Texas United States Member #1477 May 7, 2003 2311 Posts Offline

Posted: August 23, 2010, 9:34 pm - IP Logged

Quote: Originally posted by jwhou on August 23, 2010

I calculated that if I bought five lines a draw in the Texas Two Step, the median time to win a jackpot would've been 2,400 years.

I also had a random number generator that was structured so that a minimum of the lines you bought in a given draw overlapped hence giving you the maximum chance at the small prize that your money could buy, it also did MegaMillions, PowerBall and numerous others. Interesting thing is that though it was very random if you told it to generate a thousand numbers, it wasn't so random if you started it 1,000 times to generate one number each time even though the seed for the RNG was taken by two calls to the system RNG generator. There were two specific number combinations that kept popping out far more than a standard deviation above the others, sorta a fractal attractor. Historical data shows that those two number combinations didn't have an advantage over any others but maybe in the games with RNG's...

I noticed the attractors when I added Lotto Max to the scripts repertoire.

Big difference between your generator and mine is that mind is a "Controlled" generator. I set filters and it only generates specific type of combinations. When I pick numbers I am looking for something very specific. Comes from looking at the past 31 draws and being able to eliminate certain numbers from play.

Tonight I don't think anything above 28 will play. First number will be Odd. Bonus Ball will involve digits 0-1-2 or 3 which would eliminate 20 numbers.

There are 53,900 possible four number combinations. Last Thursday I cut them to 296.

United States Member #51269 April 3, 2007 529 Posts Offline

Posted: August 24, 2010, 2:19 am - IP Logged

Quote: Originally posted by jwhou on August 23, 2010

What a down payment on the national debt?

When the Spanairds brought back all that gold from South America, the bottom fell out of the gold market and they didn't make nearly as much per trip as they had promised the investors.

You don't want to be bringing all that iron, nickel, cobalt and platinum back to Earth. But you do want to build space habitats with them and have the people come out to the habitats. It's all about real estate.

Texas United States Member #1477 May 7, 2003 2311 Posts Offline

Posted: August 24, 2010, 2:44 am - IP Logged

Was slightly off on the Two Step tonight but still in the ballpark. Really thought a small Odd would be first number. Played 3 and 5......4 played :( Played 12-18 and 17 for Teens and 18 played. Then played 27 for 20's and 25 played. Then dropped to 30's where used 32 and 31 played :(

Had cut the Bonus Balls down to 11 possibles with digits 0-1-2 or 3 involved. Bonus Ball was 2 My favorites were 1-10 and 20. 20 Sum Totals 2 :(

United States Member #83701 December 13, 2009 225 Posts Offline

Posted: August 24, 2010, 2:48 am - IP Logged

Quote: Originally posted by Clipper on August 23, 2010

Big difference between your generator and mine is that mind is a "Controlled" generator. I set filters and it only generates specific type of combinations. When I pick numbers I am looking for something very specific. Comes from looking at the past 31 draws and being able to eliminate certain numbers from play.

Tonight I don't think anything above 28 will play. First number will be Odd. Bonus Ball will involve digits 0-1-2 or 3 which would eliminate 20 numbers.

There are 53,900 possible four number combinations. Last Thursday I cut them to 296.

The arguments for filtering tend to be rather vague as to causality. i.e.: any particular physical reason for believing numbers above 28 won't play? or why would the first number be odd? Why would the bonus ball involve digits 0-1-2 or 3?

People tend to filter out number sets that had already been drawn on the theory that once drawn it's unlikely to be drawn again. Well did you know that in the 2,734 draws since 1982 in Canada's Lotto 6/49, 18 number sets had been drawn twice each?

Did the numbers drawn last Thursday hit in your 296 field? Close doesn't count in lotto, there's no such thing as just one number off.

The idea with my script was to maximize the number of combinations covered for subsidiary prizes, those can be calculated and doesn't require the draws to know about previous draws and doesn't rely on any belief that certain numbers are more likely than others to come out. It did "control" it's selection versus the numbers it had already selected so that each line selected did not duplicate subsidiary combinations too much (after a certain number of lines, some duplication is unavoidable). It does a good job of optimizing for winning a small prize but winning a small prize excludes winning any other prizes even other small prizes so I place little value on it's use unless you want to just say I've won one of the small prizes without mentioning how much you've spent on the tickets.

I really don't want to get into a mine is better than yours crap because that's all just based on individual theories often loosely or not at all substantiated, I was just saying I was doing heuristic number generation not too long ago and noticed a few things. So mine selected the numbers according to sound laws of probability, combinations and permutations focusing on coverage while yours based them upon historical patterns focusing on prediction without rationale, so what.

What I found odd when working on the generator is that there were fractal attractors, certain number combinations that came up more frequently then they should, this didn't happen so long as the number generation was within my own RNG but happened when I seeded from the system RNG (which also meets rigourous tests for randomness) for each set of numbers and was unexpected due to the size of the internal state representation versus the range of the numbers being produced. Thing is, if the lottery commission fires up their RNG program for a draw and then shuts it down between draws, they are reseeding it each time they fire it up but if they are only checking the chi square on a continuos run then they may be missing a similar behaviour. Of course, this is another reason to prefer mechanical draws.

Texas United States Member #1477 May 7, 2003 2311 Posts Offline

Posted: August 24, 2010, 3:59 am - IP Logged

Quote: Originally posted by jwhou on August 24, 2010

The arguments for filtering tend to be rather vague as to causality. i.e.: any particular physical reason for believing numbers above 28 won't play? or why would the first number be odd? Why would the bonus ball involve digits 0-1-2 or 3?

People tend to filter out number sets that had already been drawn on the theory that once drawn it's unlikely to be drawn again. Well did you know that in the 2,734 draws since 1982 in Canada's Lotto 6/49, 18 number sets had been drawn twice each?

Did the numbers drawn last Thursday hit in your 296 field? Close doesn't count in lotto, there's no such thing as just one number off.

The idea with my script was to maximize the number of combinations covered for subsidiary prizes, those can be calculated and doesn't require the draws to know about previous draws and doesn't rely on any belief that certain numbers are more likely than others to come out. It did "control" it's selection versus the numbers it had already selected so that each line selected did not duplicate subsidiary combinations too much (after a certain number of lines, some duplication is unavoidable). It does a good job of optimizing for winning a small prize but winning a small prize excludes winning any other prizes even other small prizes so I place little value on it's use unless you want to just say I've won one of the small prizes without mentioning how much you've spent on the tickets.

I really don't want to get into a mine is better than yours crap because that's all just based on individual theories often loosely or not at all substantiated, I was just saying I was doing heuristic number generation not too long ago and noticed a few things. So mine selected the numbers according to sound laws of probability, combinations and permutations focusing on coverage while yours based them upon historical patterns focusing on prediction without rationale, so what.

What I found odd when working on the generator is that there were fractal attractors, certain number combinations that came up more frequently then they should, this didn't happen so long as the number generation was within my own RNG but happened when I seeded from the system RNG (which also meets rigourous tests for randomness) for each set of numbers and was unexpected due to the size of the internal state representation versus the range of the numbers being produced. Thing is, if the lottery commission fires up their RNG program for a draw and then shuts it down between draws, they are reseeding it each time they fire it up but if they are only checking the chi square on a continuos run then they may be missing a similar behaviour. Of course, this is another reason to prefer mechanical draws.

To answer your question.....yes the combination that played was among the 296 as was the combination that played tonight out of 20 possible combinations generated once I knew the kind of game that would probably appear. I don't filter out "number sets" either I don't need to go back more then 31 games to figure out what is about to happen. Anything beyond that is not necessary.

Yes, there was physical reason for believing numbers above 28 would not play. Simple fact, determination there was high probability that only 3 Denominations out of 4 would be used and by tracking what Denomination would not play it indicated 30's would not play if only 3 Denominations appeared. If anything played higher then 28 there was indication 32 would play.....31 played. I ruled out all 33-34 and 35.

Why would the bonus ball involve digits 0-1-2 or 3? I track the Bonus Ball separately using personal Pick 3 software . It indicated to me any digit used would come from within the last 4 games which eliminated 4-5-6-7-8-9. The Bonus Ball was 2.

Why would the first number be Odd? I track Alpha Patterns and there was indication the first two numbers would go Odd-Even. It went Even/Even. I played 3 and 5 (but not together)......4 played.

I think that covers your questions. I've tracked for 15 years. Been working on the Two Step for the past 6 months. Have had 3 Numbers of the 4 several times and on other occassions the Bonus Ball several times. One night I had the BB on 11 tickets and that is the amount of money I spent...$11.00. I use the same technique for larger games and one weekend I had 3 numbers on the MegaMillion and the following Monday 3 numbers on the Two Step.

My predictions are not based upon historical patterns persay. Just because you don't understand my concept does not by any means there is no rationale to it. Just because numbers are being tracked doesn't mean that mathematical formulas, etc are the only thing that will produce the right numbers to use.

A minimum of 13 numbers can be eliminated each game if you know what to look for. Most of the time 19 numbers can be eliminated.

Don't get me wrong now. I really don't want to get into a mine is better than yours crap either. I was simply pointing out the difference between yours and my generator. Most people use Generators that are Random. Mine happens to be controlled. I've only had it about two weeks now so all previous wins were done without the use of a generator and simply using other software to track things for me.

United States Member #83701 December 13, 2009 225 Posts Offline

Posted: August 24, 2010, 11:46 am - IP Logged

Quote: Originally posted by Clipper on August 24, 2010

To answer your question.....yes the combination that played was among the 296 as was the combination that played tonight out of 20 possible combinations generated once I knew the kind of game that would probably appear. I don't filter out "number sets" either I don't need to go back more then 31 games to figure out what is about to happen. Anything beyond that is not necessary.

Yes, there was physical reason for believing numbers above 28 would not play. Simple fact, determination there was high probability that only 3 Denominations out of 4 would be used and by tracking what Denomination would not play it indicated 30's would not play if only 3 Denominations appeared. If anything played higher then 28 there was indication 32 would play.....31 played. I ruled out all 33-34 and 35.

Why would the bonus ball involve digits 0-1-2 or 3? I track the Bonus Ball separately using personal Pick 3 software . It indicated to me any digit used would come from within the last 4 games which eliminated 4-5-6-7-8-9. The Bonus Ball was 2.

Why would the first number be Odd? I track Alpha Patterns and there was indication the first two numbers would go Odd-Even. It went Even/Even. I played 3 and 5 (but not together)......4 played.

I think that covers your questions. I've tracked for 15 years. Been working on the Two Step for the past 6 months. Have had 3 Numbers of the 4 several times and on other occassions the Bonus Ball several times. One night I had the BB on 11 tickets and that is the amount of money I spent...$11.00. I use the same technique for larger games and one weekend I had 3 numbers on the MegaMillion and the following Monday 3 numbers on the Two Step.

My predictions are not based upon historical patterns persay. Just because you don't understand my concept does not by any means there is no rationale to it. Just because numbers are being tracked doesn't mean that mathematical formulas, etc are the only thing that will produce the right numbers to use.

A minimum of 13 numbers can be eliminated each game if you know what to look for. Most of the time 19 numbers can be eliminated.

Don't get me wrong now. I really don't want to get into a mine is better than yours crap either. I was simply pointing out the difference between yours and my generator. Most people use Generators that are Random. Mine happens to be controlled. I've only had it about two weeks now so all previous wins were done without the use of a generator and simply using other software to track things for me.

This is truly fascinating.

Yes, I don't understand the rationale behind your concepts but that doesn't mean they are not of interest to me. Let's look at the simplest and I would say thereby most difficult to develop a rationale for and that is the bonus ball. It's a single draw from a pool of numbers. There's very little sample size for any meaningful statistics (that can be said about lotteries in general) and the only possible source of bias would be the sensitivity to boundary conditions such as ball order, time when ball is released into the machine, position and movement of paddles and time when the gate is opened to extract a ball. This sensitivity is far too complicated for people to model at this time so it's reasonable to believe that there is only an assumption of random like distribution with probabilistic confirmation from statistics (most likely the chi square test). That is, there might be something useful there but it would be very hard to sort out, so hard that the randomness is generally accepted.

You mentioned Alpha Tracking, I'm not certain as to what you are talking about when you say Alpha Tracking. The usage of the term Alpha is usually for classification hence with neurobiology, alpha waves are the most obvious and dominant frequencies when in a certain conscious state. In almost every field, Alpha would simply be the most obvious. With demographics, there's the concept of Alpha diversity which is a rating used to determine variations in geographic resources by analysing populations of various species and though they would cite Shannon & Weaver as a source, the equations they use look very similar to the earlier works in signal processing at Bell Labs upon which Shannon's work is based. Shannon is considered seminal to the field of informatics and much of the work involves determining if and how much information can be contained in a pattern, prior to Shannon, Nyquist's concept of the Nyquist frequency was the defining limit for information. Of course once the pattern is known then there is no information in the pattern. This is very interesting because informatics is used in gaming and investing theory so there may be applications in lotteries, I'll have to look further into this but I don't have access to my texts for the moment and DSP was always a weak area of mine in my opinion. Even so, in informatics the term Alpha pattern would be more likely used to summarize entropy levels. In image and sound processing, it may be a process used to simulate certain effects, essentially a type of fractal but what fractal primitive could possibly be associated with a ball being drawn from a pool of other balls? An impulse function perhaps? The early fractal images of mountains were generated from triangular or rather cone and pyramid shaped primitives which make sense because the overall shape of a mountain is triangular but that is fundamentally wrong as it results in a landscape full of lakes. A subtractive channel from a given point to a defined base would be an appropriate mountain fractal primitive given that mountains are defined by erosion basically a branch primitive similar to those used to produce tree images but used subtractively. A triangular sawtooth with a rise and then a sharp dropoff may be a good primitive for the stock market as it reflects the general life cycle of a business. What could possibly be the basis of your use of the term "Alpha Pattern" and how would even or odd numbers matter to such a pattern? There's no inherent significance of even or odd, it's not as if balls have chirality though the machine does but only inso far as the insertion point so the concept of even and odd only exists in the order which the balls enter the chamber.

The use of common Pick 3 software is also interesting seeing how Pick 3 is ideally three separate draws from pools of 10 but tends to be functionally by RNG. However, I'll be satisfied to just sort out what you mean by Alpha Pattern for the time being.

I'm sorry, it's possible to apply theorems without knowing the underlying relationships and that's perfectly acceptable as long as that's acknowledged but an actual reason is presumed so there should be some possible hypothesis for a given method to apply. I'd like to know what that hypothesis may be. Please excuse my curiosity.

Texas United States Member #1477 May 7, 2003 2311 Posts Offline

Posted: August 24, 2010, 12:47 pm - IP Logged

Quote: Originally posted by jwhou on August 24, 2010

This is truly fascinating.

Yes, I don't understand the rationale behind your concepts but that doesn't mean they are not of interest to me. Let's look at the simplest and I would say thereby most difficult to develop a rationale for and that is the bonus ball. It's a single draw from a pool of numbers. There's very little sample size for any meaningful statistics (that can be said about lotteries in general) and the only possible source of bias would be the sensitivity to boundary conditions such as ball order, time when ball is released into the machine, position and movement of paddles and time when the gate is opened to extract a ball. This sensitivity is far too complicated for people to model at this time so it's reasonable to believe that there is only an assumption of random like distribution with probabilistic confirmation from statistics (most likely the chi square test). That is, there might be something useful there but it would be very hard to sort out, so hard that the randomness is generally accepted.

You mentioned Alpha Tracking, I'm not certain as to what you are talking about when you say Alpha Tracking. The usage of the term Alpha is usually for classification hence with neurobiology, alpha waves are the most obvious and dominant frequencies when in a certain conscious state. In almost every field, Alpha would simply be the most obvious. With demographics, there's the concept of Alpha diversity which is a rating used to determine variations in geographic resources by analysing populations of various species and though they would cite Shannon & Weaver as a source, the equations they use look very similar to the earlier works in signal processing at Bell Labs upon which Shannon's work is based. Shannon is considered seminal to the field of informatics and much of the work involves determining if and how much information can be contained in a pattern, prior to Shannon, Nyquist's concept of the Nyquist frequency was the defining limit for information. Of course once the pattern is known then there is no information in the pattern. This is very interesting because informatics is used in gaming and investing theory so there may be applications in lotteries, I'll have to look further into this but I don't have access to my texts for the moment and DSP was always a weak area of mine in my opinion. Even so, in informatics the term Alpha pattern would be more likely used to summarize entropy levels. In image and sound processing, it may be a process used to simulate certain effects, essentially a type of fractal but what fractal primitive could possibly be associated with a ball being drawn from a pool of other balls? An impulse function perhaps? The early fractal images of mountains were generated from triangular or rather cone and pyramid shaped primitives which make sense because the overall shape of a mountain is triangular but that is fundamentally wrong as it results in a landscape full of lakes. A subtractive channel from a given point to a defined base would be an appropriate mountain fractal primitive given that mountains are defined by erosion basically a branch primitive similar to those used to produce tree images but used subtractively. A triangular sawtooth with a rise and then a sharp dropoff may be a good primitive for the stock market as it reflects the general life cycle of a business. What could possibly be the basis of your use of the term "Alpha Pattern" and how would even or odd numbers matter to such a pattern? There's no inherent significance of even or odd, it's not as if balls have chirality though the machine does but only inso far as the insertion point so the concept of even and odd only exists in the order which the balls enter the chamber.

The use of common Pick 3 software is also interesting seeing how Pick 3 is ideally three separate draws from pools of 10 but tends to be functionally by RNG. However, I'll be satisfied to just sort out what you mean by Alpha Pattern for the time being.

I'm sorry, it's possible to apply theorems without knowing the underlying relationships and that's perfectly acceptable as long as that's acknowledged but an actual reason is presumed so there should be some possible hypothesis for a given method to apply. I'd like to know what that hypothesis may be. Please excuse my curiosity.

That's just it....there is no need to apply theorems as there is no mathematical solution needed or required. Just using simple Basic Math ideas and common sense. Over a period of time and education some people have a problem reverting to basic math and their concepts.

Alpha Patterns is nothing nore the "letter tracking". E for Even and O for Odd. I use my Pick 3 software and assign 1 to Odd and 2 to Even. For instance last night the 1st three numbers that played were 4-18-25. This equates to 221 entry in the Pick 3 software. Each pattern has a different Sum Total.

111 = 3 Always indicates all Odd Numbers

112 = 4 This could appear as 121 or 211 but still Sum Totals 4 & 2 Odd/1 Even

221 - 5 This could appear 212 or 122 but still Sum Total 5 * 3 Even/1 Odd

222 = 6 Always indicates all Even Numbers

I can block all digits 3-4-5-6-7-8-9 in my Pick 3 software, check to see what the next Sum Total will be to determine how many Odd or Even numbers will be used and with other manipulation determine what order the 1's and 2's will be in as they have their own Alpha Pattern. OOO-OOE-OEO-EOO-EE0-EOE-OEE-OOO.

When I track the numbers in pairs a constant "2" always repesents the 1st entry in the Pick 3 software which could be be entered as 022 for last nights 4-18 but since the 0 is also Even I just use 2 and make allowance for it in the Sum Total. Makes for 1 less digit to track too.

Once I determine how many Odd/Even numbers are likely to play I can set my Generator software to produce only combinations that have "x" number of Odd/Even Numbers.

As far as the Bonus Ball, I use the same Pick 3 software. Since there are 3 digits required for the software I have to add one or two constant "0's" to make entry in the database. For Singles 1-2-3-4-5-6-7-8-9 I would enter 001, 002, 003, 004, 005, 006, etc. Other entries would require one"0"...011, 012, 013,...021, 022, 023....Ten would be 010; Twenty would be 020; Thirty would be 030.

I then track them like Pick 3 game. So I deal only with 3 numbers combinations 001 through 035. Odds 1 in 35. Knowing facts about the Pick 3 of which 1 is that a digit that just played usually plays again before it gets 5 games out. Last night all I had to do was look at that chart to see how many times it's been since numbers had played and it indicated that no digit past 4 games would be used. The only digits that would be effected were 0-1-2-3 which could create only 15 possible numbers. Then use Sum Total Chart and Alpha Charts for further determination and elilmination. I eliminated numbers 3-13-23-33. Notice I didn't eliminate the 30 though. There are a few other things about this number that can be tracked as well....like what Denomination it will be. Single, Teen, Twenty or Thirty.

Just simple, basic tracking of numbers without all the mathematical formulas and statistics. Actually I have tracked numbers longer then 15 years as I started when the Lotto first started in Texas. Once the Pick 3 started I realized how much easier it was to track so I tracked it and have done most of my work on the Pick 3 where I realized a lot of facts. One being that if you can figure out 1 digit that is going to play AND what position that digit will be in you can reduce your 1 in 1000 to 1 in 100. The 100 includes all PURES and Doubles connected to that 1 digit. How do you figure out what digit? I learned that 50-65% of the time that a number that just played will repeat in the next game so I tracked incidence of repeat. Some times it would repeat once and then no repeat (which I call Break) and other times a number (not necessarily the same number) would repeat several times in a row before a "Break".

I also learned that most digits will not go out further then 4 games before they replay. So I have a charts that track that information. One of which tracks how many digits come from 1-4 past games and how many 5+ games out.

I use Alpha Patterns in this game and Sum Totals in the game. 2 Odd/1 Even will always Sum Total Even; 2 Even/1 Odd will always Sum Total Odd. Of course, all Odd Sum Odd and all Even Sum Even. Figure out which way it Sum Totals and you can cut half the 1,000 Exacts. It doesn't take involved mathematical calculations to do this.

When I started I didn't even have a computer nor knowledge of what was available to track numbers. So I had to come up with a way to track numbers and simply studied the games that did appear and started noticing some facts that have never changed in the past nearly 20 years that I have been tracking. I rely on those "facts" to guide me in number selection because they have proven time and again that they aren't going to change. They may flucuate but then again....nothing is constant. You just have to allow for the change and go with the flow.

It does not matter what kind of lottery machines the Lottery Commission uses. Whether it is computerized or ping pong ball. Doesn't matter the weight of the balls, paddles are anything else. With this system if a Number is going to play.......it's going to play and there is nothing that can be done to change that. The lottery commission would also have to be able to make such determinations and do something to change the outcome which would be illegal since it's stated the generations are Random.

United States Member #83701 December 13, 2009 225 Posts Offline

Posted: August 24, 2010, 7:11 pm - IP Logged

Quote: Originally posted by Clipper on August 24, 2010

That's just it....there is no need to apply theorems as there is no mathematical solution needed or required. Just using simple Basic Math ideas and common sense. Over a period of time and education some people have a problem reverting to basic math and their concepts.

Alpha Patterns is nothing nore the "letter tracking". E for Even and O for Odd. I use my Pick 3 software and assign 1 to Odd and 2 to Even. For instance last night the 1st three numbers that played were 4-18-25. This equates to 221 entry in the Pick 3 software. Each pattern has a different Sum Total.

111 = 3 Always indicates all Odd Numbers

112 = 4 This could appear as 121 or 211 but still Sum Totals 4 & 2 Odd/1 Even

221 - 5 This could appear 212 or 122 but still Sum Total 5 * 3 Even/1 Odd

222 = 6 Always indicates all Even Numbers

I can block all digits 3-4-5-6-7-8-9 in my Pick 3 software, check to see what the next Sum Total will be to determine how many Odd or Even numbers will be used and with other manipulation determine what order the 1's and 2's will be in as they have their own Alpha Pattern. OOO-OOE-OEO-EOO-EE0-EOE-OEE-OOO.

When I track the numbers in pairs a constant "2" always repesents the 1st entry in the Pick 3 software which could be be entered as 022 for last nights 4-18 but since the 0 is also Even I just use 2 and make allowance for it in the Sum Total. Makes for 1 less digit to track too.

Once I determine how many Odd/Even numbers are likely to play I can set my Generator software to produce only combinations that have "x" number of Odd/Even Numbers.

As far as the Bonus Ball, I use the same Pick 3 software. Since there are 3 digits required for the software I have to add one or two constant "0's" to make entry in the database. For Singles 1-2-3-4-5-6-7-8-9 I would enter 001, 002, 003, 004, 005, 006, etc. Other entries would require one"0"...011, 012, 013,...021, 022, 023....Ten would be 010; Twenty would be 020; Thirty would be 030.

I then track them like Pick 3 game. So I deal only with 3 numbers combinations 001 through 035. Odds 1 in 35. Knowing facts about the Pick 3 of which 1 is that a digit that just played usually plays again before it gets 5 games out. Last night all I had to do was look at that chart to see how many times it's been since numbers had played and it indicated that no digit past 4 games would be used. The only digits that would be effected were 0-1-2-3 which could create only 15 possible numbers. Then use Sum Total Chart and Alpha Charts for further determination and elilmination. I eliminated numbers 3-13-23-33. Notice I didn't eliminate the 30 though. There are a few other things about this number that can be tracked as well....like what Denomination it will be. Single, Teen, Twenty or Thirty.

Just simple, basic tracking of numbers without all the mathematical formulas and statistics. Actually I have tracked numbers longer then 15 years as I started when the Lotto first started in Texas. Once the Pick 3 started I realized how much easier it was to track so I tracked it and have done most of my work on the Pick 3 where I realized a lot of facts. One being that if you can figure out 1 digit that is going to play AND what position that digit will be in you can reduce your 1 in 1000 to 1 in 100. The 100 includes all PURES and Doubles connected to that 1 digit. How do you figure out what digit? I learned that 50-65% of the time that a number that just played will repeat in the next game so I tracked incidence of repeat. Some times it would repeat once and then no repeat (which I call Break) and other times a number (not necessarily the same number) would repeat several times in a row before a "Break".

I also learned that most digits will not go out further then 4 games before they replay. So I have a charts that track that information. One of which tracks how many digits come from 1-4 past games and how many 5+ games out.

I use Alpha Patterns in this game and Sum Totals in the game. 2 Odd/1 Even will always Sum Total Even; 2 Even/1 Odd will always Sum Total Odd. Of course, all Odd Sum Odd and all Even Sum Even. Figure out which way it Sum Totals and you can cut half the 1,000 Exacts. It doesn't take involved mathematical calculations to do this.

When I started I didn't even have a computer nor knowledge of what was available to track numbers. So I had to come up with a way to track numbers and simply studied the games that did appear and started noticing some facts that have never changed in the past nearly 20 years that I have been tracking. I rely on those "facts" to guide me in number selection because they have proven time and again that they aren't going to change. They may flucuate but then again....nothing is constant. You just have to allow for the change and go with the flow.

It does not matter what kind of lottery machines the Lottery Commission uses. Whether it is computerized or ping pong ball. Doesn't matter the weight of the balls, paddles are anything else. With this system if a Number is going to play.......it's going to play and there is nothing that can be done to change that. The lottery commission would also have to be able to make such determinations and do something to change the outcome which would be illegal since it's stated the generations are Random.

Interesting.

You may not be aware of it but reducing the numbers down to even or odd is essentially the mathematical operation modulus (represented by the % symbol), basically you're taking the modulus 2 of the number (though a modulus 2 would render even numbers to be 0 and odd numbers to be 1). By taking the modulus 2 of three numbers to come up with three binary values, you're classifying them to 8 possibilities, by summing them, you are classifying them into four categories, two of which are each three times more prevalent then each of the other two i.e.: with your notation the sums of 4 and 5 would each appear three times more often then the sums of 3 and 6.

I don't know what your pick 3 number software does but the distribution bias introduced by summation should give it a pattern to feed on. Whether it's meaningful or not is another matter. It's as if you've inadvertently introduced a pattern for the pick 3 software to use, it may even be possible to quantify the periodicity of the pattern you've introduced. I usually find such software distasteful since it's not always clear what it's doing, best to write one's own.

There is a significant difference between computerized draws and ping pong balls. Indeed computerized draws may lend credence to your methods especially if they simply used the basic linear congruential generator built into the compiler which is actually very likely. There is random and then there is random. The lottery commissions go by what's legally accepted as random and that is a far broader definition than one would expect. Thats what gives people who attempt to analyze the lottery hope.

A lot of people would use the concept of skip interval to describe what you are referring to as repeat and break. To a certain extent that is just a matter of distribution, sorta like the mythological law of averages.

I don't see any direct relationships that would indicate that any of this would have any relevance but as long as you're willing to admit it's just a pattern match then it seems quite suitable. With some work, it could be related to some basis at some point in the future. Your methods are not without precedent, the reduction of data through modulus is analogous to some stock tracking techniques such as candlestick graphs and point and follow graphs, usually those techniques are focused on detecting change to determine if there is a fundamental change in driving factors. It would be difficult to rationalize that to a lottery because the lottery attempts to keep as many variables as possible the same from draw to draw so there should be no fundamental change to detect. It's been 14 years since I took a look at stock analysis techniques, perhaps it's time to revisit the subject. Hmm, they have magazines on technical analysis of stocks and there's always a discussion of an obscure technique which is essentially just applying systematic processes sometimes without clear rationale, maybe it's time for similar articles on lottery analysis. Unfortunately, I don't think most people's techniques are actually analytical and are more likely to be heuristic.

Near as I can tell, you have a representation transformation in terms of the reduction to odds and evens. Then you use somebody's pick 3 software whose internals we can only guess at and then apply a set of heuristics based on your observations presumably since 1992 since that's when Texas first introduced the lottery. Do you apply all of your observations systematically or do you apply them selectively as needed?

To be frank, although I find this interesting, I would never pay for anyone's book on how they analyze the lottery.

United States Member #93947 July 10, 2010 2180 Posts Offline

Posted: August 25, 2010, 1:36 am - IP Logged

Quote: Originally posted by jwhou on August 24, 2010

Interesting.

You may not be aware of it but reducing the numbers down to even or odd is essentially the mathematical operation modulus (represented by the % symbol), basically you're taking the modulus 2 of the number (though a modulus 2 would render even numbers to be 0 and odd numbers to be 1). By taking the modulus 2 of three numbers to come up with three binary values, you're classifying them to 8 possibilities, by summing them, you are classifying them into four categories, two of which are each three times more prevalent then each of the other two i.e.: with your notation the sums of 4 and 5 would each appear three times more often then the sums of 3 and 6.

I don't know what your pick 3 number software does but the distribution bias introduced by summation should give it a pattern to feed on. Whether it's meaningful or not is another matter. It's as if you've inadvertently introduced a pattern for the pick 3 software to use, it may even be possible to quantify the periodicity of the pattern you've introduced. I usually find such software distasteful since it's not always clear what it's doing, best to write one's own.

There is a significant difference between computerized draws and ping pong balls. Indeed computerized draws may lend credence to your methods especially if they simply used the basic linear congruential generator built into the compiler which is actually very likely. There is random and then there is random. The lottery commissions go by what's legally accepted as random and that is a far broader definition than one would expect. Thats what gives people who attempt to analyze the lottery hope.

A lot of people would use the concept of skip interval to describe what you are referring to as repeat and break. To a certain extent that is just a matter of distribution, sorta like the mythological law of averages.

I don't see any direct relationships that would indicate that any of this would have any relevance but as long as you're willing to admit it's just a pattern match then it seems quite suitable. With some work, it could be related to some basis at some point in the future. Your methods are not without precedent, the reduction of data through modulus is analogous to some stock tracking techniques such as candlestick graphs and point and follow graphs, usually those techniques are focused on detecting change to determine if there is a fundamental change in driving factors. It would be difficult to rationalize that to a lottery because the lottery attempts to keep as many variables as possible the same from draw to draw so there should be no fundamental change to detect. It's been 14 years since I took a look at stock analysis techniques, perhaps it's time to revisit the subject. Hmm, they have magazines on technical analysis of stocks and there's always a discussion of an obscure technique which is essentially just applying systematic processes sometimes without clear rationale, maybe it's time for similar articles on lottery analysis. Unfortunately, I don't think most people's techniques are actually analytical and are more likely to be heuristic.

Near as I can tell, you have a representation transformation in terms of the reduction to odds and evens. Then you use somebody's pick 3 software whose internals we can only guess at and then apply a set of heuristics based on your observations presumably since 1992 since that's when Texas first introduced the lottery. Do you apply all of your observations systematically or do you apply them selectively as needed?

To be frank, although I find this interesting, I would never pay for anyone's book on how they analyze the lottery.

jwhou,

So where were you when I needed you during my pummeling in my Fooled By Randomness thread in Lottery Systems? Your post above and its immediate predecessor provide enough insight [for me] to feel even more secure in my belief that most lottery machines and algorithms are "sufficiently random" for their assigned task. I too have worked with several types of RNGs while doing OS simulations and some not too critical database encryptions. I found the basic linear conguential generator invaluable during debugging, because of its repeatability. For database encryption, my model needed the repeatability of the sequence, using it in conjunction with username/passwords in both directions. When I wanted longer(truly? random) sequences for simulations, I would "kick" the congruential generator randomly forward using physical measures accessible to the program like the clock or temperature. I suspect your encryption applications required much more sophistication than my work, but I think my methods were "sufficiently random" for our work!

The above is easy for me to spit out, but what I really want to express here is more difficult. Given the analysis in your last 2 posts, especially your delineation of the even/odd issue above, I find it bordering on impossible to believe that anyone would be able to [essentially] devise a decryption scheme that decrypts things like the effects of a corrupt programmer's code segment or tiny variations in balls that would have any utility. To me, it's even more difficult to accept when you remember that the very large numbers generated by RNGs are, for the most part, being scaled down to integers under 60! To me, the analogies which can be drawn between these nuances of RNGs and the near infinitessimal mechanical differences in the ball machines is telling.

A point that I have been trying to make over my 5 weeks of posting here is that EVEN IF there is fraud, or wear and tear, playing matrix manipulation games and/or adding constants to previous draws using modulus arithmetic, will NOT enable anyone to predict a future draw! As you suggest, I too play sets that attempt to produce as many small hits as possible; it would get pretty boring if you only rarely got to cash a winner, even though your overall odds for a Jackpot are the same.

"To be frank, although I find this interesting, I would never pay for anyone's book on how they analyze the lottery."

Based on this, your last line above, I suspect you might agree with me, at least to some degree.

Having said this, I offer an example where programmers of lottery commission software, could, in fact, pull off a major fraud, and do it undetected. It's the concern that brought me to this site, and is my first post. I'd welcome your opinion on it.

United States Member #83701 December 13, 2009 225 Posts Offline

Posted: August 25, 2010, 10:54 pm - IP Logged

Quote: Originally posted by jimmy4164 on August 25, 2010

jwhou,

So where were you when I needed you during my pummeling in my Fooled By Randomness thread in Lottery Systems? Your post above and its immediate predecessor provide enough insight [for me] to feel even more secure in my belief that most lottery machines and algorithms are "sufficiently random" for their assigned task. I too have worked with several types of RNGs while doing OS simulations and some not too critical database encryptions. I found the basic linear conguential generator invaluable during debugging, because of its repeatability. For database encryption, my model needed the repeatability of the sequence, using it in conjunction with username/passwords in both directions. When I wanted longer(truly? random) sequences for simulations, I would "kick" the congruential generator randomly forward using physical measures accessible to the program like the clock or temperature. I suspect your encryption applications required much more sophistication than my work, but I think my methods were "sufficiently random" for our work!

The above is easy for me to spit out, but what I really want to express here is more difficult. Given the analysis in your last 2 posts, especially your delineation of the even/odd issue above, I find it bordering on impossible to believe that anyone would be able to [essentially] devise a decryption scheme that decrypts things like the effects of a corrupt programmer's code segment or tiny variations in balls that would have any utility. To me, it's even more difficult to accept when you remember that the very large numbers generated by RNGs are, for the most part, being scaled down to integers under 60! To me, the analogies which can be drawn between these nuances of RNGs and the near infinitessimal mechanical differences in the ball machines is telling.

A point that I have been trying to make over my 5 weeks of posting here is that EVEN IF there is fraud, or wear and tear, playing matrix manipulation games and/or adding constants to previous draws using modulus arithmetic, will NOT enable anyone to predict a future draw! As you suggest, I too play sets that attempt to produce as many small hits as possible; it would get pretty boring if you only rarely got to cash a winner, even though your overall odds for a Jackpot are the same.

"To be frank, although I find this interesting, I would never pay for anyone's book on how they analyze the lottery."

Based on this, your last line above, I suspect you might agree with me, at least to some degree.

Having said this, I offer an example where programmers of lottery commission software, could, in fact, pull off a major fraud, and do it undetected. It's the concern that brought me to this site, and is my first post. I'd welcome your opinion on it.

I occasionally take prolonged breaks from reading lotterypost and I believe your "fool by randomness" thread was just prior to my return. For all practical purposes, what the lotteries do with their RNG is obviously enough, they rake in the money and people keep playing. That's not to say that there isn't any flaws with their RNG's, there almost certainly is. Of course, with the exception of Keno, chances are the software isn't in the same RNG sequence from draw to draw so there's very little sample size to sync into and certainly no opportunity to exploit anything if you did.

Regardless of whether there is or isn't a basis to what people do with their "systems", it's still quite interesting and many are simply making things up even though they may already exist such as the even odd representation already being a modulus function.

I haven't come across Conover yet, thanks for the link. I only recently came across Shannon even though my academic background is in his field. I'm kind of kicking myself for not coming across him earlier and now that I think about it, I think his work did come up a few times as well.

As to your ADM post, I believe there was an exploit of Keno in Las Vegas where a Casino insider stole the code, modified it to run on his laptop and allow him to enter a sample of the Keno draws that day, he had a few equations that he developed to reduce the search space for the seed and the code would find a few seeds that matched and play them forward to catch up to the real Keno draws. Early on in the 80's there was also a story of how a programmer in Vegas had planted exploits into the chips for the slot machines for the mob but his life was in danger for talking about it. Of course, at least one poster to your thread pointed out that the mechanical slot machines could also be exploited. Everyone who has tried lock picking knows that the human sense of touch and timing can generate amazingly consistent results, a microsurgeon is a classic example of this where fractions of a millimeter count, a human with a scalpel is the only way to go. If I were to plant an ADM exploit, I would not do so in the code that I was developing as that would leave an audit trail straight to me but I would plant a virus that attacked the RNG library which almost certainly will be a commercial product from Microsoft simply for legal indemnification. The virus would lay moot until certain days that are not that frequent and on those days it would reduce the seed space to a limited but still large selection of seeds. I would use Keno as that would have the most draws and would run throughout the day. As the programmer, you can replicate the code and database the sequences in that limited seed space. The virus would only kick in very infrequently at random but often enough say once a quarter. You would enter in a couple of the Keno draws each day, your computer would check to see if that sequence is one of the limited seed space sequences and if it is, you go to town. The virus would after a few months unencrypt a common viral tag so that a antivirus checker would identify it as a common virus but since that would occur long after you've parted company, suspicion would be raised on someone else. That way if your unusual winning streak is investigated which it almost certainly will, the virus popping up while you clearly had no opportunity to introduce it would get you off the hook. You must remember to remove your laptop hard drive and destroy it before claiming your winnings. Such an exploit would always pass all possible statistical testing and would require decades of data before there might be even the slightest hint the seeds had been compromised, when the virus mutated to be found, it would show that it would use completely different seeds than the ones you used hence it would be proof that you did not cheat but someone else was going to. After that it would be a matter of whether or not you had the opportunity to replicate the code externally to play it forward and backwards hence the destruction of all your hard drives. To exploit a Jackpot draw, the virus would be set to use one of say 10 seeds, once and only once. It would remove itself and all you have to do is buy the same ten tickets for every draw till you win. It would be almost impossible to track the exploit to you.

United States Member #93947 July 10, 2010 2180 Posts Offline

Posted: August 25, 2010, 11:24 pm - IP Logged

Quote: Originally posted by jwhou on August 25, 2010

I occasionally take prolonged breaks from reading lotterypost and I believe your "fool by randomness" thread was just prior to my return. For all practical purposes, what the lotteries do with their RNG is obviously enough, they rake in the money and people keep playing. That's not to say that there isn't any flaws with their RNG's, there almost certainly is. Of course, with the exception of Keno, chances are the software isn't in the same RNG sequence from draw to draw so there's very little sample size to sync into and certainly no opportunity to exploit anything if you did.

Regardless of whether there is or isn't a basis to what people do with their "systems", it's still quite interesting and many are simply making things up even though they may already exist such as the even odd representation already being a modulus function.

I haven't come across Conover yet, thanks for the link. I only recently came across Shannon even though my academic background is in his field. I'm kind of kicking myself for not coming across him earlier and now that I think about it, I think his work did come up a few times as well.

As to your ADM post, I believe there was an exploit of Keno in Las Vegas where a Casino insider stole the code, modified it to run on his laptop and allow him to enter a sample of the Keno draws that day, he had a few equations that he developed to reduce the search space for the seed and the code would find a few seeds that matched and play them forward to catch up to the real Keno draws. Early on in the 80's there was also a story of how a programmer in Vegas had planted exploits into the chips for the slot machines for the mob but his life was in danger for talking about it. Of course, at least one poster to your thread pointed out that the mechanical slot machines could also be exploited. Everyone who has tried lock picking knows that the human sense of touch and timing can generate amazingly consistent results, a microsurgeon is a classic example of this where fractions of a millimeter count, a human with a scalpel is the only way to go. If I were to plant an ADM exploit, I would not do so in the code that I was developing as that would leave an audit trail straight to me but I would plant a virus that attacked the RNG library which almost certainly will be a commercial product from Microsoft simply for legal indemnification. The virus would lay moot until certain days that are not that frequent and on those days it would reduce the seed space to a limited but still large selection of seeds. I would use Keno as that would have the most draws and would run throughout the day. As the programmer, you can replicate the code and database the sequences in that limited seed space. The virus would only kick in very infrequently at random but often enough say once a quarter. You would enter in a couple of the Keno draws each day, your computer would check to see if that sequence is one of the limited seed space sequences and if it is, you go to town. The virus would after a few months unencrypt a common viral tag so that a antivirus checker would identify it as a common virus but since that would occur long after you've parted company, suspicion would be raised on someone else. That way if your unusual winning streak is investigated which it almost certainly will, the virus popping up while you clearly had no opportunity to introduce it would get you off the hook. You must remember to remove your laptop hard drive and destroy it before claiming your winnings. Such an exploit would always pass all possible statistical testing and would require decades of data before there might be even the slightest hint the seeds had been compromised, when the virus mutated to be found, it would show that it would use completely different seeds than the ones you used hence it would be proof that you did not cheat but someone else was going to. After that it would be a matter of whether or not you had the opportunity to replicate the code externally to play it forward and backwards hence the destruction of all your hard drives. To exploit a Jackpot draw, the virus would be set to use one of say 10 seeds, once and only once. It would remove itself and all you have to do is buy the same ten tickets for every draw till you win. It would be almost impossible to track the exploit to you.

Hopefully, your description of several more ingenous ways beyond the one I warned about that IT people could rip off the lottery will add support to the efforts underway to eliminate computerized draws. The possibility that someone could design software for their employer that is vulnerable to virus attacks, and then attack it, is reason enough to shut down all computer drawings. Amazing!

Texas United States Member #92330 June 5, 2010 887 Posts Offline

Posted: August 30, 2010, 3:40 am - IP Logged

Quote: Originally posted by Coin Toss on August 23, 2010

$100,000,000 is mere chump change to NASA.

Sylvia Browne, in one of her books, said Earth is the worse place in the entire universe.

Some say this planet is another planet's hell, or perhaps hell for the universe.

Others say it's the insane asylum for the universe.

I'd really like to know what the Bill Gates, Carlos Slim and Warren Buffet did to have to live in this hell. Earth may be hell for some people but it is heaven to others, all depends on what side of the money you're on. Just look at those living in Hollywood compared to those living in Zimbabwe, Ethiopia or Botswana.

But back to the dream of having $100 million and dealing with the ladies... I think I could do pretty well. Part of the reason I want to win big is so that I can have these girls come up to me getting all Anna Nicole Smith and just shut 'em down. I know those that are my friends and those that would never have given me the time of day

Aside from the lady situation, my dream of what to do with the money is just to buy a decent house, nice car, nice furnishings and not have to worry about the next paycheck. I would actually prefer winning something more along the lines of $15 million (after tax) just so it would be easier to manage. Don't have any huge dreams of ten million dollar mansions, one million dollar cars or million dollar trips. Just quietly living the good life.