Welcome Guest
Log In | Register )
You last visited December 9, 2016, 6:13 pm
All times shown are
Eastern Time (GMT-5:00)

What constitutes a good reduction filter?

Topic closed. 22 replies. Last post 4 years ago by SergeM.

Page 2 of 2
3.52
PrintE-mailLink
lottoarchitect's avatar - waveform

Greece
Member #2815
November 18, 2003
502 Posts
Offline
Posted: July 15, 2012, 12:40 am - IP Logged

And what is the performance of the individual reduction elements in your normal filtering application? To get to a 12 number scheme you must have reduction method that has an associated error rate. If you look at the Maddog challenges in the jackpot section, the majority of people are having a very rough time with the 12 number scenario. In which case, it does matter what wheeling method you use if you don’t have any reliability in the 12 numbers you’ve chosen.

 


JKING, back in 2003 I developed a fully automated filtering process which I call ECD (Error Correction Design). Its target was to identify the errors produced by filtering combinations in such a way so not to remove the winning combination if possible and it was self-adaptive in time. The idea of this system occcurred simply because one filter alone is not enough for good filtering (unless if set too agressive) and setting up several filters have the problem, 90% of the filters to set them correctly but the other 10% will filter out the winning combination. My tests were mainly performed on a 5/45 game (1221759 combinations) and the results were elimination 95-99% of the tickets and keep the winning combination 20% of the time. The problem here is that the remaining 10000-50000 tickets are still too many to be played. If my initial combinations are a full wheel, this method returns proportionally to its size acceptable tickets. So, my conclusion is, such an approach is only viable to pools of players who want to spent good money but within one month they are due to win the jackpot. For players who want to spend a few dollars, the only other method I have seen producing results is what my users do (wheel around 12 numbers, match 4-5 correct numbers every so often and the cost to play is about $8-12 for the wheel).

If you have something to do, at least do it well...

    JKING's avatar - Kaleidoscope 3.gif

    United States
    Member #5599
    July 13, 2004
    1185 Posts
    Offline
    Posted: July 15, 2012, 2:55 am - IP Logged

    Hi,

       Lakerben…

    I've been around the lottery scene for a long time.  My question is what do you base your assumptions on : by experience??  And what system do you use ?  I've been testing  and analyzing systems for over 25 years, and I have learned many different methods from Wheels,matrix of all sizes, grids you name it.

     

    I base my own designed system on historical data. The core of it is the amount of occurrences that happens for each lottery number over different windows in time. The associated positional occurrence distribution patterns for each window in time set s the limits for each filter…..

    Window in time #1 - 3 game pick occurrence tracking for a 5/39 game.

    Y axis = amt of occurrances X axis= sorted accumulation of historical values.

    OCC3-(3)

    1

    2

    3

    4

    5

    TOT

    0

    473

    454

    373

    216

    56

    1572

    1

    0

    19

    100

    253

    328

    700

    2

    0

    0

    0

    4

    86

    90

    3

    0

    0

    0

    0

    3

    3

    4

    0

    0

    0

    0

    0

    0

    5

    0

    0

    0

    0

    0

    0

     

    OCC3-(3)%

    1

    2

    3

    4

    5

    TOT

    0

    100

    96

    78.9

    45.7

    11.8

    332

    1

    0

    4.02

    21.1

    53.5

    69.3

    148

    2

    0

    0

    0

    0.85

    18.2

    19

    3

    0

    0

    0

    0

    0.63

    0.63

    4

    0

    0

    0

    0

    0

    0

    5

    0

    0

    0

    0

    0

    0

     

    I also run about 20 other filter variations based on occurrence themes.

    Part 2 of my programs go to the more traditional numerical methods you see at the LP, like Rl’s Double odd and double even filters. I have a couple other things going on that I won’t get into here. But it is a by the numbers approach based on historical fact.

    I hope this answers your question. *S*

    Lottoarchitect…

    JKING, back in 2003 I developed a fully automated filtering process which I call ECD (Error Correction Design). Its target was to identify the errors produced by filtering combinations in such a way so not to remove the winning combination if possible and it was self-adaptive in time. The idea of this system occcurred simply because one filter alone is not enough for good filtering (unless if set too agressive) and setting up several filters have the problem, 90% of the filters to set them correctly but the other 10% will filter out the winning combination. My tests were mainly performed on a 5/45 game (1221759 combinations) and the results were elimination 95-99% of the tickets and keep the winning combination 20% of the time. The problem here is that the remaining 10000-50000 tickets are still too many to be played. If my initial combinations are a full wheel, this method returns proportionally to its size acceptable tickets. So, my conclusion is, such an approach is only viable to pools of players who want to spent good money but within one month they are due to win the jackpot. For players who want to spend a few dollars, the only other method I have seen producing results is what my users do (wheel around 12 numbers, match 4-5 correct numbers every so often and the cost to play is about $8-12 for the wheel).

    Practically everything you say about filtering is true. However, I have achieved a much better range of remaining tickets by using different filtering criteria than you probably used (1k to 2k on a 5/39 game with about a 5% chance of eliminating the winning combination). And by using 0% error filters I have reduced the amount of filter setting to a minimum (that 5%). Like what you ran into, it is still too many combinations to bet on for stand- alone system. That is part of the reason for this thread.

    By the way, I’ve always appreciated all your posts. Believe me when I say that even though I am trying a different approach, that I don’t have the highest respect for you and the quality work you do.

     

    To get back on topic…

    Is a filter/reduction method good if it doesn’t clearly indicate the best numbers or combinations to bet on?

    You are a slave to the choices you have made.  jk

    Even a blind squirrel will occasioanlly find an acorn.

      RL-RANDOMLOGIC's avatar - usafce

      United States
      Member #59354
      March 13, 2008
      3983 Posts
      Offline
      Posted: July 15, 2012, 3:43 am - IP Logged

      JKING

      What you seem to be asking has no answer,  a filter is a filter.  Reduction filters all do one thing, they reduce.

      The only good filter is a filter who's value can be predicted and at the end of the day every filter does the exact

      same thing.  It does not really matter how or what a filter does  it only matters that it is set correctly.  Math

      cannot predict the lottery but we can gain some good insight as to what is expected.  If given a choice of using

      math or instinct I would choose instinct over math for any lottery.   You cannot change the odds for the game

      just like you can't squeeze apple juce from a peach.   I believe the human element plus faith plus hope plus a 

      number of other things are all factors in how one does.   Two people can use the same software or method and

      one will do very well and other very poorly.  This has no bearing on how smart a person is or how much math is

      applied.  If you want to take lottery prediction to the next level try a supernatural or mistical approach.   There

      are math based systems that will give you the most bang for the buck but all require some sort of reduction in

      the first stages, this includes wheels.  Wheeling 12 numbers taken from a pool of 59 is reducing the pool by 47

      numbers as the first stage.  From that point on math can be used to ensure a certian prize if part or all of the

      winning numbers are in the 12 selected.  I win many prizes and have trapped many JP's within a very small

      percent of the total possible but out of all those times I have only done it once when I held the tickets in my

      hand.  I do far better than the odds would suggest and it is not because of the software I use it is because I

      make good guesses.  I also think that anyone can do this that does not rely on math,  as I have said before

      If you use math to make a selection then expect the results to follow the odds.  It seems to me that everyone

      with a system falls just short of defining the only real bit of data that is truly needed.  It's not the system it's 

      the ability of the user to make good guesses for the static set of choices the system offers.  Any system should

      be able to reduce any matrix to one winning line providing the choices are correct.  For me it is the last 5% that

      seems to form the barrier that keeps us seperated from the JP's

      RL

      Working on my Ph.D.  "University of hard Knocks"

      I will consider the opinion that my winnings are a product of chance if you are willing to consider

      they are not.  Many great discoveries come while searching for something else

      USAF https://en.wikipedia.org/wiki/Prime_Base_Engineer_Emergency_Force

        US Flag Trump / 2016 & 2020  

        RL-RANDOMLOGIC's avatar - usafce

        United States
        Member #59354
        March 13, 2008
        3983 Posts
        Offline
        Posted: July 15, 2012, 4:03 am - IP Logged

        JKING

        After reading your last post I think I know what you are attempting.  I followed a similar approach

        but hit a wall where I could make no futher progress.   This would be the ultimate lottery program

        if it's possible.  I just don't have the enegry or abilities to see it to the end, I wish you success.

        RL

        .

        Working on my Ph.D.  "University of hard Knocks"

        I will consider the opinion that my winnings are a product of chance if you are willing to consider

        they are not.  Many great discoveries come while searching for something else

        USAF https://en.wikipedia.org/wiki/Prime_Base_Engineer_Emergency_Force

          US Flag Trump / 2016 & 2020  

          JKING's avatar - Kaleidoscope 3.gif

          United States
          Member #5599
          July 13, 2004
          1185 Posts
          Offline
          Posted: July 15, 2012, 6:37 am - IP Logged

          JKING

          What you seem to be asking has no answer,  a filter is a filter.  Reduction filters all do one thing, they reduce.

          The only good filter is a filter who's value can be predicted and at the end of the day every filter does the exact

          same thing.  It does not really matter how or what a filter does  it only matters that it is set correctly.  Math

          cannot predict the lottery but we can gain some good insight as to what is expected.  If given a choice of using

          math or instinct I would choose instinct over math for any lottery.   You cannot change the odds for the game

          just like you can't squeeze apple juce from a peach.   I believe the human element plus faith plus hope plus a 

          number of other things are all factors in how one does.   Two people can use the same software or method and

          one will do very well and other very poorly.  This has no bearing on how smart a person is or how much math is

          applied.  If you want to take lottery prediction to the next level try a supernatural or mistical approach.   There

          are math based systems that will give you the most bang for the buck but all require some sort of reduction in

          the first stages, this includes wheels.  Wheeling 12 numbers taken from a pool of 59 is reducing the pool by 47

          numbers as the first stage.  From that point on math can be used to ensure a certian prize if part or all of the

          winning numbers are in the 12 selected.  I win many prizes and have trapped many JP's within a very small

          percent of the total possible but out of all those times I have only done it once when I held the tickets in my

          hand.  I do far better than the odds would suggest and it is not because of the software I use it is because I

          make good guesses.  I also think that anyone can do this that does not rely on math,  as I have said before

          If you use math to make a selection then expect the results to follow the odds.  It seems to me that everyone

          with a system falls just short of defining the only real bit of data that is truly needed.  It's not the system it's 

          the ability of the user to make good guesses for the static set of choices the system offers.  Any system should

          be able to reduce any matrix to one winning line providing the choices are correct.  For me it is the last 5% that

          seems to form the barrier that keeps us seperated from the JP's

          RL

          Hi,

            Your replies are appreciated and respected as always.

            You might want consider filter/reduction methods from this point of view...

            Even time that you use a filter/reduction method you are saying that you are willing to lose a certain amount of games for the reduction achieved. Unless, of course, you take the 100% approach.

            Some filter/reduction methods give you very little reduction for the risk taken. So, in essence, there is a risk management issue to be delt with.

            Then there is the aspect of the how interaction of multiple filter/reduction criteria impact each other. If the filter/reduction types are to similiar, the reduction amount from testing a filter/reduction inividually, may be drastically reduced from what is expected because the other filter/reduction types have overlapping properties with the initial filter/reduction method. You are then left with ineffective reductions with all the risk.

            By reducing the amount of filter values that have to be guess at, the better the chances of winning are (assuming you are working with a winning system).

            When it comes to the odds, I think we both agree that you can enhance your chances by taking advantage of consistencies in the measurable data from previous draws. Any lottery system can only take you so far until uncertainty takes over. I'm not quite sure where that point is yet. Maybe its that last pesky 5% you are having trouble with. *L* In addition if you add the fact that the lotteries use mutiple machines and computers it makes it even more interesting. The flip side is that in thier monitoring for bias in the draws, there are constraints to what they willing to let happen over multiple draws. Anyway, my goal is create an automated and optimized system, as much as possible, that takes you up to that point of uncertainty. If we are all lucky, that uncertain area won't require more than 12 numbers or a small amount of combinations to cover.

          You are a slave to the choices you have made.  jk

          Even a blind squirrel will occasioanlly find an acorn.

            Avatar
            bgonçalves
            Brasil
            Member #92564
            June 9, 2010
            2125 Posts
            Offline
            Posted: July 15, 2012, 8:40 am - IP Logged
            Hello, jking is an example of a lottery 50/5 ok = is an example to demonstrate the útilidade crossing the filter "sum of numbers, with the endings."
            First, you must choose at least 5 to 6 terminations, mark them in the table, all the numbers chosen endings.
            Second, choose at least 5 to 6 sums, and just check the numbers that match the treminações, e.g. choose termination 1 (1, 11, 21, 31, 41), then chose 2 and 5 checksums, so matching numbers are 11 and 41, right.
            Finally just select and transcribe them down, in order to unfold or how to understand play with them.
            If the numbers drawn match the endings, with the sum of the numbers ", they are in this group.
            To test in the last draw:
            paragraphs 4, 16, hit: 17, 20.29
            Outputs: 0-endings 4-6-7-9
            Sum of paragraphs: 2 (20.29)
            4 (04)
            7 (16)
            8 (17)
            In the final paragraphs selected: 4.7 .16 .17 .20 .26 .29 .34 .40 .44 .47 .49. total 12n paragraphs.
            In this case, 5 4 útilizamos endings and sums, it is certain that more more paragraphs are selected.
              RL-RANDOMLOGIC's avatar - usafce

              United States
              Member #59354
              March 13, 2008
              3983 Posts
              Offline
              Posted: July 15, 2012, 11:00 am - IP Logged

              Hi,

                Your replies are appreciated and respected as always.

                You might want consider filter/reduction methods from this point of view...

                Even time that you use a filter/reduction method you are saying that you are willing to lose a certain amount of games for the reduction achieved. Unless, of course, you take the 100% approach.

                Some filter/reduction methods give you very little reduction for the risk taken. So, in essence, there is a risk management issue to be delt with.

                Then there is the aspect of the how interaction of multiple filter/reduction criteria impact each other. If the filter/reduction types are to similiar, the reduction amount from testing a filter/reduction inividually, may be drastically reduced from what is expected because the other filter/reduction types have overlapping properties with the initial filter/reduction method. You are then left with ineffective reductions with all the risk.

                By reducing the amount of filter values that have to be guess at, the better the chances of winning are (assuming you are working with a winning system).

                When it comes to the odds, I think we both agree that you can enhance your chances by taking advantage of consistencies in the measurable data from previous draws. Any lottery system can only take you so far until uncertainty takes over. I'm not quite sure where that point is yet. Maybe its that last pesky 5% you are having trouble with. *L* In addition if you add the fact that the lotteries use mutiple machines and computers it makes it even more interesting. The flip side is that in thier monitoring for bias in the draws, there are constraints to what they willing to let happen over multiple draws. Anyway, my goal is create an automated and optimized system, as much as possible, that takes you up to that point of uncertainty. If we are all lucky, that uncertain area won't require more than 12 numbers or a small amount of combinations to cover.

              JKING

              I have tried many different approaches to overcome uncertainty and have adjusted my thought process

              to exclude it.  How can we be certain about anything in a random process.  I got to a point where I needed

              to be correct which placed so much stress on me that I often found myself doing anything to avoid making a

              setup where I might make a mistake. 

              Since anything can show in the next draw I try to look at the data like selecting a peice of fruit.  Would I like

              an apple or a banana?   I like both but which one would I enjoy most at that moment.   Once I moved into this

              sort of mindset I began to make far better choices and playing the lottery became fun again.  Taking on random

              is like fighting an opponent with 10 arms, every time you think you have a chance a fist comes out of the blue and

              knocks you for a loop.  This may not be for everyone but it works for me.  The difference between a 4of5 and a 5of5

              is huge and I often I miss it by one setting. 

              My goal is to play until I get everything correct.  I have done it in the past so in my thinking I can do it again.  I still

              have to analyze the data but don't feel the need to be correct before I can play.  I have had several games in a row

              where I have matched at least 3 of 5 and hope it continues.  I have not won that much in the last several games

              but I am still ahead and have done way better than what the odds would suggest.   Filters are the killers of jackpots

              and a person should seek to understand every aspect of using them.  Many of the filters I use in my software overlap

              and some can be backed off all the way while adding just a few lines to the total.  If people would play with their filters

              setups after the draw they would often find one filter that would have made a big difference.  I have a program that will

              generate a list of all the 4of5 for the last draw along with the filter values for each.   Many would be amazed at how many

              filters can be missed and still hit a second level prize.  Missing 5 to 8 filters can often can produce more prizes then missing

              one with the exception of digits and groups. 

               

              RL

              Working on my Ph.D.  "University of hard Knocks"

              I will consider the opinion that my winnings are a product of chance if you are willing to consider

              they are not.  Many great discoveries come while searching for something else

              USAF https://en.wikipedia.org/wiki/Prime_Base_Engineer_Emergency_Force

                US Flag Trump / 2016 & 2020  

                SergeM's avatar - slow icon.png
                Economy class
                Belgium
                Member #123700
                February 27, 2012
                4035 Posts
                Offline
                Posted: July 15, 2012, 1:10 pm - IP Logged

                My preference for playing goes to a different kind of wheel.

                I filtered a 12 number wheel in pick 6.

                N1N2N3N4N5N6
                123789
                1247810
                1257811
                1267812
                1347910
                1357911
                1367912
                14571011
                14671012
                15671112
                2348910
                2358911
                2368912
                24581011
                24681012
                25681112
                34591011
                34691012
                35691112
                456101112

                 

                S.M.