Welcome Guest
Log In | Register )
You last visited December 10, 2016, 9:06 pm
All times shown are
Eastern Time (GMT-5:00)

Neural Net Lottery Picker

678 replies. Last post 11 hours ago by lakerben.

Page 44 of 46
511
PrintE-mailLink
MillionsWanted's avatar - 24Qa6LT

Norway
Member #9517
December 10, 2004
1272 Posts
Online
Posted: September 30, 2016, 8:59 am - IP Logged

Congrats!!!.

You have found something interesting in your bets with "lowest total error".

Regards.Smile

Thanks.

Regarding lowest Total Error. I took notice of this post from Stoopendaal about errors about half a year ago. Before that I used NNLP to find 4-5 lines and then start it over again. Now I let it run until it has found 8-10 lines instead, before starting the predictions again.

It's all about getting as low total errors as possible. The faster the processor, the better. Unfortunately I am using a rather mediocre AMD processor in my PC(FX-6300).

    Avatar
    Krakow
    Poland
    Member #86302
    February 2, 2010
    860 Posts
    Offline
    Posted: October 1, 2016, 2:55 pm - IP Logged

    Just my opinion but I think a filtered RNG would do as well as this program.  As I see from the post, there are several

    user settings and if a person does manage to pull out some kind of a win then don't expect the wins to continue.  The

    lottery is random and no NN is going to predict random.   Play with the settings long enough and it's bound to produce

    a few wins but it's not the NN that is predicting, it's a combination of the user settings and chance.   I have written many

    such programs and used a few commercial programs in the past and none of them were worth the effort.  I have not seen

    the author posting anything for quite some time and I think it's for good reason, he / she came to the same conclusion. 

    Not saying it won't win here and there as even a broken clock gives the correct time twice a day.   Just my two cents here

    but I think it's a waste of time.   Maybe if the author could add the ability to track progress so that the user could know

    exactly what settings would have produced a better outcome would be a step up but given the millions of different choices

    I don't think that would be possible.   IMHO if you can't track where you went wrong then there is no way to do better the

    next game.

    RL

    I couldn't agree more with that that if there's no way to check what settings or change of a parameter would have given a good result it can in now way make our play better.

    Adam

      wander73's avatar - Lottery-038.jpg
      Philadelphia, PA
      United States
      Member #153774
      March 24, 2014
      925 Posts
      Online
      Posted: October 7, 2016, 9:43 am - IP Logged

      Neural Networks and the applicability to lotteries has been discussed many times here on Lottery Post.

      Here some comments on NN here on Lottery Post of the past years: 

      "You are quite correct in your comments that neural networks will not predict the winning numbers everytime. They are however very powerful prediction tools that when used with other ai toolsets can reduce your odds and give you more winnings than other methods."

      "If the lottery is not random enough they might be able to find a pattern if not, it's just like guessing"

      "Most of the work of neural networks is going into pattern recognition and simulations. Neural networks are trained rather than programmed, and potentially will have the most adaptability of any expert system, but like humans they can make mistakes."

      "Neural networks are particularly adept at solving problems that cannot be expressed as a series of steps. Neural networks are particularly useful for recognizing patterns, classification into groups, series prediction and data mining."

      "There is no such thing as AI or NI when it comes to lotteries, anyone saying something different I challenge them to prove it"

      "AI and NN along with many other methods of prediction can be applied to lotteries, just don't expect
      them to work. I have written at least a dozen of each and the lottery is just too random. Trying to
      predict the lottery is a waste of time IMHO but that does not mean that we can't play smarter. Study
      the data and then make your best guess. Forget the odds, don't apply to much math and trust your
      gut instincts. AI and NN can be used to make forecast for some events like sales, stocks, profits etc...
      but these events are not based on random. Math can be used to test the randomness of a game but
      it will never be able to predict random"

      The last couple of months I did a research on this subject and wrote a simple program (freeware) using a neural network generating lottery picks (predictions?) using the history draw results of a lottery game (via import feature in the tool).

      So now you have the opportunity testing a Neural Network algorithm for your lottery game(s) Dance

      Information and instructions can be found in the tool.

      You can download tool here: http://intelbet.somee.com/tools/nnlp.rar (zipped file)  or  http:/intelbet.somee.com/tools/nnlp.exe (exe file)

       

      Some screen shots (From Windows 10):

      Generate Picks

      Generate Picks 2

      Neural Network Settings

      Note: I'm not asserting you can predict the/some winning numbers for you lottery game with this tool.

      About the tool:

      - Obusfacted with Confuserex

      - No installation required, just put the executable somewhere

      - No changes are made to your registry

      Good luck!

      I already have figured out the megamillions and powerball.  Funny that the current month date and yr have something to do with it.   Nothing to download.  Simple math.

        Avatar
        Krakow
        Poland
        Member #86302
        February 2, 2010
        860 Posts
        Offline
        Posted: October 7, 2016, 3:24 pm - IP Logged

        I already have figured out the megamillions and powerball.  Funny that the current month date and yr have something to do with it.   Nothing to download.  Simple math.

        Ok, another guy with new-fanged lottery math.

          lottoburg's avatar - wiggle
          NYC
          United States
          Member #54483
          August 20, 2007
          886 Posts
          Offline
          Posted: October 8, 2016, 12:29 am - IP Logged

          Ok, another guy with new-fanged lottery math.

          LOL ROFL LOL

          • Saying Lotto #s Can be Predicted means that Lotto's ODD can be reduced down  to an economical level by a system.
          • Saying a Lotto System Works means that we can win constantly (not each draw)  and economically (get a real profit) by using the System.
          • Practice is the only criterion for testing truth.
            MillionsWanted's avatar - 24Qa6LT

            Norway
            Member #9517
            December 10, 2004
            1272 Posts
            Online
            Posted: October 17, 2016, 9:15 am - IP Logged

            Back with a new strategy. Based on only playing lines with as low Total Error rates as possible.

            I generate for example 120 lines in a lottery and only pick the lowest 40(33.3%) to 48(40%) lines for playing. It is a risk to take, because it's not impossible there will be winners among the 80 lines which you ignored/didn't play. If I generate 150 lines I would pick the lowest 50 to 60 lines for playing.

            I used this strategy for Saturday lotto(7/34) on October 15th. Generated 120 lines and picked the 40 lines with lowest Total Error.

            I won three 4/7 and one 5/7. Total prize win was NOK 260(about US$31). I spent NOK 200(about US$24). I used the settings for VikingLotto on page 18, which seems to work fine for Saturday lotto as well.

              Avatar
              Krakow
              Poland
              Member #86302
              February 2, 2010
              860 Posts
              Offline
              Posted: October 19, 2016, 9:19 am - IP Logged

              Back with a new strategy. Based on only playing lines with as low Total Error rates as possible.

              I generate for example 120 lines in a lottery and only pick the lowest 40(33.3%) to 48(40%) lines for playing. It is a risk to take, because it's not impossible there will be winners among the 80 lines which you ignored/didn't play. If I generate 150 lines I would pick the lowest 50 to 60 lines for playing.

              I used this strategy for Saturday lotto(7/34) on October 15th. Generated 120 lines and picked the 40 lines with lowest Total Error.

              I won three 4/7 and one 5/7. Total prize win was NOK 260(about US$31). I spent NOK 200(about US$24). I used the settings for VikingLotto on page 18, which seems to work fine for Saturday lotto as well.

              Actually not so new as rather consistent with your opinion that the lowest total errors give the best chances for a hit. Perhaps the new thing is that you take more lines to play. Still it does not change my view that with no way to track the best settings for the games you cannot improve anything in your play.

                lottolot's avatar - Sphere animated2.gif
                Tanhauser Gates
                Holy See (Vatican City State)
                Member #139281
                February 18, 2013
                173 Posts
                Offline
                Posted: October 19, 2016, 8:35 pm - IP Logged

                Back with a new strategy. Based on only playing lines with as low Total Error rates as possible.

                I generate for example 120 lines in a lottery and only pick the lowest 40(33.3%) to 48(40%) lines for playing. It is a risk to take, because it's not impossible there will be winners among the 80 lines which you ignored/didn't play. If I generate 150 lines I would pick the lowest 50 to 60 lines for playing.

                I used this strategy for Saturday lotto(7/34) on October 15th. Generated 120 lines and picked the 40 lines with lowest Total Error.

                I won three 4/7 and one 5/7. Total prize win was NOK 260(about US$31). I spent NOK 200(about US$24). I used the settings for VikingLotto on page 18, which seems to work fine for Saturday lotto as well.

                Congrats!!!Thumbs Up

                5/7 is a very good prize.

                Regards

                  MillionsWanted's avatar - 24Qa6LT

                  Norway
                  Member #9517
                  December 10, 2004
                  1272 Posts
                  Online
                  Posted: October 19, 2016, 8:45 pm - IP Logged

                  Congrats!!!Thumbs Up

                  5/7 is a very good prize.

                  Regards

                  Yes Nod Yeah... But it is now the hunt for 6/7 and 6+Bonus/7 starts. Only playing lines with low Total Error rates might help. We'll see.

                    Avatar

                    United States
                    Member #139779
                    March 3, 2013
                    115 Posts
                    Offline
                    Posted: October 20, 2016, 3:28 am - IP Logged

                    Actually not so new as rather consistent with your opinion that the lowest total errors give the best chances for a hit. Perhaps the new thing is that you take more lines to play. Still it does not change my view that with no way to track the best settings for the games you cannot improve anything in your play.

                    "Still it does not change my view that with no way to track the best settings for the games you cannot improve anything in your play." Just a thought but NNLP saves settings to note pad when you save numbers/predictions? Mabe importing the text files to spreadsheet might be a way to sort of keep track and sort out which settings work best for what conditions. Just an idea thou.

                      Avatar
                      Krakow
                      Poland
                      Member #86302
                      February 2, 2010
                      860 Posts
                      Offline
                      Posted: October 20, 2016, 3:34 am - IP Logged

                      "Still it does not change my view that with no way to track the best settings for the games you cannot improve anything in your play." Just a thought but NNLP saves settings to note pad when you save numbers/predictions? Mabe importing the text files to spreadsheet might be a way to sort of keep track and sort out which settings work best for what conditions. Just an idea thou.

                      Not quite what was on my mind.

                      Say you use learning rate of 1E-05 ,threshold of 0.24 and hidden neutrons of 200, then after the draw you get the report on what training data range you should have used to get a 5 of 6 or so.

                        Avatar

                        United States
                        Member #139779
                        March 3, 2013
                        115 Posts
                        Offline
                        Posted: October 20, 2016, 4:08 am - IP Logged

                        Not quite what was on my mind.

                        Say you use learning rate of 1E-05 ,threshold of 0.24 and hidden neutrons of 200, then after the draw you get the report on what training data range you should have used to get a 5 of 6 or so.

                        I was thinking in terms of using historical data to get a general consensus of which settings have been most productive. Then using those results of the lowest total error rates to see if they match anything promising. But then again I am not using it to look for the actual numbers just the sums. There are fewer sums than there are number combinations possibly making it easier to point in the direction of a hit.

                          MillionsWanted's avatar - 24Qa6LT

                          Norway
                          Member #9517
                          December 10, 2004
                          1272 Posts
                          Online
                          Posted: November 12, 2016, 7:44 pm - IP Logged

                          Bumping thread.

                            lottolot's avatar - Sphere animated2.gif
                            Tanhauser Gates
                            Holy See (Vatican City State)
                            Member #139281
                            February 18, 2013
                            173 Posts
                            Offline
                            Posted: November 13, 2016, 4:59 am - IP Logged

                            Bumping thread.

                            Thumbs Up.

                            Easy thing is not to read things that a person does not like ..

                            Some people only write in  the threads to criticize and discredit the creator:Bad method, Bad program,contains viruses... 
                            They get that generous people to get tired and stop contributing with programs,methods,etc.
                            If someone does not like this thread, it is very easy not to read it.
                            Regards (Specially to MillionsWanted and Harmen,of course).
                              Gambler667's avatar - shapes swish.jpg

                              Australia
                              Member #166177
                              May 8, 2015
                              29 Posts
                              Offline
                              Posted: November 18, 2016, 5:32 pm - IP Logged

                              I tottaly agree with you, lottolot. It seems people here does not like sharing their methods and only give critics to those who contributes with programs etc. 

                              Gambler667

                                 
                                Page 44 of 46