Welcome Guest
Log In | Register )
You last visited January 21, 2017, 10:35 pm
All times shown are
Eastern Time (GMT-5:00)

SoniQ BOOM, a new P3/P4 system (Excel)

Topic closed. 116 replies. Last post 11 years ago by xavier102772.

Page 6 of 8
PrintE-mailLink
Avatar
Calgary
Canada
Member #9419
December 7, 2004
228 Posts
Offline
Posted: January 5, 2006, 2:41 pm - IP Logged

More interesting things about this system is the fact you can modify the draw history to check back various amounts and increase the number of wins substantially.  I just did a small test on using only the past 2000 draws.  Not a rolling 2000 draws like soniq's new sheets are, but just 2000 draws.  It came back with approximately the same amount of hits, with approximately 1/4 of the hits being the same as the original draw sheet and the others being new hits.  So essentially I had 47 hits in my original file of 3240 and 31 hits in the 2000 draw file.  Extrapolated they are about the same number of hits. 

Say you had 10 files with differing draw history lengths that you updated.  This would give you approximately 30 hits in each file of 2000, 7-9 of which are the same as in the other files.  Anyway, you would likely be around 200 hits for 2000 draws, which is approximately 1 hit for every 10 draws or less(10*10=100 layout for a $500 win).  Sure you'd have to buy 10 tickets every day and essentially it would work out to being the same as buying 1, $10 ticket every day for 100 draws(10*100=1000 layout for a $5000 win), but the frequency of hits would be very high so you wouldn't be waiting long for a hit. 

Anyway, just something I was thinking about. 

    Avatar
    Calgary
    Canada
    Member #9419
    December 7, 2004
    228 Posts
    Offline
    Posted: January 5, 2006, 3:01 pm - IP Logged

    Comparing the two files some more.  The same 3 digit pick3 number appears to be the same 1 out of 7 draws or so. The wins are 7 that are the same picked number in each file. 

    Draws played = 2000/7 =  285

    Amount Wagered = 285*2 = 570 

    Amount Won = 14*500 = 7000

    Win/Loss Ratio = 7000/570 = 12.28 

     

    Amazing. 

      Avatar
      Calgary
      Canada
      Member #9419
      December 7, 2004
      228 Posts
      Offline
      Posted: January 5, 2006, 3:53 pm - IP Logged

      Looking at the skips in my original file.  If you just played 12 draws after a hit, you would have spent $484(played only 484 draws) and would have won 14(14*$500=$7000) times.  That's a 7000/484=17.67 win/loss ratio. 

      Can't wait to see the new spreadsheets from soniq.  : )

        Avatar
        Anna, TX
        United States
        Member #26720
        November 21, 2005
        263 Posts
        Offline
        Posted: January 6, 2006, 1:02 am - IP Logged

        More interesting things about this system is the fact you can modify the draw history to check back various amounts and increase the number of wins substantially.  I just did a small test on using only the past 2000 draws.  Not a rolling 2000 draws like soniq's new sheets are, but just 2000 draws.  It came back with approximately the same amount of hits, with approximately 1/4 of the hits being the same as the original draw sheet and the others being new hits.  So essentially I had 47 hits in my original file of 3240 and 31 hits in the 2000 draw file.  Extrapolated they are about the same number of hits. 

        Say you had 10 files with differing draw history lengths that you updated.  This would give you approximately 30 hits in each file of 2000, 7-9 of which are the same as in the other files.  Anyway, you would likely be around 200 hits for 2000 draws, which is approximately 1 hit for every 10 draws or less(10*10=100 layout for a $500 win).  Sure you'd have to buy 10 tickets every day and essentially it would work out to being the same as buying 1, $10 ticket every day for 100 draws(10*100=1000 layout for a $5000 win), but the frequency of hits would be very high so you wouldn't be waiting long for a hit. 

        Anyway, just something I was thinking about. 

        I was actually thinking about that and was going to test it later. I did try applying vtrac numbers instead and it  came with the same results.

        2000 heh? I wasn't sure where to start. Thanks for that.

        When I  get time I will post you my hit ratio as well as what happens with the new test of 2000. \

        Melissa 

          Avatar
          Calgary
          Canada
          Member #9419
          December 7, 2004
          228 Posts
          Offline
          Posted: January 6, 2006, 1:19 am - IP Logged

          Going from retiring on this system to being dejected by it. :( I think I'm just gonna wait till hypersoniq gives us an update on his progress, because the more I look at it, the more the counting is off. Just for laughs, I modified the counting formula to not count the following digit in it's current counts to get an accurate count. I believe I had an accurate formula. What happened? The hits dropped down to 6 in 3240 draws from the original 47 hits. Not quite break even. I could still be off with my formula. But initially it looks correct. So, I'm with Pogo and that the counting of the future digit in the spreadsheet is predisposing false predictions. Anyone else finding this? Soniq, did you see this effect when you adjusted the counting formula? How are the updates? Let me know. Thanks.

            Avatar
            Anna, TX
            United States
            Member #26720
            November 21, 2005
            263 Posts
            Offline
            Posted: January 6, 2006, 1:23 am - IP Logged

            Going from retiring on this system to being dejected by it. :( I think I'm just gonna wait till hypersoniq gives us an update on his progress, because the more I look at it, the more the counting is off. Just for laughs, I modified the counting formula to not count the following digit in it's current counts to get an accurate count. I believe I had an accurate formula. What happened? The hits dropped down to 6 in 3240 draws from the original 47 hits. Not quite break even. I could still be off with my formula. But initially it looks correct. So, I'm with Pogo and that the counting of the future digit in the spreadsheet is predisposing false predictions. Anyone else finding this? Soniq, did you see this effect when you adjusted the counting formula? How are the updates? Let me know. Thanks.

            hmm.  I think it Bill Gates fault. Maybe we should switch to the other side.

            Ha! 

              Pogo's avatar - wfl
              NC
              United States
              Member #29378
              January 1, 2006
              552 Posts
              Offline
              Posted: January 6, 2006, 1:29 am - IP Logged

              Startling News for my trials... ...VA numbers I'm using between 6 months (160 draws) & 9 months (235 draws).

              Without using my corrected data, I'm averaging 82 out of 160 draws & 120 out of 235 draws.

              I told you previously about adding an additional line to highlight when a draw has a tie. Well, when I use this data it corrects the few numbers (usually 1 number wrong for each pick 3) and I end up averaging 160 out of 160 & 235 out of 235.

              Again all of this data is using only 6 or 9 months worth of back data.  The true test will be tomorrow - I'm wagering $12.00 on 12 different straight number combinations I came up with from Soniq's formulas (midday - pick 3). And, I'm wagering $27.00 on 27 different straight number combinations for Evening - Pick 3.

              Wish me luck!!! I hope it works - I'll let ya'll know something tomorrow... :-) Pogo 

                Avatar
                Anna, TX
                United States
                Member #26720
                November 21, 2005
                263 Posts
                Offline
                Posted: January 6, 2006, 1:31 am - IP Logged

                Startling News for my trials... ...VA numbers I'm using between 6 months (160 draws) & 9 months (235 draws).

                Without using my corrected data, I'm averaging 82 out of 160 draws & 120 out of 235 draws.

                I told you previously about adding an additional line to highlight when a draw has a tie. Well, when I use this data it corrects the few numbers (usually 1 number wrong for each pick 3) and I end up averaging 160 out of 160 & 235 out of 235.

                Again all of this data is using only 6 or 9 months worth of back data.  The true test will be tomorrow - I'm wagering $12.00 on 12 different straight number combinations I came up with from Soniq's formulas (midday - pick 3). And, I'm wagering $27.00 on 27 different straight number combinations for Evening - Pick 3.

                Wish me luck!!! I hope it works - I'll let ya'll know something tomorrow... :-) Pogo 

                explain? more?

                  Avatar
                  Calgary
                  Canada
                  Member #9419
                  December 7, 2004
                  228 Posts
                  Offline
                  Posted: January 6, 2006, 1:37 am - IP Logged

                  Pogo, I really wouldn't be wagering anything just yet unless you fixed the way the spreadsheet does it's counting. I hope you win Pogo, but I'm pretty sure you won't be winning tomorrow. Not to be cynical or negative, but I think patience and looking more closely at soniq's spreadsheets and waiting till soniq posts back on here would be prudent.

                    Pogo's avatar - wfl
                    NC
                    United States
                    Member #29378
                    January 1, 2006
                    552 Posts
                    Offline
                    Posted: January 6, 2006, 1:49 am - IP Logged

                    Well, not really much to tell, but I found, if you provide about 6 to 9 months worth of data prior to the date you want it to select numbers for. After you hit F9 and it recalculates, you pretty much have a choice of 2 numbers for each digit. And, when you have only 2 picks for each digit, it only takes 8 combinations to cover all possible combinations. Therefore, if the first digit is a 1 or 2, the second digit is a 3 or 4, and the last digit is a 5 or 6. You get this for combinations: 135, 136, 145, 146, 235, 236, 245, 246. Maybe this will work or maybe not, but I think just taking the most recent data will work better than taking 5 years worth of data. That's why near the beginning of Soniq's spreadsheet you have a hug number of hits and then it trails off. Basically, I'm thinking either the data is stale or corresponding numbers change with time. Either way Soniq's piece of software is a success as long as tomorrow's trial goes to plan. If not, we got some more tweaking to do.

                     Regardless, Soniq - an awesome piece of work! Thanks!

                    If everything works out tomorrow, I'll post the changed excel file on my friend's blog and link it here. 

                     

                      Pogo's avatar - wfl
                      NC
                      United States
                      Member #29378
                      January 1, 2006
                      552 Posts
                      Offline
                      Posted: January 6, 2006, 1:51 am - IP Logged

                      That's a good idea. I'll just make tomorrow a dry run.

                      But, I have been hand calculating everything to ensure his coding is correct, and thus far - it is dead on.

                      Love the input everyone!

                      Thank You,

                      Pogo 

                        Avatar
                        Calgary
                        Canada
                        Member #9419
                        December 7, 2004
                        228 Posts
                        Offline
                        Posted: January 6, 2006, 2:05 am - IP Logged

                        Every time I've done a manual count it's been off. And not just off by the one count of one digit, but in multiple digits counted and it's quite difficult to figure out why the formula isn't working. The reason why there is such a huge number of hits in the beginning is because of the offcounting, especially the future draw count. This skews the results majorly. In the beginning, the huge number of hits occur because the only digits counted are gonna be the following draws digits, and therefore a lot of false false positive hits, up until some actual hits start occurring. FYI: the huge number of hits were in soniq's original spreadsheet formula. He gave us the updated formula to plug in in one of his posts which adjusts the counting so that there are not such a huge amount of false hits at the beginning, but still the formula is not quite right yet. Hopefully soniq will shed some light on the issue later.

                          Pogo's avatar - wfl
                          NC
                          United States
                          Member #29378
                          January 1, 2006
                          552 Posts
                          Offline
                          Posted: January 6, 2006, 3:08 am - IP Logged

                          Thanks for the info. Xavier, Pogo

                            hypersoniq's avatar - 8ball
                            Pennsylvania
                            United States
                            Member #1340
                            April 6, 2003
                            2450 Posts
                            Offline
                            Posted: January 6, 2006, 6:02 am - IP Logged

                            finally back online... installed a new LinkSYS cablemodem. (old modem failing intermittently)

                            went from 1.7 megabits/sec up to 6 megabits/sec... blazing speed for under $80 ;-)

                            I really need to figure the count error out. I have an idea, lose the -1 and just look one row above where the formula is, for example, calculating row 2000 make search range from 3 to 1999 (lose the recalc time of the IF)

                            I also would like to get rid of the staleness that kicks in, what if the 0-9 count order (important when a tie exists) was variable? maybe 2 sets of numbers could be used

                            1. count of the announcers for the vertical pair

                            2. frequency count of each digit 0-9 over the last 100 games in each position

                            I should figure out how to code this in visual basic so more test parameters could be tweaked faster... add that to long-term goal list

                            anyway... update plan...

                            I have to work soon, tonight I will replace the sheets on my homepage with the 4 variants of the SB3, so anyone who wants to see what is on there now, get it before 8pm tonight because they need to go to make room.

                            pogo/xavier/hareless thanks for working with the sheet so in-depth and finding errors, that's the exact reason I decided to release it before it was finished! It's great that we're making progress... every little bit helps... I think for now, we will concentrate entirely on optimizing the pick3, since the pick4 is an easy update later (the inherent benefit of a positional system)

                            thanks for the patience while I got my internet probs sorted out... at 6mbps it's ALL good now ;-)

                            Playing more than one ticket per game is betting against yourself.

                              Zoozie's avatar - back
                              South Fort Myers
                              United States
                              Member #26835
                              November 23, 2005
                              454 Posts
                              Offline
                              Posted: January 6, 2006, 7:34 am - IP Logged

                              Hypersoniq - Great Job!  I do hope the new version of your program includes "basic" set up/use for the non-excel folks like me.  Seeing all of you put your heads together on this project ROCKS!

                              I would help Hypersoniq... but excel is not my cup of tea.  Looking forward to your new update.

                               

                              Zoozie  Dance

                               

                              "Whatever the mind can conceive and believe, it can achieve"
                              Napoleon Hill