Registered: Oct 2003
Posts: 31

I have written a CA generator and analyzer that attempts to predict lotto numbers based on previous winning numbers using CA rules.

I need access to a super computer because I keep running out of memory on my home PC.

Can anybody help me?

I have a attached the readme file from my project to give you an idea of what I have been doing.

Yes this is whimsy and folly but I'm having a lot of fun with it. And that is the whole point of a hobby isn't it?

Attachment: readme.txt
This has been downloaded 2764 time(s).

Report this post to a moderator | IP: Logged

Registered: Oct 2003
Posts: 38

One thing to note is that a CA with significant randomness will generate a large fraction of all possibile sequences. So for instance the center column of rule 30 will contain all past and future lottery winners in some form or other.

An extreme example of this issue is the person J. Schmidhuber, a rabid anti-nks-ite. He claims that Wolfram's ideas on fundamental physics is based on "his" idea that one can compute all possible universes using a simple program that is analogous to generating all possible sequences.

It is true that the universe exists somewhere in the list of all possible sequences. But that is a pretty vacuous statement. Primarily because it cannot be used to make predictions. Whereas if you have a model, it means you have an understanding of how a system operates - and are capable of making predictions about what will happen.

Concerning the issue at hand, i think its not possible for a cellular automaton to be a realistic model for lottery winners. There is just no motivation for it.

However, i do think its an amusing possibility that the lottery machine generates intrinsic randomness, much as may be the case with fluid turbulence. However, the fact that humans operate these machines means that randomness is always inserted from the outside, and that probably dominates the outcome.

__________________
Everything is an expression.

Report this post to a moderator | IP: Logged

Registered: Jun 2004
Posts: 1

I have access to the computer cluster at Lawrence Berkeley Labs (LBL) in California. There are over 6000 processors that may be run in parallel. Trust me, this cluster can crunch numbers fast! The entire operation can be controlled from a remote terminal, so we can use it while located across the country. I don't claim to know anything except a general knowledge of NKS, so you must do the brunt of the theory and program writing. I, however, cannot let you login to the cluster. I hope this is understandable. In other words, you will have to show me how to run your program, and I'll have to compile it and run it on the LBL cluster. Please contact me at brspurri@syr.edu. I am more than willing to help you because I think this would be an excellent learning opportunity for myself on NKS. -Brett

Report this post to a moderator | IP: Logged

Registered: Oct 2003
Posts: 31

Concerning the statement "that humans operate these machines means that randomness is always inserted from the outside", I have an alternate theory. Dr. Wolfram shows that we can see evidence of CA rules in the patterns on shells on the beach. These were at one time thought to be random. Extending the theory (which isn't necessarily valid but is fun to ponder) CA rules are the mechanisms behind nature's randomness. So the question isn't "does human influence mean it's truly random?" the question is "what natural phenomenon might be influencing this seemingly random event?"

To predict lotto balls I tried to imagine what would be unique about the instant a ball was committed to going down the chute. It had to be something unique in time and space. Then I realized that Dr. Carl Sagan had already solved this problem. The plaque on the Voyager spacecraft had to identify where and when it was launched, pulsar spin rates. Pulsars are declining in spin rotation. Dr. Sagan put on the Voyager plaques the speed of the nearest pulsars to earth at the time of launch. THis uniquely placed the time and location of Earth when the Voyager probes left.

Ok so my theory was if I could get the pulsar spin rates in terms of hydrogen transition units to a great enough precision that they could uniquely identify the time and space that the lotto ball was committed to going down the chute. It had to be the location and time of the draw site. Then I could find the CA rule that matched the ball that came out and voila, the lotto was mine for the taking. I abandoned this theory when I realized that I just had to search the CA output for the numbers and look for the pattern. Who cares what pulsar declination spin rate crosshairs started it?

The point of the theory was that if CA rules are natures way of generating randomness then we need to find the natural events that might be related to the random part of nature that we are trying to predict. Just because humans are involved doesn't mean it is unpredictable. We just don't understand the implications of CA rules in nature fully yet.

"People say I'm crazy, but I have a good time." - Joel Walsh in Life's Been Good To Me So Far.

Report this post to a moderator | IP: Logged

Registered: Oct 2003
Posts: 31

Ok! Sounds great! I will zip up my Lotto predictor and post it to this site. It's not workable right now because I am in the middle of switching out as many "long" declarations for "short" ones in an effort shrink the footprint. The limitation I am introducing is that it will be limited to 32,767 rows of CA output that can be referenced. However, that should be more than enough to find some good lotto number possibilities. I am finding it takes no more than 500 rows to find all the numbers.

I'll try to post it this weekend.

Report this post to a moderator | IP: Logged

Registered: May 2004
Posts: 27

I worked as a Keno writer at the Golden Gate casino, the Golden Nugget Casino, and Bob Stupack's Vegas World Casino when I was in my early 20's (I'm now in my late 40's), and I became fascinated with the game.

One of the things that I noticed is that the geometric texture of the patterns that would show up on the Keno board would repeat over many games. For example, if blockish patterns show up, you will see them over many games. If spread-out patterns show up, you'll likewise see spread-out patterns over many games. If horseshoes show up, horseshoes. If lines show up, lines. Etc.

I lack the math to figure it out, but I nonethless think that I understand why it occurs:

There's a natural limit on true randomness during the shuffling of a finite number of physical elements in a closed system. Most simply, one Keno ball cannot magically pass through any others in its way. Therefore, a general texture or matrix relationship will persist because of this natural physical limit on true randomness

Said otherwise, pure randomness is a mathematical concept that's physically unattainable in a closed universe of interrelated, finite, physical elements.

A very, very intelligent lady by the name of Ionia Dumitriu at M.I.T. may have by now finished her planned doctoral thesis on "random matrixes" which she intends to solve using an approach with linear algebra, which I understand is a method of finding the underlying rules of a pattern so as to map out future patterns according to the same rules.

My disagreement with the approach, from my mathematically unsophisticated point of view, is that the rules themselves are in flux in a loose system like Keno (or lotto), etc. Math often gets applied in an all-or-nothing sort of way, which ignores reality and perhaps misses opportunities, and I can only hope that Ms. Dumitriu will bring something brilliant to the table when she's done cooking the numbers.

Very Respectfully,
Ray Donald Pratt

Report this post to a moderator | IP: Logged

Registered: Oct 2003
Posts: 31

As promised, here is my Lotto number predictor based on CA. I have included all of the source code, the input files and I have updated the readme.txt to include a roadmap through the functionality and a few caveats.

It is currently set up to run on a larger system than mine and generate good output. That is, I have set the internal storage beyond what my PC can handle but large enough that it will successfully run to conclusion.

The megami.txt contains the latest drawn numbers. To keep the file up to date consult www.megamillions.com They have a video replay of the previous 20 drawings. Be sure to load the numbers in the order drawn, it is crucial to the theory.

Please respect my copyrights. If you want to make modifications to this code, leave my name on as the author of version 1.0

Attachment: lottocapredictor.zip
This has been downloaded 2753 time(s).

Report this post to a moderator | IP: Logged

Registered: May 2004
Posts: 27

Not long ago, I learned a painful lesson about sharing gambling secrets on a public board, and I want to share that experience with anyone who actually discovers an advantage in any game.

I had been losing while playing Basic Strategy in blackjack against 4-deck continuous shuffling machines, and it bothered me that I was becoming so familiar with the losing patterns that I could feel when and exactly how I was going to lose, but I could not bring myself to depart from Basic Strategy and streak-playing and money management techniques common to gambling.

One evening, however, I noticed something strange about how the dealer was shuffling and loading new decks of cards into the machines. Each of the four decks was shuffled just one time and then placed in a stack until all four decks were in the stack, and then the stack was loaded into the machine.

I then realized why I was experiencing long runs of small cards and then bursts of high cards: The shuffling machine was altering the cards very little from the very orderly patterns loaded into it, and the number of players affected and altered how the cards were falling until the cards were falling into long runs of low cards and bursts of high cards.

After paying attention and getting a feel for the pattern, I would wait until a run of low cards would culminate in one or more aces. At that point, I knew from observation that 10-valued cards would come thickly. I also determined that the best way to capitalize on the situation, when the ace or aces showed up, would be to be playing first base and to spread out to two hands if there were less than four players, and especially if I was playing heads up against the dealer. This would lessen the possibility that the dealer would get an equal or better hand with the 10-valued cards coming out thickly.

I would wait for the pattern and then pump my $5 bets up to $25 or $50 and wack 'em. If the pattern failed, I would immediately leave the table. (The advantage proved false.)

I recouped about $2,000 of previous losses with this method, but one day I got in a stinking math argument with some educated idiot visiting the John Patrick website, and I confronted his arguments with what I was doing and experiencing.

The very next day, the casinos had completely changed their shuffling practices.

NEVER AGAIN WILL I EVER REVEAL ANY SPECIFIC DETAILS OF AN ADVANTAGE AT GAMBLING ON A PUBLIC WEBSITE. NEVER, EVER AGAIN.

The shuffling practice has never returned, and although I have plans in mind for learning to identify other favorable and unfavorable patterns, I will never publish any successful method.

It's just a word to the wise.

Very Respectfully,
Ray Donald Pratt

Last edited by Ray Donald Pratt on 06-21-2004 at 08:56 AM

Report this post to a moderator | IP: Logged

Registered: Oct 2003
Posts: 31

Good advice. While I would love to win a multi-million dollar lottery game, I do not expect to. So far I have not spent one dollar on a ticket my predictor has output. I just love the challenge of trying to beat the game. It's a fasicinating hobby.

As to your point about publicly airing the supposed flaw I am trying to exploit. Yeah, you're right I probably should not have explicitly laid it out. I suppose the lottery commisions could change the way the game is played to foil me. However, I would consider that an honor. To think that someone thought I was actually close enough to have to change the rules would be a huge compliment. I'd feel like the '77 Pittsburgh Steelers. Their defense was so good the league had to change the rules to make the game competitive again.

Report this post to a moderator | IP: Logged

Registered: May 2004
Posts: 27

I respect your intellectual and moral purity in striving to improve mankind's understanding of presupposedly random events.

However, having had the taste of blood from advantage, I can assure you that I'd rather get PAID than be immortalized and honored with a statue, praised with an "attaboy," and given absolutely nothing of substance for my trouble.

And besides, what could be more moral than taking away as much of the mob's legitimized income as possible before being fully exposed?

Very Respectfully,
Ray Donald Pratt

Report this post to a moderator | IP: Logged

Registered: Nov 2006
Posts: 5

What you call ‘new science’ is actually old shoes! I’ve been in the gambling mathematics since 1988! Sorry, guys, I know scientists have big egos! That’s good, if a big ego is accompanied by high self-esteem. The worst a human can do is mixing a big ego with a very low self-esteem.

You can read a lot of gambling and lottery theory at my website:

www.saliu.com

Not only that, but you can also download a lot of free software. A lot means a lot! You can get gambling software, lottery software, probability software, combinatorics software…plus some other things. All for free!

If you want to generate truly random and unique numbers (including lotto), I offer free source code for the Power Basic compiler (PBCC). It is lighting fast and works with huge arrays. Your case of 32K is a limitation of interpreters. The EXT data type in PBCC works with numbers up to 18 digits in size. The source code is here (including a compiled program that mimics the delay of lotto drawings conducted by humans):

www.saliu.com/random-numbers.html

Best of luck!

Ion Saliu,
Programmer At-Large
www.saliu.com/infodown.html
(The entry point to the free downloads)

Report this post to a moderator | IP: Logged

Registered: Nov 2006
Posts: 5

You want to generate all sequences of C(N, M), or permutations N, or arrangements A (N, M), or the largest possible sets – Exponents (N, M)? My free program PermuteCombine.EXE does it easily and very quickly.

There is the reverse to it. Find the rank or index of a sequence, or generate the exact sequence for a given index. For example, the sequence #1000000 in a lotto 6/49 game is represented by the combination:
1, 9, 14, 26, 31, 37

The program that does such calculations is also free: LexicographicSets.EXE

Presentations of sets generating and indexing:

www.saliu.com/permutations.html
www.saliu.com/lexicographic.html

Best of luck!

Ion Saliu,
Pacesetter At-Large

Report this post to a moderator | IP: Logged

Registered: Nov 2006
Posts: 5

Ed Meier et al:

I admit I had a hard time figuring out what CA stood for. First, I thought of ‘computer algorithm’. I googled and I was (mis)directed to ‘California’, ‘Canada’, ‘chemical associations’, etc.

Using abbreviations as if complete words points to religious behavior. I can see now that NKS is more like NKR: New Kind of Religion! I hope you take it as man — or as men! It’s not about gender, but about Reason…

I too use quite often an abbreviation: FFG. But I always try to start with ‘Fundamental Formula of Gambling (FFG)’.

Finally, I decrypted your ZIP. The .java files don’t run directly. Probably I need a java compiler. I don’t have it and don’t want one. The performance drives me nuts!

First off: CA stands for Cellular Automata. A new god, looks like!

I read the contents of your files. I am not a Java programmer, but I was able to understand your code.

What you call ‘spacers’ are better known as ‘deltas’. One can sort a lotto drawing in ascending order. That is what ‘combination’ is all about. If you work with ‘drawn order’, it is called ‘arrangements’. I use combinations here. You can find at least 5 deltas in a lotto-6 combination. For example, a lotto drawing such as:
3, 13, 20, 24, 33, 49
can be represented by the following deltas:
10, 7, 4, 9, 16

There is a multitude of deltas possibilities. My free program ‘MDIEditor And Lotto WE’ has a filter named Ion1. It is based on deltas, although more complex than what I presented above.

Delta is just one filter. It is never sufficient. One filter is way too little. I don’t know why you think you need a supercomputer to run your algorithm! It is just one filter, bud! I have console (command prompt) software that employs lots of filters. The 5-deltas filter I told you above reaches values of over 400,000. Yes, four hundred thousand “drawings” (combinations). No, you don’t have that many drawings in real life. You need to simulate lots of combinations to improve the probability of winning the lotto games. I can tell you this. That one 5-deltas filter is more complex. In 70% of the situations, no 5-deltas repeat in 400,000 combinations. So, the software can eliminate all 5-deltas possibilities. My software is compiled with Power Basic Console Compiler (PBCC). It is lightning fast. I can generate all the qualifying combinations (based on 5-deltas) in some 20 minutes. I run a 733 MHz PC — a far cry from a supercomputer! There are a few thousand qualifying combinations. Your figure — hundreds — is erroneous.

You are right in one aspect. Lottery can be analyzed as a branch of mathematics. Some say: “Lottery is random. You can’t analyze it!”

But what is non-random, super crocodilule and reasonable colleague of mine? Everything is RANDOM. The Universe is random. If it is not random, it must be ordered. It must follow a Plan. A Plan requires a Planner! That would be what humans have called god(s) for millennia. God’s fundamental attribute is Absolute Certainty. Mathematically, Absolute Certainty represents Absolute Absurdity. There is a formula that proves that absolute absurdity. I call it FFG: The Fundamental Formula of Gambling.

N = log(1 - DC) / log(1 – p)

where N is the numbers of trials for an event of probability p to appear with the degree of certainty DC. For example, it takes 7 coin tosses to get ‘heads’ with a degree of certainty equal to 99%. If you play blackjack, as your good colleague Ray Donald Pratt does, you noticed that you could lose 7 consecutive hands now and then. Ray Donald Pratt is a casino guy — yet he lost 10 hands in a row several times! I know, he won’t admit it immediately! I know well the casino guys…

Some people started to call FFG the ‘formula of god’. Sometimes I call FFG the ‘Fundamental Formula of TheEverything’.

You can read these materials (at least):

The Fundamental Formula of Gambling (FFG)
Mathematics Of The Fundamental Formula Of Gambling
Theory of Probability: Best introduction, formulae, algorithms, software
Almighty Number, Randomness, Universe, God, Order, History, Repetition

The founder of this NKS is right about one thing, for sure. Computer software can simulate reality, real phenomena. Matter of fact, there are situations when humans would “kill” one another because of opposing ideas. Only the unbiased computer can solve the fighting. Computer software can validate or invalidate ideas better than the biased humans.

As you can see, your lotto simulation software needs a whole lot more elements, if you want just hundreds of combinations in the output.

”Bright software I write —
It turns the night into light,
Explores deeply underground,
To the square root of sound.”
(Creativity song inspired by “Liru, liru, crocodilu II”).


Ion Saliu,
C-Automatically At-Large

Report this post to a moderator | IP: Logged