Lottery Sorcerer trial run

[PAB]

Hi lottoarchitect,

I have just picked out a couple of comments from your previous posts:-

The lottery does not follow statistics or observable patterns in first place.

Actually, they are not completely random, lottery draws performed in a barrel have reduced randomness. I have designed a program that examines this specific issue and tries to take advantage of it. So far the results are quite promising given the wins made by the users and from my end, based on their results, I can say I have serious and strong clues that indeed lottery draws ARE NOT PURELY RANDOM (which can't be proved by any formal method however) and there is a way to extract this reduced randomness they embed so to produce even better number sets than pure luck.

Does your first sentence not imply Randomness?
Out of interest, where did you get your data from regarding Lotto draws performed in a barrel have reduced randomness?
If you could elaborate a bit more please on how you extract this reduced randomness they embed so as to produce even better number sets than pure luck that will be interesting.

Actually I expect over the very very long run of lotto draws these to behave statistically as pure random events i.e. eventually all the combinations will be drawn approximatelly an equal amount of times. Of course we'll need way more test events than the total possible combinations of our lotto game to verify the pure randomness of this particular experiment but I hope we agree on that aspect, that we expect overall the results to show pure random behavior.

At some point of a very huge set of test events (draws), even these observations would occur and still this will be considered a normal purely random outcome. We can't conclude anything at the moment about the pure (or not) random nature of lotto draws simply because we don't have somehow such a huge history (multiple times the total combinations of the experiment) to evaluate.

In another thread, I did actually run 96,153.85 years worth of draws at two draws a week, this experiment equated to 10,000,000 ( 10 MILLION ) draws in total and calculated the results for the numbers 01-49 for the total times each of them were drawn along with their respective percentage.

The MINIMUM value for a drawn number was 1,223,016 times.
The MAXIMUM value for a drawn number was 1,226,040 times.
The RANGE from MINIMUM to MAXIMUM value for a drawn number was ONLY 3,024.

The 3,024 difference is miniscule in the overall concept of such a large test experiment.
All that this seemed to prove was that over a substantial period of time for that particular test was that everything leveled out to be on a virtually level par.

This thread has turned out to be very interesting.

Regards,
PAB


-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏
12:45, restate my assumptions.
(1) Mathematics is the language of nature.
(2) Everything around us can be represented and understood through numbers.
(3) If you graph the numbers of any system, patterns emerge. Therefore, there are patterns, everywhere in nature.

 

[lottoarchitect]

Hi PAB,

...The lottery does not follow statistics or observable patterns in first... place.
I have just picked out a couple of comments from your previous posts:-
Does your first sentence not imply Randomness?

No, statistics is a way to evaluate data within confidence intervals and being humanly understandable to judge an experiment. Data can look random when viewed at a distance and not so random when viewed "locally". Statistics can't evaluate the locality because by design require a quite larger data set to be applicable. So even if statistics suggest something behaves randomly, this doesn't forbid non-random behavior locally (a much smaller subset). Statistics simply can't trap this.
Similarly, a pattern is a very well defined/strict behavior. Pure randomness does not have patterns obviously. Reduced randomness does not have very well defined behavior. In fact, it has a "loose" behavior of a pattern. Therefore this type of pattern is not observable by humans or simple methodologies.

Out of interest, where did you get your data from regarding Lotto draws performed in a barrel have reduced randomness?

My engine and the results produced suggest that.

If you could elaborate a bit more please on how you extract this reduced randomness they embed so as to produce even better number sets than pure luck that will be interesting.

I'll not because this novel design is what I sell. Suffice to say, it has never been attempted before, I may also patent it.

In another thread, I did actually run 96,153.85 years worth of draws at two draws a week, this experiment equated to 10,000,000 ( 10 MILLION ) draws in total and calculated the results for the numbers 01-49 for the total times each of them were drawn along with their respective percentage.

The MINIMUM value for a drawn number was 1,223,016 times.
The MAXIMUM value for a drawn number was 1,226,040 times.
The RANGE from MINIMUM to MAXIMUM value for a drawn number was ONLY 3,024.

The 3,024 difference is miniscule in the overall concept of such a large test experiment.
All that this seemed to prove was that over a substantial period of time for that particular test was that everything leveled out to be on a virtually level par.

It is hard to believe you have done by hand that many draws using a system similar to what is being used by the commission. So I suspect you produced a PRNG system. If this is what you have done, it lacks those physical elements and factors affecting the normal draws in a barrel. Statistically both may overall look random, no dispute at that, however the locality behavior is what we look for. Draws in a barrel do have this behavior more or less. I can't tell what is happening for any other system however.

 

[PAB]

Hi lottoarchitect,

...a pattern is a very well defined/strict behavior. Pure randomness does not have patterns obviously.

I disagree with this (see number three in my signature) because if you graph the numbers of ANY system, including RANDOM numbers, patterns WILL emerge, maybe not at instantly recognisable intervals or structure, but they WILL be there.

It is hard to believe you have done by hand that many draws using a system similar to what is being used by the commission. So I suspect you produced a PRNG system. If this is what you have done, it lacks those physical elements and factors affecting the normal draws in a barrel. Statistically both may overall look random, no dispute at that, however the locality behavior is what we look for. Draws in a barrel do have this behavior more or less. I can't tell what is happening for any other system however.

I obviously coded the experiment to produce the results, which you say, "lacks those physical elements and factors affecting the normal draws in a barrel", unfortunately we only have your word and opinion on this because you are unable to share the true concept of your system with us, fair enough.
It's almost as if you are saying that the Lotto is fixed and predictable, which is obviously NOT true.

I think this has probably exhausted this particular avenue of discussion because I can understand why you do not want to give too much away with regard to your system, and to be honest, that is what is needed to enhance and backup your theories because I think there may be quite a few following this thread who are not convinced by your arguments without hard facts and proof.

Out of interest, using your system, what is your percenatge return against your outlay per draw/week, I would assume it is quite high if as you say, your system has honed in on the numers to play within the "Locality" method?

Regards,
PAB


-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏
12:45, restate my assumptions.
(1) Mathematics is the language of nature.
(2) Everything around us can be represented and understood through numbers.
(3) If you graph the numbers of any system, patterns emerge. Therefore, there are patterns, everywhere in nature.

 

[lottoarchitect]

Hi lottoarchitect,


I disagree with this (see number three in my signature) because if you graph the numbers of ANY system, including RANDOM numbers, patterns WILL emerge, maybe not at instantly recognisable intervals or structure, but they WILL be there.

I said something at the previous posts, what a human interprets of something he observes, it doesn't mean it is there. Your example of finding patterns in pure random events is unfortunately such a case. You believe you observe some pattern but in fact there isn't any. Existence of a pattern suggests there is some sort of logic/structural mechanism in the produced results, therefore the results cannot be purely random. So, in any pure random experiment, we cannot have patterns, these are mutually exclusive. We might think we see patterns, but this is our interpretation, it doesn't mean they are there. To be honest however, I don't believe we have completely pure random events in nature. However I have to rule out your 3) at your signature as wrong judgement in the case of pure random events. Reduced randomness on the other hand exhibits some sort of patterns, "loose" ones.

I obviously coded the experiment to produce the results, which you say, "lacks those physical elements and factors affecting the normal draws in a barrel", unfortunately we only have your word and opinion on this because you are unable to share the true concept of your system with us, fair enough.
It's almost as if you are saying that the Lotto is fixed and predictable, which is obviously NOT true.

Not sure how you concluded I suggest lotto draws are fixed and predictable (at least the way you mean it). Fixed? Absolutely not! Predictable? Up to an extend yes, they embed some sort of information which can be extracted and used to improve natural odds e.g. by a factor 2x, 3x 10x etc. The only reason I suggest lotto draws in a barrel have reduced randomness is due to the way these draws are performed. Whatever these factors might be, they contribute to the result of having reduced randomness.

I think this has probably exhausted this particular avenue of discussion because I can understand why you do not want to give too much away with regard to your system, and to be honest, that is what is needed to enhance and backup your theories because I think there may be quite a few following this thread who are not convinced by your arguments without hard facts and proof.

Out of interest, using your system, what is your percenatge return against your outlay per draw/week, I would assume it is quite high if as you say, your system has honed in on the numers to play within the "Locality" method?

True, that's why I let users talk for this with actual hard facts of their own. The wins they produce suggest a good improvement of odds. I'll give some examples: a user plays a 6/49 lotto in Australia and picks 12 numbers every time. He managed with this system to predict 4 times in a row a 4 match. The natural chance for such a sequence of events to occur is around 1/3 million. Of course you can attribute this to pure luck.
Another user plays his German 6/49 lotto. Again he uses 12 numbers in his playing strategy. Within 4-5 months he managed to match twice a 5 hit. The natural waiting time for a 5 hit is 9 years (when picking 12 numbers) for this particular game. We had twice that event in 4-5 months. Again of course you can attribute this to pure luck.
Another user plays the Greek 5/45 + 1/20 (PowerBall) game. He runs the system for about 3-4 months now. He uses the system in a more advanced approach (using matrix constructions) with results obtained by GAT. GAT produced a 5-hit for the main field in two groups and also found the PB at the same draw. So after about 30 draws of actual GAT usage (since the user bought the system), and around 20 euros of cost of play with his matrix construction he matched a 4+1 hit which delivered 27000% profit. Again of course you can attribute this to pure luck.
Several other examples which are posted at my forum by the users (including the above), most if not all, suggest odds improvements. Of course even all these you can attribute to pure luck. However there are many such experiences of favorable results which make it hard to believe they occurred due to pure luck. So even if I have just a few hundred users so far, several have achieved outstanding results. So the outcome is we can have odds improvements in lotto draws. These results achieved by utilization of GAT engine and posted by the actual users.

 

[Icewynd]

Anecdotal evidence is all well and good, but of course people don't generally take the time to report poor results. I think what PAB had in mind was to have you post some predictions for any lottery of your choice so that we could get an idea of the usefulness of the "locality" analysis.

I don't think that doing so would in any way compromise your proprietary program, which we all understand has taken considerable effort to develop, market and maintain.

 

[lottoarchitect]

I don't try to convince anyone for anything. I'll not enter that challenge because I have the users who provide the hard evidence needed - just take the time and read what they say. Since they win (quite) more than naturally expected, no need to do anything myself beyond that to prove anything to anyone.
Furthermore, you can backtest the system yourself in your favorite lotto game using the compare feature. Just run it after a draw is made without entering that draw in the history and check what the program delivers. Just make sure you understand how the system works. No need to buy it to test it that way. Because I know you'll see the same favorable results more or less, that's why I offer this test capability so to appreciate this new novel prediction without displaying actual predicted numbers. So as you can see, absolutely no need from my end to do any test runs to show results of this engine. I don't have the time anyway.
Regarding poor results, indeed some users haven't won something to notice so far. I can't guarantee there will be wins for every game everywhere - actually I can't guarantee wins to begin with. This depends on the lottery game in question. If it has quite reduced randomness, many good wins will emerge. The best games so far for that engine are the Australian ones and the German one where most hits have been produced. The Greek one is tougher for the engine so I suppose this has lesser reduced randomness. However, if let's say 5% of the existing users (might be more, haven't counted them) have produced quite better results than odds suggest, beating chances much worse than 1/1000 in each case, then can you say "it is just luck"? I doubt.

 

[PAB]

Hi lottoarchitect,

I think what PAB had in mind was to have you post some predictions for any lottery of your choice so that we could get an idea of the usefulness of the "locality" analysis.

I'll not enter that challenge because I have the users who provide the hard evidence needed - just take the time and read what they say. Since they win (quite) more than naturally expected, no need to do anything myself beyond that to prove anything to anyone.

That's why I offer this test capability so to appreciate this new novel prediction without displaying actual predicted numbers. So as you can see, absolutely no need from my end to do any test runs to show results of this engine.

That's a shame really because you could have backed up your comments and theories, but I appreciate that time is limited for you at the moment.

I have done some research into your comments and theories, and to be honest have found it quite interesting.
Although this is mainly computer based, the principle is the same, there is an article here...

http://en.wikipedia.org/wiki/Locality_of_reference

...that discusses locality of reference, which describes some of the basis of your comments in previous posts and is accredited to the area of computer science, which I believe is what you are studying for a masters degree in at the moment, hence the fact that you have no time to make improvements to your Lotto software or for further releases?

I also did a Google search and came across several articles that talk about locality of reference & principle of locality. Is this what you are thinking of patenting or is it a variation of this based along the same sort of lines?

There is an interesting snippet of an article here...

http://www.readcube.com/articles/10.1038/nature09008?locale=en

...but unfortunately you have pay for the full pdf to get the rest of the article.

In the article...

http://www.ams.org/samplings/feature-column/fcarc-random

...one paragraph says...

"Some readers may remember the draft lottery used during the Vietnam war. In 1969, the days of the year were ordered by drawing pieces of paper, each with a day of the year written on it, from a bin. Young men born on those days drawn first were more likely to be drafted. A statistical analysis later found that, because they were placed in the bin later, days toward the end of the year were more likely to be drawn. In other words, the days were not ordered randomly, and a more mathematical solution was proposed for subsequent years' lotteries."

I also found...

http://www.random.org/analysis/

...which was quite interesting reading.

I assume that you have covered all these and many, many more, not only for your degree but for your Lotto software.
I would love to spend more time on investigating the theories of whether there is or there isn't such a thing as true randomness, whether it be in a Lotto or in nature itself, but I feel that from what I have read up to now that it would be very time consuming, and anyway, I believe the jury is still out on that one.

Anyway, good luck with your degree.

Regards,
PAB


-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏
12:45, restate my assumptions.
(1) Mathematics is the language of nature.
(2) Everything around us can be represented and understood through numbers.
(3) If you graph the numbers of any system, patterns emerge. Therefore, there are patterns, everywhere in nature.

 

[lottoarchitect]

Hi PAB,

my backup is what the users suggest. I let them talk for the capabilities so I really don't understand why you say "not backed up". If actual users win more than expected, this is already the best "proof" I can have of this concept. In fact, I deliberately don't post any predictions myself because what matters is what the users say themselves, not me. If they win more than naturally expected, it is more than sufficient to me to suggest it is possible to predict above odds.

As for the theories you post, surely they'll have something in common to what I say (computed related article has nothing to do with it however). In fact a specific article on those you posted suggests what I say, pure random events aren't there, only at the quantum level possibly. As far as I know, noone ever attempted my approach to the locality, not sure what happens in a secret lab underground but so far I don't know of any implementation of my approach. Actually, even if you read those articles, you'll not see any actual implementation or even clues on how to approach it; they discuss on properties of this supposed randomness but nowhere you'll find any actual way to extract meaningful data to take advantage of those properties. So I retain the novel, unique and never attempted before status of my engine and because of that uniqueness I consider patenting it. By the way, I have completed my degrees a few years ago.

regards
lottoarchitect

 

[PAB]

Hi lottoarchitect,

I must admit that I have found this discussion quite interesting and thought provoking. As I said previously, I would love to spend more time on investigating the theories of whether there is or there isn't such a thing as true randomness, whether it be in a Lotto or in nature itself. I always like to try and fully understand to the best of my ability both sides of a discussion in order to make an informed and logical decision on it, this is such a case. I might keep nibbling away at this when time permits.

Thanks for sharing your views and concepts with us, I am sure those following this thread are finding it interesting too and I wish you all the best if you decide to go ahead and patent your system.

I must admit that I still believe that the Lotto is a random event, that Lotto balls have no memory, and that any combination has as much chance as any other combination of being drawn. But that said, my opinion might change in the future if it can be irrevocably proved to the contrary.

I think that anyone who gambles, whether it's the Lotto, Horse Racing, Dog Racing etc has a system to which they adhere to for their bets. For the Lotto, I personally use statistics and patterns (or lack of) derived from past results in an effort to try and give myself a slight edge and hopefully a better than pure luck chance of winning a prize.

Regards,
PAB


-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏-∏
12:45, restate my assumptions.
(1) Mathematics is the language of nature.
(2) Everything around us can be represented and understood through numbers.
(3) If you graph the numbers of any system, patterns emerge. Therefore, there are patterns, everywhere in nature.

 

[lottoarchitect]

Hi lottoarchitect,
I must admit that I still believe that the Lotto is a random event, that Lotto balls have no memory, and that any combination has as much chance as any other combination of being drawn. But that said, my opinion might change in the future if it can be irrevocably proved to the contrary.

Just to add my point of view to these too. Indeed, lotto balls do not have memory. This is also the reason any attempt to analyze lotto games using statistics/hit charts etc that rely on the assumption lotto balls do have memory fail. Reduced randomness (the locality) doesn't need numbers to have memory because the real data is not what each individual number does and the life span is much shorter anyway and what we try to trap here. So really, attempting to predict using the actual numbers as the starting point is already the root of the failure for any approach that rely on that assumption.
Also to add at the examples I posted above, the same user who plays that 5/45+1/20 game, he managed within one month to get twice the 5 hit (no PB this time) in his numbers selection. I was recently informed of this new win.