Welcome Guest
Log In | Register )
You last visited December 11, 2016, 8:42 am
All times shown are
Eastern Time (GMT-5:00)

Hypothetical question...

Topic closed. 34 replies. Last post 11 years ago by tntea.

Page 2 of 3
PrintE-mailLink

United States
Member #1759
June 29, 2003
1156 Posts
Offline
Posted: August 2, 2005, 11:20 pm - IP Logged

There is no "Holy Grail" of a system that would work for everyone just "Good" systems created by individuals who may or may not share it that may or may not work for others.


    United States
    Member #17555
    June 22, 2005
    5582 Posts
    Offline
    Posted: August 2, 2005, 11:29 pm - IP Logged

    There is no "Holy Grail" of a system that would work for everyone just "Good" systems created by individuals who may or may not share it that may or may not work for others.

    True, but it is fun trying to find one. There will never be a system of all systems, but tools are needed, or analyzation to get one step closer.

    The chase to me, will always be better than the catch.....unfortunately the chase never pays.....

      Pinback's avatar - driver
      Ohio
      United States
      Member #5810
      July 22, 2004
      120 Posts
      Offline
      Posted: August 2, 2005, 11:33 pm - IP Logged

      Ok Pac...

      I just finished creating a database with 10 million random 3-digit numbers.  During a dull conference call tomorrow I'll write a program to run through the Ohio Pick 3 history and for every two drawings check to see if a matching set is in the 10 million database, then see if the next drawing in the lottery matches the next entry in the 10 million database (either straight or boxed).  Then we'll see how many hits and misses we get over time.

       


        United States
        Member #17555
        June 22, 2005
        5582 Posts
        Offline
        Posted: August 2, 2005, 11:43 pm - IP Logged

        Good deal Pinback...It's always fun working with you...it's amazing how you can write a complex program on such a short notice. Let me know....It took my friend 2 hours to do what you just did. Wow. Amazing Pinback.

          lottaloot's avatar - AvatarZ56
          Redford/MI
          United States
          Member #3396
          January 18, 2004
          4867 Posts
          Offline
          Posted: August 3, 2005, 12:20 am - IP Logged

          The Lottery Post is a great place to share ideas and methods.  But... what would YOU do if YOU did it... discovered the "Holy Grail" of playing the lottery.  The method that would not only get you wins on a regular basis, but also win you the JACKPOT at least a couple times a year.  If you discovered that... would you share it here?

          Pick 3 and Pick 4 games aren't an issue, because they pay the same no matter how many people win.  But for Pick 5 games and up, the amount you get depends upon how many people win at the same time.  Would you risk lowering your own winnings because you shared your method and other people using it are winning with you?

          Just curious...

           

          I most certainly would not share it with anyone.

           

          Sorry guys--but the money would be mine

            Pinback's avatar - driver
            Ohio
            United States
            Member #5810
            July 22, 2004
            120 Posts
            Offline
            Posted: August 3, 2005, 12:51 am - IP Logged

            Good deal Pinback...It's always fun working with you...it's amazing how you can write a complex program on such a short notice. Let me know....It took my friend 2 hours to do what you just did. Wow. Amazing Pinback.

            That's why I make the big bucks...  <g>

             

              BobP's avatar - bobp avatar.png
              Dump Water Florida
              United States
              Member #380
              June 5, 2002
              3106 Posts
              Offline
              Posted: August 3, 2005, 2:58 am - IP Logged

              If I had to guess a method that might work, it would be a random seed generator that would generate sets of numbers of a user selectable wheelable field size in an amount equal to a range of history plus one.  It would check these sets against the draws for a user selected quality of match, say 4#'s correct or better in a user selected percentage  of wins against past draws. It would then provide the one additional set that would have been matched against the missing next draw.  The idea is to predict the next step in the random walk based on having correctly predicted a percentage of winning numbers leading up to the next draw.

              For example, it generates 101 sets of 12 numbers against the last hundred draws until it has at least a 4# match in at least 50% of the draws, you then play the last set on a 4if4in12number wheel.  The software could have a graph showing the match ratio as each set were matched against a coresponding draw in succession and automatically start over if the percentage of matches failed.  An AI could be used to tweak the seed generation.


              If you make such a software, you owe me a copy. BobP

                stoopendaal's avatar - archer

                Netherlands
                Member #3476
                January 24, 2004
                212 Posts
                Offline
                Posted: August 3, 2005, 1:03 pm - IP Logged

                Hi BobP,

                A while ago I used the method/idea you mentioned in a program I developed.

                But it didn't make no difference if using the method or just  generate the combinations randomly.

                The was no improvement at all........

                (If you want you can have the program to test it) 

                 

                Bye,

                Harm 

                 

                 

                 

                  Pinback's avatar - driver
                  Ohio
                  United States
                  Member #5810
                  July 22, 2004
                  120 Posts
                  Offline
                  Posted: August 3, 2005, 3:55 pm - IP Logged

                  Very finely put Pinback. I know it's frustrating when a solution or an attempt of one is not readily found. I know it can be difficult at times....I've been there and done that....I can't wait till the person has the time to iron out the wrinkles.

                  This friend only has a limited time to configure my request. He has done a quite fantastic job, despite all the distractions. Thursday is the day which more progress will be implemented, just for the sake of timeliness.

                  To answer Pinback's question....I've asked a friend of mine to run a program of 10,000,000 simulated results in a given lottery with a random number generator at work.

                  The program is still in the debugging-mode. I hope that soon enough my buddy will come through and tell me that the program is really looking back at 10,000,000 situations...

                  The reason for the statement above is that according to him I got 12 three number sets, instead of maybe a hundred or more....There seems to be more de-bugging than I imagined.

                  The command given to the program is the following.....

                  Generate 10,000,000 numbers, randomly....

                  Look back and see what the results would be if considering the last ....two night's drawings....

                  What was initially found was a boxed number , matching the next day's results. Not Bad!!!

                  That is what I wanted the program to look for. The most likely set to arrive after so and so....

                  According to the builder of the program, Not quite confident...still in the works.....

                  Many filters can be applied.

                  Many variables left to fill in.....

                  The reason for doubt on my friend's behalf relies on th notion that the program , to which he made me aware of might not be really looking back at 10,000,000.

                  The reason I say this is because the person who wrote the program wasn't sure if it infact looked 10,000,000 back.

                  To reiterate.... more testing needed..

                   

                  The last two night's drawing results (or more accurately, ANY two night's drawing results) will appear together MULTIPLE times in the 10 million number database (on average, at least 10 times).  Is your plan to look for just the first match, or check EVERY match?

                  I'll have the results this evening... and I'll do some foreshadowing by mentioning that you aren't going to like it...  <g>

                   


                    United States
                    Member #17555
                    June 22, 2005
                    5582 Posts
                    Offline
                    Posted: August 3, 2005, 4:18 pm - IP Logged

                    Thanx for the info...Pinback....I'm looking to see what those 10 or so matches are, and see if any numbers match the actual lottery results. In other words.....

                    587--two nights ago

                    377--last night ....both in florida.

                    I would like the program to look back at those two numbers in that order and tell me what numbers came after 587 and 377. Let's say that it spit out 10 numbers....107..402...516...567.....308...529....329...925....729....778......

                    Will any of those 10 above match tonight's drawing? Of course I wouldn't wait until tonight but would run the system starting with last month and checking to see if there were any matches with whatever the first two days of the month were and start there all the way down, separately,every day to see if it's even working before I spend any money. Obviously those two numbers would be changing every night because you will have two new numbers to input in the program every night.

                    Try inputting,just for fun the actual numbers of the florida's results that are listed above. 587..377.If you get ten or so numbers, we'll see if one of those numbers matches tonight's winning number.

                      Pinback's avatar - driver
                      Ohio
                      United States
                      Member #5810
                      July 22, 2004
                      120 Posts
                      Offline
                      Posted: August 3, 2005, 4:42 pm - IP Logged

                      Unfortunately, I think you'll quickly find this whole idea falls into the same catagory as the Union Jack method... what I like to call "The Stuff That Somebody Made Up But Has No Basis In Mathematics Or Statistics" methods.

                      I ran the idea against the Ohio Mid-day Pick 3 drawing.  Running it against all 1854 drawings would have taken many hours, but it cleary wasn't necessary.

                      Just to clarify, I ran each "pair" of winning numbers against the ENTIRE 10 million number database.  As I mentioned, each pair will hit multiple times, so I tested for a win against all the matching pairs in the 10 million number database.  (For example, the pair of wins 587 and 377 would appear together multiple times in the 10 million number database... so I checked for winners each time.)

                      After just 76 pairs of winning lottery numbers were checked, there were a total of 752 matching pairs in the 10 million number database.  I then compared the winning drawing following the winning pair against the number following the matching pair in the 10 million number database.  I just checked for a "box" win (not straight), since I wanted to know the maximum number of wins possible.

                      And the results were:  2 wins and 750 losses.

                      Ugh.  Only twice out of 752 tries did the number after the pair match even a box win against the actual winning numbers that were drawn.  I don't need to run the test any further in the hopes that things will "even out" at some point...  obviously there isn't a correlation there.

                      Sorry... I know that isn't what you were looking for!


                        United States
                        Member #17555
                        June 22, 2005
                        5582 Posts
                        Offline
                        Posted: August 3, 2005, 5:35 pm - IP Logged

                        I personally think that I'm not making my self clear. That's obviously not your fault. I appreciate your time though. Thanx....

                          Pinback's avatar - driver
                          Ohio
                          United States
                          Member #5810
                          July 22, 2004
                          120 Posts
                          Offline
                          Posted: August 3, 2005, 5:43 pm - IP Logged

                          I personally think that I'm not making my self clear. That's obviously not your fault. I appreciate your time though. Thanx....

                          You've been perfectly clear.  But don't worry... I wouldn't want to trust someone else either!  By all means, do it yourself manually for awhile... but please let us know the results!  I'd also be very interested in the results of the other programmer.  Of course, being a programmer myself... if he doesn't get the same results I did I'll insist on seeing his code...  <g>

                           

                            paurths's avatar - underground
                            Switching between Fairfax, VA and Belgium
                            Belgium
                            Member #19287
                            July 29, 2005
                            2254 Posts
                            Offline
                            Posted: August 3, 2005, 6:12 pm - IP Logged

                            1000 possible, first, combinations.

                            For each combination there are another 1000 combinations that might follow, makes 1 000 000.

                            For combination number 3, that makes a total of 1 000 000 000 combinations-sets in your database.

                            For each set of 4, makes 1 000 000 000 000.

                            and so on and on... and you'll never know which one to pick... (unless you only go for box then there will be less sets needed, but the amount of possible sets of combinations will stille be amazing...)


                              United States
                              Member #17555
                              June 22, 2005
                              5582 Posts
                              Offline
                              Posted: August 3, 2005, 6:19 pm - IP Logged

                              I wasn't trying to be sarcastic in any way. I really do  appreciate the time you've spent on this and a couple of others.

                              My friend is slow to conclusion based on what I said earlier. His wife..the baby...etc... He printed a copy of a list of 12 sets of numbers from this past couple of week's drawings, that the program spit out. Unfortunately he re-hit the program 20 times before finding a match, which was 718 for the next day. I asked him why he did that and should have printed the first punch....the true punch and see the results whether they matched or not.

                              Of course you will find 718 if you keep re-generating numbers by hitting the proccess button over and over again. I wanted the first list and only the first one.

                              Another problem I confronted him about the list is how short it was. I don't think the program is looking back 10 million drawings, and he says it's possible that I'm right.

                              If you analyze the 5 thousand or so Florida lottery pick-3 base of winning numbers going back about 14 years, you'll find, as I did a couple of two set numbers in any given situation give or take one. That's just 5 thousand. Imagine 10 million? I'm not a statistical mathematician or any thing close, but I will say that common sense would dictate that more number results than just 10 or 12 would show up looking back at 10 million.

                              He's looking into that. Another problem seems to be the seeding part that he puts so much emphasis on. I don't really care when you start the sequence, just as long the sequence stays the same.

                              I don't want to re-generate a new sequence with a new seed every time. I guess that woudn't matter, but for some reason I feel the same list of 10 million would suffice. 

                              There are so many filters I would like for him to apply but he's just trying to get the basics down and to see if the program is actualy looking back 10 million, or if he has to make adjustments. He said he will run a thousand first, therefore you can actually see the number balance displacement. If the outcome is about the same as far as spitting out a somewhat equal amount of times then you'll know it is looking back at a thousand and that the program is really looking back at 10 million, respectively.

                              We'll see.