Welcome Guest
Log In | Register )
You last visited January 22, 2017, 7:16 pm
All times shown are
Eastern Time (GMT-5:00)

Question for Ion Saliu

Topic closed. 25 replies. Last post 12 years ago by duk56.

Page 2 of 2
PrintE-mailLink
Avatar
Auckland
New Zealand
Member #2247
September 4, 2003
33 Posts
Offline
Posted: November 12, 2004, 3:30 pm - IP Logged

Hi foxy,

I have tested your data file and get the same Ion values for 100 draws or 200 draws?.

Am using Windows 98SE.

Spot9

    Avatar
    Auckland
    New Zealand
    Member #2247
    September 4, 2003
    33 Posts
    Offline
    Posted: November 12, 2004, 5:11 pm - IP Logged

    I get the same values with WindowsXP as I got with Windows 98SE for 100 and 200 draw statistics reports.

    Trying the 49 number data file, the Ion4 value has to be multiplied by 12 and the Ion5 value by 24 for a 49 number lottery to give the actual value.

    Its best not to set the maximum value for the Ver4 filter, for sometimes it has to be set much higher than the minimum value of the Ver5 filter to pass the winning combination.There must be something wrong with the Ver4 filter?.

    Spot9

      Avatar
      New Member

      Canada
      Member #8436
      November 5, 2004
      17 Posts
      Offline
      Posted: November 15, 2004, 10:20 am - IP Logged

      Ive been getting some private messages regarding Ion5 issues. The following document contains all the answers. Pay close attention to Tip section at the end of the document.

      http://saliu.com/mdi_lotto.html


        United States
        Member #1940
        July 29, 2003
        16 Posts
        Offline
        Posted: November 15, 2004, 10:38 am - IP Logged

        " There are a few issues. They may not be real issues all the time, after all.

        1) Ion5 creates confusion with small data files. I recommend a SIM-6 file of at least 200,000 lines (two hundred thousand). If a filter shows the same value many times, it is a sign of insufficient data file.

        2) The user gets the correct WS files (winning reports). Then, the user checks if the software generates the winning combination. For example, the user writes down the stats for line #1 (the 1st combo in the data file). Sometimes, to the dismay of the user, the winning combination is not generated. It happens rarely, but it does. My software always uses internal (inner) filters. If the winning combo consists of all odd, or all even, it is automatically removed by the inner filters.

        3) Also, the random generation implemented in my software is optimized. In addition to the inner filters, the software eliminates all 6 (or all 5) groups in the data file. In other words, it is possible sometimes (rarely) that the winning draw is already in the SIM file.

        4) I decided to keep my software work that way, for being the optimal configuration. Yes, there are a few moments when the winning combo is eliminated. But I dont think is reasonable to expect the software win every time. Moreover, the user should not think of using the software once and only one time.

        Please also check as many pages as you can at my web site. Some pages are referred to many times. They deserve special attention. Most issues occur because of superficial treatment of the tutorials. Incredibly, there are still problems with creating the data files! As simple and easy as the data files for my software are, some users still have problems! Sorry, I cant do anything about that.

        Some people (I hope they are joking!) fabricate errors! There are no errors, but they modify the reports or data files to show errors! I think I have good sense of humor problem is, time is insufficient for joking too much!

        Best of luck to you all!

          Avatar
          New Member

          Canada
          Member #8436
          November 5, 2004
          17 Posts
          Offline
          Posted: November 15, 2004, 3:48 pm - IP Logged

          For those who are still interested in MDIEditor;

          I just wanted to share some observations of filter performance in MDIEditor.

            15  17  18  34  39  45
            15  31  32  34  42  47
            35  36  37  46  47  48

          Combinations above are real draws from Canadian Atlantic 649 (starting March 30, 2003). They were drawn sequentially (bottom to top). I am sure you will find similar numbers in your game of choice.

          Statistical report generated on a data file with 228,263 (real + simulated) lines produces the following results for Ion5.

          9503
          9503
          9503+

          Simulated data has been purged against real drawings (all filters disabled) and checked for duplicates. Please note that data file size is sufficient according to authors recommendations.

          If you try to insert combinations mentioned above into your data file you may get different values for corresponding Ion5 values, though I am quite certain that they will be out of range. Test it for yourself.

          We can only guess how this may affect number generation process (filters in particular). I ran tests on three occurrences of a pattern in the lotto game mentioned above. In all three cases Ion5 value which actually passed the winning combination did not match the stat report value (6). In fact, they had to be multiplied by 16, 20, and 12 respectively.

          Results of my tests may be incorrect due to the problem described earlier. You can draw your own conclusions based on information presented here. I personally think that Ion5 does not work correctly when irregular draws are present in data files. I am still looking into this however&

          p.s. Please note that all three combinations have at least two sequential numbers (17-18, 31-32, and 35-36). A brief look at my data file with reference to the latest stat report reveals that every out of range value for Ion5 correspond with a combination which has sequential numbers.

          PM me if you would like a copy of the data file I referred to in this post. Also, this is not a joke.

            Avatar
            New Member

            Canada
            Member #8436
            November 5, 2004
            17 Posts
            Offline
            Posted: November 15, 2004, 6:25 pm - IP Logged

            In your previous post (and your message board) you said that abnormal Ion5 values are due to insufficient data file sizes. It's happening in my data files, even though the size is beyond 200,000 lines. My 'observation' was that it was only happening when irregular combinations were present in the data file (real game data).

            My understanding is that if statistics report produces a filter value it should be read the same way when purging and/or generating combinations. That is not the case.

            I am not interested in inner-workings of the filters. I just want to get consistent results.

              Avatar
              Auckland
              New Zealand
              Member #2247
              September 4, 2003
              33 Posts
              Offline
              Posted: November 16, 2004, 7:56 pm - IP Logged

              Hi,

              These are three results from the NZ lotto:

                      2 20 23 27 35 37

                      1 6 16 20 38 40

                      22 24 30 31 34 40

              They all pass the purge test and the first two generate the optimized winning combination ONLY, within 10 minutes, but the last combination will never generate any combinations even if the computer is run for hours the only difference is that it contains an adjacent pair 30-31. If the 31 is changed to 32 and the 22 changed to 21 ( to avoid having all even numbers and rejection by the internal filters )it generates the winning combination within 2 minutes!.

              Spot9


                United States
                Member #1940
                July 29, 2003
                16 Posts
                Offline
                Posted: November 17, 2004, 10:38 am - IP Logged

                First, my software did not enable inner filtering. Then, I noticed the statistical configuration of the lottery drawings. I added inner filtering, a feature that enhanced random generating tremendously. I was happy with the tradeoff. I knew that a tiny percentage of combinations would be left out.

                In hindsight, I should have come up with the brilliant idea of creating a switch. The software should have asked the user to enable the inner filters OR not. I think about that now. There is no deadline, however. I dont know when (even if) Id enable such a feature in my software.

                Until then, I advise to put to good use filtering with more normal values (within the normal probability rule). Lets put aside extremely extreme filter values, even if some of them still work. I wonder why a combo like this one in NZ lotto 6/40 (is it 6/40?) was left out:
                22 24 30 31 34 40
                It should be in the field. Its sum-total is inside the bell; it only has 2 consecutive numbers (the soft allows for 3, sometimes 4 consecutive numbers). Perhaps there is another
                22 24 30 31 34 40 in the SIM-6 file; or a combo like 21 23 29 30 33 39.

                I want to bring closure to this thread. Thanks everyone for your feedback!
                 

                  Avatar
                  New Member
                  San Francisco
                  United States
                  Member #7795
                  October 12, 2004
                  4 Posts
                  Offline
                  Posted: November 22, 2004, 10:59 pm - IP Logged

                  Ion

                  I am moving to Florida and my hope is that some day you will update and have a Pick432.exe on your site. I have said this before to everyone if you just keep going thru the software you will find how to work the ins & outs of all of the program. It's free, it's fun and you just might make money if you stick with it.

                   

                                                                John Civitan

                   

                    Avatar
                    New Member
                    San Francisco
                    United States
                    Member #7795
                    October 12, 2004
                    4 Posts
                    Offline
                    Posted: December 14, 2004, 6:00 pm - IP Logged

                    Ion

                    Thanks for sending the programs. You and I have allways had a good working understanding with each other. I wish you good luck with the updating of your programs. Thanks Again.

                                                                          John

                     

                      Avatar
                      New Member

                      Greece
                      Member #10145
                      January 3, 2005
                      1 Posts
                      Offline
                      Posted: January 9, 2005, 1:19 am - IP Logged

                      A question about new Ions Pick332.I use the Lotwon3 but the problem is with the Pick332 Wreports,the values of sync1 ,tot1,rev1,any1 are diferent in the Pick332 for the same draw.So my strategies are useless for the Pick332.Anybody,Ion,any sugestion?