know if the actualization of the draws is made from the program or from the csv file?
Thank you for your patience and for your possible responses.
 

Patrick123

Member
Hi Excellence,
The invalid draw date was most probably related to me converting the date into a date value and localization(day month year vs month day year) was most probably causing the problem. I don't really use this so have changed it into a string to suite any format you require(display purposes only).

Ensure that the Num Balls setting is set to 49 or whatever number of balls your lotto uses, as well as the Balls Drawn setting 6 for lotto 7 to include the bonus ball.

The format of the Comma Separated Value file is:

Draw Number, Date, Ball1, Ball2, Ball3, Ball4, Ball5, Ball6, Bonus Ball - you may leave out the Bonus Ball if not using it.

It is important to change these settings first before importing the file as the structure is created up front.

Please re-download the updated program from the earlier link and let me know if you have more hassles.

Regards
Patrick
 

Springbok

Member
Right folks we are going to look at the n + 4 <= H(highest skip in history) to not only evaluate the probability of hitting in the next 4 draws but which draw it will hit and the probability of this happening and how to avoid unreliable pairs.
Let's look at some pairs in my zulu file and see what they are communicating to us.
*Zulu2 is hot. 100% probability of hitting in the next four draws.60% probability of hitting in the next draw. Why? In its 200 draw history the skip of 9 has occured five times. On three of those occasions it hit in the next draw. So 3 / 5 = 60%.
* Zulu5 is the rattlesnake number. You don't know what it will do next. Its current skip of 17 is the highest in its history. What will it do next? Who knows, It may bite or you it may not.0% probability.
*Zulu6. 100% probability in the next 4 draws and 66% probaability to hit in the next draw.
*Zulu10. 0% probability as there is no history to guide us.Another rattlesnake number.
*Zulu22. Probability of 100% for the next 4 draws and 66% for the next draw,
So what happened in the end. Well Zulu5 and 10 never appeared in the following 4 draws and zulu2,6 and 22 all hit in the next draw.

* zulu # 2 * Hits: 54
------------------------
* Skips >>>>> 9 10 0 4 11 1 3 2 1 0 2 0 2 4 6 1 1 1 2 1 1 1 2 0 2 2 2 6 3 3 3 5 3 0 1 3 1 2 0 4 4 1 0 9 2 1 3 1 0 3 4 2 9 2 * Sorted Skips: 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 4 4 4 4 4 5 6 6 9 9 9 10 11 ** Skip Median: 2
* zulu # 5 * Hits: 55
------------------------
* Skips >>>>> 17 0 6 6 3 8 1 0 0 3 1 1 1 7 3 3 0 3 3 5 3 0 1 0 1 1 0 2 3 3 0 1 0 0 1 1 6 2 2 2 6 16 6 4 2 1 0 1 0 0 0 0 1 3 0 * Sorted Skips: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 4 5 6 6 6 6 6 7 8 16 17 ** Skip Median: 1
* zulu # 6 * Hits: 45
------------------------
* Skips >>>>> 13 10 2 8 4 0 0 0 3 4 16 0 1 0 4 0 1 0 6 4 9 2 0 1 0 1 2 4 0 4 1 13 1 1 6 1 0 0 7 0 3 5 4 1 4 * Sorted Skips: 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 2 2 2 3 3 4 4 4 4 4 4 4 4 5 6 6 7 8 9 10 13 13 16 ** Skip Median: 2
* zulu # 10 * Hits: 59
------------------------
* Skips >>>>> 9 0 3 0 1 3 1 3 2 0 1 0 4 1 3 4 0 8 3 3 0 0 3 2 0 0 3 3 1 1 1 3 4 1 0 0 0 5 7 2 3 2 7 4 3 3 2 2 0 4 6 5 6 1 1 0 5 0 1 * Sorted Skips: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 5 5 5 6 6 7 7 8 9 ** Skip Median: 2
* zulu # 22 * Hits: 54
------------------------
* Skips >>>>> 9 2 0 1 0 2 3 2 0 1 3 0 2 2 1 2 3 3 1 11 6 1 6 1 0 1 1 4 2 0 0 0 1 6 4 0 1 6 6 1 1 9 6 2 4 0 3 4 4 7 1 0 1 3 * Sorted Skips: 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 6 6 6 6 6 6 7 9 9 11 ** Skip Median: 2
 

Springbok

Member
To Excellence

Read my previous post to this to see the patterns we look for in the history. Ignore medians with this particular technique. They are not relevant. Don't worry too much with theory. The history of the number pairs know nothing about theory.
 
Patrick I am glad to say that your program works. I have successfully imported the csv file and after that the program worked fine. I assume that the second column “Total” from the draw probability represents the number weights right? Anyway as long as I will use it I will probably understand how it works (hopefully).
Springbok I wait for the results of your strategy. As far as I see you pick numbers that have a skip close to their maximum skip in the last 200 draws, and even numbers that have skips equal to this threshold. In addition you argument your choices based on statistical observations. I don’t have the skips for all the pairs, therefore I cannot say what is the difference between these numbers and the others (I have seen that in most cases the current skip is bigger then at least the last 2 before). I now try to work at two strategies of mine, and since I know only Excel it takes a little more. Anyway I hope you continue publishing results and especially examples of how this pair is right to be played and how another is not.
I think that it would be interesting to know for a certain number of draws = k, how many skips = n will appear for a certain number (work on it). I think your strategy is close to this goal, but you support your choices on statistic. Hope you are right. Good luck!
 

Springbok

Member
As we expected according to logic we had a 100% strike rate. Delta16 hit in draw 1 with the added bonus of both balls striking simulataneously. Delta23 hit in the 2nd draw, followed by echo17 in the third and finally charlie18 hit in the 4th. Even the rank outsider bravo19 made a strike. Was the number 5 significant being in bravo19 and echo17? Certainly was. This gives an idea with what to expect when you use sangoma using the n+4 <= H format with a caution about using numbers whose current skip is equal to their historical highest skip(please not this Excellence). Although this method is fairly crude it gives you the advantage of psychologically getting a grip on a chaotic event. No more will you believe the nonsense about all numbers having an equal chance to appear in a lottery. This rubbish is even spouted out by academics in universities who haven't a clue what they are talking about. Chaotic events have their boundaries and we operate on these boundaries.
To put it more dramatically, we are like a lion studying its prey in the bushveld, seeking out the easiest and weakest one to catch. Once you espy your prey you leap forward, catch it an eat it. Weak fuzzy numbers are our prey.
 
I think I understand a lot more now. I might add the next formulation regarding the current skip of the numbers selected: H – 4 <= n < H (I lived under the impression that there is more). I don’t have skips for pairs but I assume that creating 3 or 4 groups assures you some numbers that respect the above criterion. I also noticed that so far you’ve choused 4 pairs first time, 5 the second time. Now the systems is more clearer to me. Better later then never. I would like to say that obviously, the system would not have a 100% success rate even if we would play only numbers that have n = H – 4. I would consider forming groups of even three numbers and play a smaller interval, like H – 3 or even H – 2. Just a proposal. Anyway it seems that this system will prove to be very profitable. Thank you for sharing it!
 

Springbok

Member
Forecasts for Monday and Tuesday. All based on n+ 4 =< H
n(current skip): H(historical high-200 draws)

-----------------------------current skip---------historical high------probability-------probability next draw
ALPHA10{21, 35}........4........................8.....................100%....................63%
DELTA17{6, 21}...........5.......................9......................100%....................50%
DELTA21{35, 45}.........9.......................11.....................100%....................50%
FOXTROT16{31, 32}.....4.......................8.......................100%....................40%
FOXTROT23{45, 46.......9.......................13.....................100%...................33%

Gee Whiz, I hope you can read this.
 

Springbok

Member
Patrick123

Hi Patrick123
Been looking at your programme. I think am going to need some instructions. Will it load a text file as well as a CSV file? All my data files are text files. I recognise the any filters but cannot figure out anything else.
Later on in the week I will explain how to pick numbers that hit in the previous draw and the ones that are sitting on their median. In other words the focus will be on hitting numbers in the very next draw . This is different from the shotgun effect that is being used at the moment. I see you use the New Zealand lottery as a test bed. Have you tried the 5/34 section of the UK Thunderball as a test bed?
Could you add a facility to your programme so the people out there can create their own alpha, beta etc pairings and save them so it can scan the real results and convert them for analysis by the MDIED stats report. This would make your programme extremely powerful. You could even market it on the SA market once the local lottery restarts(once the local corrupt politicians get their filthy hands off it).
I must say your application looks very professional. My own VB 6 stuff has the feel of a plaasjapie who has inhabited a cave in the Sneeuberg since the Great Trek. But it works brilliantly.
 
Hi Springbok,
I know that you can open any text file whit excel and after you can save it in csv extension. If your results are well ordered in text files (number bellow number) you should not have problems. I don't know if after you opened the text file whit excel you will need to use Concatenate or not. Patrick will clear all out.
 

Springbok

Member
Thanks for the tip,Excellence. Although I have Excel on my computer, I never use it as it is not powerful enough to handle lotteries. I use far more powerful programmes. Excel could be useful in handling sangoma but sangoma does not need powerful programmes to run it.
 

Patrick123

Member
Hi Springbok, thanks for the advice, at this stage, the program does not load text files, but I'll make provision for it.
I should actually write the manual first then I can see where all my hogwash is.

The initial Settings before importing the your data:
Max Alpha: This is the number of alpha, beta, and delta sets you would like to create.
Num Balls: Number of balls in lottery.
Balls Drawn: the number of balls drawn. Setting this to 7 will include the Bonus Ball.
Alpha, Beta & Delta Balls: number of balls in each group respectively. Setting this to two, creates the pairs as you describe. The only difference is that the last group as you have it contains three balls for an odd numbered lotto. what I do is re-use a randomly selected ball so that no pair or tuple has a higher number of balls based on the assumption that with 50 alpha sets, values should smooth out.
Use Median: This will check if the last skip value is greater than the median and weight the ball pair accordingly.
Use Mode: This will check if the last skip value is greater than the mode and weight the ball pair accordingly.
Median Hit:In my revised version, I'm using this for the n+4 <= H formula.
Use Alpha, Beta, Delta, Actual: The Weighting only makes use of the checked options.
Actual
are the individual balls' stats.
To allow you to check historical drawings, and to limit how much historical lines you want to analyse, you can Skip x amount of draws and then load y amount of draws to analyse.
Manually adding a draw in is not active at this time.

Ball Analysis
-------------
Draw List: Incorrect naming, should actually be called the Hit/Skip List.
Vert & Horizontal Hit List: This needs revising. In NZ you can win by predicting the ball and draw position (1,2,3 or 4) of the first 4 draws. This was my attempt to make sense of that.
Draw Probability: This gives the totals of the Actuals & Alpha, Beta & Delta Groups, the Prev Draw Column tries to weight the balls that have occured more than three times in the last 4 draws or twice in the last 2 draws.
As Excellence correctly assumed, the Total Column, the this combined weight excluding the Prev Draw.

Selecting then either Median or Mode, then Alpha, Beta, or Delta from the combination Box, and the range 1-50 or 51-100, you then click on one of the number buttons to show that pairings hit/skip values.
I will get the save button working so that you can export it into MDIED.

Drawings Analysis
-----------------
Any: shows you where the balls came from in previous draws.
Group 7: simply grouped the balls 1-7, 8-14,15-21 and so on with a Total based off their position. i.e.
3 from group 1, 2 from Group 2 and 2 from group 3 would give me a total 3*1 + 2*2 + 2*3 = 13.
Prime/Non-Prime and Odds/Evens: Seperating the Prime vs Non-Prime & total.
Any Summary, this gives you the stats of the Any detail showing (NZ stats Bonus Ball included) that any 1 ball was drawn from the last 2 draws 93% of the time, analysing for the last 10 draws only.
Prev. Occurs and Prev Occurs BB Inc shows that a draw that contained x amount of balls has occured y amount of times. I use this to try and optimise the number order for the wheeling.

Wheels
------
Here I generate a basic wheeling system based on how many balls you require, how many balls on a ticket, the Minimum Match, the number of balls drawn, and the number of tickets you require.

I populate the grid and run the Grain value of random numbers against them, adjusting the numbers all the time and holding the best probability wheel.

YOU NEED TO CLICK 'STOP' TO STOP OPTIMISING.

Predict Numbers
---------------
This is still in the experimental stage.
After weighting the balls and sorting them based on weight:
Head: Selects the top x number of balls.
Tail - is still under development at this stage, to be used for selecting balls from the previous draws.
Once you have the number of balls selected that you want, on the right, you select how many subsets of the wheel you want, i.e. if you are using a 16 number wheel and you have 24 numbers, By selecting a subset value of 2, you will get 2 wheels of 16 numbers from the 24 but with nearly all 24 numbers used.

Best reset, is used when the optimising section sees that it's not getting anywhere then starts again with the best wheel. grain, once again tells it how many checks to do, the higher the number, the finer the detail.
Click recalc to start the generation of your tickets.
Once the program has generated it's initial wheels/numbers, it will carry on trying to optimise the positioning and selection of the numbers, once it get's a better 'fit' value, it will discard the lowest valued subset.
YOU NEED TO CLICK 'STOP' to get out the optimising section or else it will carry on forever.

Probability, will give you a quick check how you numbers will fare.

Save will allow you to save your tickets in a text file.

Regards
Patrick
 

Springbok

Member
http://www.national-lottery.co.uk/player/files/Thunderball.csv

This will take you there to the download section. I separate the actual thunderball numbers into a separate file. The dates mave no consequence to me. At a later stage I will show you how to handle the actual thunderball number. I convert the csv file into a text file without commas. Go ahead and see what you make of it. Will be studying your instructions on your programme.
 

Sidebar

Top