Welcome Guest
Log In | Register )
You last visited December 5, 2016, 3:33 pm
All times shown are
Eastern Time (GMT-5:00)

Developing a winning MegaMillions System

Topic closed. 1111 replies. Last post 6 years ago by Stack47.

Page 2 of 75
4.215
PrintE-mailLink
RJOh's avatar - chipmunk
mid-Ohio
United States
Member #9
March 24, 2001
19825 Posts
Offline
Posted: September 2, 2008, 11:29 pm - IP Logged

Hey RJOh..

Not for nothing but have any of you JP players tried the angle of the ball sums(Ball 36=9 type of stuff)?Particularly 6,7,9 and 12 sums and if you have what was the impact of wheeling 3 fairly repetive sums and the outcome of those specific sums over a period of time?.

I've never tried it, but if anyone has I would be interested in their results.

 * you don't need to buy more tickets, just buy a winning ticket * 
   
             Evil Looking       

    RJOh's avatar - chipmunk
    mid-Ohio
    United States
    Member #9
    March 24, 2001
    19825 Posts
    Offline
    Posted: September 3, 2008, 9:53 am - IP Logged

    Review of system

    The distribution pattern 4.ABCCC was predicted
    Total Matches with previous drawings close but need adjustments
    *4 of 5 winning numbers in core pool of 45 numbers
    *Bonus number not in bonus pool of 20 numbers
    Beat over all odds of 1:39 of getting a hit
    Beat odds of 1:306 of getting a 3/5+0
    The additional parameters(sum range and gaps) were on target

    System need some tweaks before next MM drawing but will keep the 20 combo format.

     * you don't need to buy more tickets, just buy a winning ticket * 
       
                 Evil Looking       

      Stew12's avatar - bad egg-64x64.png
      CT
      United States
      Member #61398
      May 21, 2008
      781 Posts
      Offline
      Posted: September 3, 2008, 11:35 am - IP Logged

      Nice work RJOh.  Definitely interested to hear more about your methods.

      I have been working on a system for PB, and play with a little group of people. We wheel 22 WB numbers and 14 PB numbers arranged in 42 tickets (3if5 of 22).  I have been working with my systems a good amount and found that the 'Due' methods have been the most accurate.  I have analyzed followers, hot/cold, frequencies, deviations, etc.  I also started noticing that more frequently numbers would fall in certain areas of my DUE ranking than in others.  I have been picking the top 11 due and the bottom 11 due, and wheeling those with 14 PB numbers that are ranked in the middle of their due values. For the white balls I use a pool size of 288 draws, and I look back 247 draws for the PB number.  Out of the last 10 draws, this is what has hit from my numbers:

       

      5 numbers in pool:1
      4 numbers in pool:2
      3 numbers in pool:2
      2 numbers in pool:4
      1 numbers in pool:1
      0 numbers in pool:0

       

      And according to my ranknig system, these are where the numbers have fallen:

       

      DUE #1DUE #2DUE #3DUE #4DUE #5DUE PB
      793831362
      3402941329
      504244452634
      3522981730
      51832312326
      484552111011
      244518247
      4755145441
      112930275520
      389448326

       

       

      Note: This is for Powerball, not Mega Millions.

        RJOh's avatar - chipmunk
        mid-Ohio
        United States
        Member #9
        March 24, 2001
        19825 Posts
        Offline
        Posted: September 3, 2008, 3:23 pm - IP Logged

        Stew12,

        Your PB is a little different than what I do with MM.  I use to travel to Indiana to play PB when its jackpot was $100M+ but stop when gas prices went up.  I played it using similar parameters and limits that I used with MM.  I opted for the distribution patterns rather than wheeling because I couldn't wheel enough numbers in 20 lines or less.  Good luck to you and your group tonight.

        I have an idea for narrowing some of the parameters to get better results this Friday.  I checked back over 200 drawings and 67% of those combinations would have made it through.

         * you don't need to buy more tickets, just buy a winning ticket * 
           
                     Evil Looking       

          Avatar

          United States
          Member #37109
          April 10, 2006
          57 Posts
          Offline
          Posted: September 3, 2008, 3:31 pm - IP Logged

          For tonight's MM drawing, I will pick combinations of fives from the following 45 numbers:01 02 03 04 07 08 09 12 13 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 35 36 38 39 40 41 42 43 44 45 46 48 50 51 52 54 55 56

          and pick the bonus numbers from the following 20 numbers:01 04 05 06 08 09 10 11 12 15 17 20 21 26 29 35 36 41 42 44

          I will use the following 7 distribution patterns:
          A = 1-19   B = 20-38   C = 39-56

             1. AABBC           
             2. ABBCC           
             3. AABCC         
             4. ABCCC         
             5. AAABC           
             6. ABBBC         
             7. AACCC 

            TOTAL MATCHES WITH PREVIOUS 333 DRAWINGS

              MATCH  0 = 168-202
              MATCH  1 = 81-110
              MATCH  2 = 11-24
              MATCH  3 = 0-2     
              MATCH  4 = 0       
              MATCH  5 = 0                 

          I will also be using  some additional parameters that I have mentioned in other posts.

          RJOh

          Could you share how you come up with the numbers and what are you using to pick them?

          Doc Tor

            RJOh's avatar - chipmunk
            mid-Ohio
            United States
            Member #9
            March 24, 2001
            19825 Posts
            Offline
            Posted: September 3, 2008, 4:18 pm - IP Logged

            RJOh

            Could you share how you come up with the numbers and what are you using to pick them?

            Doc Tor

            I go back several fixed periods of time and check for the most common amounts of hits for the winning numbers during those period. For example since I have to go back 39 drawings to cover all 56 numbers in the most recent drawings I check all the 39 drawings periods of the last 334 drawings since the last matrix change.  It turns out there has been 295 such periods and the most common hits were 2-5.  If I decided to use the last 39 drawings, my core numbers pool would be the 44 numbers that have hit 2-5 in the last 39 drawings.  I check several periods of times and the ones with the best coverage is the one I use.   Sometimes it just a hunch. 

            For my bonus numbers pool I use the top bonus numbers that haven't hit recently. 

            Once I have my number pools I choose the parameters that will be used to select or reject combinations.  Those parameters act like filters and would accept at least 60-70% of previous winners.  They include number of hits, matches, sums, ranges, gaps and etc.  For the last drawing I only had four of the winning numbers in my core number pool and none of the bonus numbers in the bonus number pools but I still got a 3+0 in 20 lines.  Had all six numbers been in them, I might have done better.

            If this seems like much to do about nothing, remember this is a hobby and I do it for entertainment.

             * you don't need to buy more tickets, just buy a winning ticket * 
               
                         Evil Looking       

              Avatar
              Long Island, NY
              United States
              Member #57412
              January 3, 2008
              419 Posts
              Offline
              Posted: September 3, 2008, 4:44 pm - IP Logged

              I just dabble with Mega in my spare time.  I play other games.

              I too would like to help and contribute.  But I hesitate as I can't validate my opinion/approach with a winning claim.  However, I would like to think I have shown the LP community thus far that I have developed some degree of number picking ability.

              I've read many of your past posts and surely yours is a practical fundamental approach that provides predictions based on past draws history.

              Is it better to spend your time improving your current strategy/method..or....spend your primary time developing a new method?

              Any way I slice it.  Filtering methods commonly used/discussed here "box" you in.  I lose every time.  In that, I never capture "enough" numbers for a win.  You get a few numbers and it teases you to continue on down a road that, without luck won't get you to your destination.  So I spend alot of time:

              1)developing new filtering methods

              2)using a blend of methods commonly used.  Much as a master Chef combines commonly found ingredients/seasonings  to produce a new unique menu item.

              We all need an approach that accurately predicts what history tells us "shouldnt" happen with the other half of the numbers that come out in the next draw.  I always include intuition as one of my key ingredients.

              The Mega Millions draw history is a perfect example of a bumpy road as far as history goes.  For example, lukewarm numbers (seems more so than many other games I analyze) pop up in every draw and no "one" system seems to be able to accurately predict which ones will pop...to produce a significant win.  So I consider all mindsets when I select numbers.  Mega is a game where cold numbers players won't do as well.

              Few Ideas I don't hear about as frequently:

              1) The relations of the draws of winning(white) numbers and the mega ball draws.  If 39 came out as the Mega Ball.... when will/or has it appeared within the white balls drawn?

              2) simply add a planned amount to each drawn number.  If 6 comes out, add two to it and play eight.  Play several more lines with different amounts added.   Through analyzation you can develop amounts to add to each category (teens, 20's, 30's etc...) I love toying with this old method and it is stupid easy to use and provides as good as results as any.

              I've got many more I use every day for my picks.

                time*treat's avatar - radar

                United States
                Member #13130
                March 30, 2005
                2171 Posts
                Offline
                Posted: September 3, 2008, 5:06 pm - IP Logged

                Hey RJOh..

                Not for nothing but have any of you JP players tried the angle of the ball sums(Ball 36=9 type of stuff)?Particularly 6,7,9 and 12 sums and if you have what was the impact of wheeling 3 fairly repetive sums and the outcome of those specific sums over a period of time?.

                When you get to this level of compression, you lose too much infomation to reconstruct the original number in isolation from a bunch of other possibilities.

                I.e., 9 'blows up' into 36, but also 27, 18, and 09

                Do that for five numbers and if you get 4 for each of them, you're looking at 1,024 combos,

                In neo-conned Amerika, bank robs you.
                Alcohol, Tobacco, and Firearms should be the name of a convenience store, not a govnoment agency.

                  RJOh's avatar - chipmunk
                  mid-Ohio
                  United States
                  Member #9
                  March 24, 2001
                  19825 Posts
                  Offline
                  Posted: September 4, 2008, 11:48 am - IP Logged

                  "We all need an approach that accurately predicts what history tells us "shouldnt" happen with the other half of the numbers that come out in the next draw."

                  DownandDirty,

                  Look as last night PowerBall drawing:38 40 46 49 50 +44, there hadn't been a drawing where the lowest WB number was that high since 10/13/01 but that doesn't mean it "shouldn't" happen.  When I look at draw history, I'm looking for events or stats that commonly  happen and therefore are likely to happen again.  I've never tried to come up with any absolute rules for picking combinations, all the possible combinations are on the table.   I'm just trying to pick the best 10-20 of them to play at the time.

                  Like you, I'm just dabbling with MegaMillions in my spare time and having a little fun.   

                   * you don't need to buy more tickets, just buy a winning ticket * 
                     
                               Evil Looking       

                    Stew12's avatar - bad egg-64x64.png
                    CT
                    United States
                    Member #61398
                    May 21, 2008
                    781 Posts
                    Offline
                    Posted: September 4, 2008, 2:52 pm - IP Logged

                    Interesting thread indeed.

                    We managed to match 3 of the 5 numbers from PB last night in our 22 number selection.  38, which was the furthest out number (ranked 55 due as well) along with 46 and 50, which were ranked #9 and #7 due, respectively. However, we did not hit the powerball number.

                    I have done a bit of research with this same method for Mega Millions, and it appears picking top 11 and bottom 11 due (for a 22 number selection) seem to work better than picking either the top 22 or bottom 22.  A quick backtest shows the hits as follows:

                     

                    Date:9/2/2008  Hits:2
                    Date:8/29/2008  Hits:1
                    Date:8/26/2008  Hits:4
                    Date:8/22/2008  Hits:2
                    Date:8/19/2008  Hits:1
                    Date:8/15/2008  Hits:3
                    Date:8/12/2008    Hits:2
                    Date:8/8/2008   Hits:2
                    Date:8/5/2008  Hits:2
                    Date:8/1/2008  Hits:3
                    Date:7/29/2008  Hits:4
                      Avatar

                      Honduras
                      Member #20982
                      August 29, 2005
                      4715 Posts
                      Offline
                      Posted: September 4, 2008, 3:00 pm - IP Logged

                      Rjoh guess what? i saw your system and i thought about it too...I thought about it like 2 weeks ago, but i thought using 16 numbers and just waiting...Very good job...

                      I also notice that your system uses part of Tntea's Schooling for Powerball...

                      The Forex trades: 1.6 Trillion dollars EVERY day, that´s more than the GDP of the Carribbean Central America, COMBINED. Enough to feed every crook out there for centuries...To all Geniuses & Powers Countries of the World the Planet needs breakthroughs in all Medicine, Veterinary, Biology related fields, Psychology, Population Psychology/Sociology..They need to genetically ingeneer new plants species/types to give more variety of plants and thus have more resources for combating diseases¨


                       


                       


                       

                       


                        RJOh's avatar - chipmunk
                        mid-Ohio
                        United States
                        Member #9
                        March 24, 2001
                        19825 Posts
                        Offline
                        Posted: September 4, 2008, 3:13 pm - IP Logged

                        Interesting thread indeed.

                        We managed to match 3 of the 5 numbers from PB last night in our 22 number selection.  38, which was the furthest out number (ranked 55 due as well) along with 46 and 50, which were ranked #9 and #7 due, respectively. However, we did not hit the powerball number.

                        I have done a bit of research with this same method for Mega Millions, and it appears picking top 11 and bottom 11 due (for a 22 number selection) seem to work better than picking either the top 22 or bottom 22.  A quick backtest shows the hits as follows:

                         

                        Date:9/2/2008  Hits:2
                        Date:8/29/2008  Hits:1
                        Date:8/26/2008  Hits:4
                        Date:8/22/2008  Hits:2
                        Date:8/19/2008  Hits:1
                        Date:8/15/2008  Hits:3
                        Date:8/12/2008    Hits:2
                        Date:8/8/2008   Hits:2
                        Date:8/5/2008  Hits:2
                        Date:8/1/2008  Hits:3
                        Date:7/29/2008  Hits:4

                        How do you determine the top and bottom eleven numbers?  Do you use all the previous draw history since the last matrix change or do use a fixed amount of draw history? 

                        I've been experimenting with using several lengths of draw history and when I want use all 334 MegaMillions drawings since the matrix change I add the 334 drawings to themselves in a test file and use them like a loop, looking back 333 drawings to avoid looking a the same drawing that I'm comparing.

                         * you don't need to buy more tickets, just buy a winning ticket * 
                           
                                     Evil Looking       

                          Stew12's avatar - bad egg-64x64.png
                          CT
                          United States
                          Member #61398
                          May 21, 2008
                          781 Posts
                          Offline
                          Posted: September 4, 2008, 3:34 pm - IP Logged

                          I use a fixed draw history amount, but I built into my application functionality that uses the specific algorithm I want to test and automatically back tests it for every possible amount of draw history (loops through with a 10 draw history, 11, 12,13....).  It tells me data on each pool size (draw history size) as far as total hits, calculations, etc.  When I run this draw history checker on my 'Due' algorithm (with 22 pics) I can see that the pool size 266 ran 11 back tests, hit 4 balls twice, hit 3 balls twice, hit 2 balls 5 times, hit 1 ball twice, and did not have any hits of 0 or 5.  Then I run that poolsize/algorithm through the back-check/predictor method (shows both at the same time), and have the numbers ready for the next draw, with the old numbers/hits on the same screen.

                          When I spit the picks like this (front 11/back 11) I run the application in this fashion twice and post a number of different result sets into excel and analyze them myself, determining which method/pool size gives the highest consistent hit percentage.  Then I run the application and tell it to give me the highest ranked 11 picks, and run it a second time telling it to give me the lowest ranked 11 picks, with the desired pool size.

                            RJOh's avatar - chipmunk
                            mid-Ohio
                            United States
                            Member #9
                            March 24, 2001
                            19825 Posts
                            Offline
                            Posted: September 4, 2008, 4:02 pm - IP Logged

                            That sounds more complex than anything I've ever design.  Unless I want to use the default values which cover everything, I have to enter specific values every time I run my programs and record the results.  I have no algorithm to evaluate which result is the best, it's simple an eyes on opinion.

                            Once I decide, I enter those parameters and data in the RNG that pick my combinations and it stays within the parameters and use the data I entered.  Any questionable picks, the program stops for my an eyes on opinion before continuing.

                             * you don't need to buy more tickets, just buy a winning ticket * 
                               
                                         Evil Looking       

                              Stew12's avatar - bad egg-64x64.png
                              CT
                              United States
                              Member #61398
                              May 21, 2008
                              781 Posts
                              Offline
                              Posted: September 4, 2008, 4:14 pm - IP Logged

                              The algorithm that decides which result sets are the best isn't really that complicated itself, instead of printing out values for 1 result set with a certain drawing history it runs a 'for' loop and records each result set to an array, then when finished, prints out the results from the array with their corresponding pool sizes for review.  At that point I review the results and decide which drawing history I think might work best.  That portion of the app is simply to pick which drawing history size I would like to use.

                                 
                                Page 2 of 75