Welcome Guest
Log In | Register )
You last visited December 2, 2016, 2:44 pm
All times shown are
Eastern Time (GMT-5:00)

Ohio Classic Lotto experiment

Topic closed. 7 replies. Last post 9 years ago by RJOh.

Page 1 of 1
PrintE-mailLink
RJOh's avatar - chipmunk
mid-Ohio
United States
Member #9
March 24, 2001
19813 Posts
Offline
Posted: September 1, 2007, 2:12 pm - IP Logged

This is an experiment with Ohio Classic Lotto.

Recently Ohio started a new daily game Ten-OH! which is a keno style game in which the player picks 10 of 80 numbers and the state draws 20 of those 80 numbers and if the player matches 5-10 or none of the 20 numbers drawn he wins a prize.  I checked and the odds of matching none of the 20 numbers are about the same as matching 5 of them and that got me thinking "Is there a limit to how long one can pick losing combinations randomly in a regular lottery?".  Thus the reason for this experiment.

Classic Lotto (6/49) has had 95 drawings until today. The odds of a winning match is as follows:
 MATCH    ODDS
  6/6    1 : 13983816
  5/6    1 : 54201
  4/6    1 : 1032
  3/6    1 : 57

There are 18424 combinations of threes in a pool of 49 numbers and 20 of them are in every combinations of 6.  Statistically one of those combinations of threes in every combination of six will appear in the winning numbers an average of once every 57 drawings.  With an RNG I generated 35 ordinary combinations of sixes with only the conditions that they had never matched three in the previous 95 drawings and no combination of three was repeated within them. These 35 combinations have gone 95 drawings with out a hit, that's the equivalent of a single combination going 3325 drawings without a hit, can they continue to do it?  If they do then there must be something unique about them to be avoided with future picks.

These are the 35 combinations:

    12 17 25 41 42 44    09 11 22 27 33 37    08 10 30 35 41 43
    13 24 35 36 47 48    21 25 27 35 38 39    12 18 19 27 29 31
    23 31 32 37 39 46    10 11 16 20 25 36    05 12 15 43 46 47
    11 19 28 38 43 45    05 14 19 26 40 41    03 20 27 42 43 48
    08 15 18 33 34 40    06 18 38 44 46 48    02 03 21 32 33 41
    06 09 10 21 26 29    16 21 24 40 45 46    04 12 23 24 33 38
    01 07 14 16 27 30    05 30 33 36 44 45    01 03 05 28 29 35
    20 23 34 41 45 47    16 22 26 32 44 47    04 07 26 34 36 46
    29 32 36 38 40 42    02 24 30 34 37 42    10 13 14 39 42 45
    04 14 15 20 31 35    07 09 19 20 24 39    06 13 15 25 30 49
    02 17 18 26 39 43    01 11 13 17 32 34    07 08 11 29 47 49
    07 12 13 21 22 28    03 06 08 14 17 23

I picked 10 of these combinations randomly and posted them on the prediction board.  I will review the results later after tonight's drawing

 * 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
    19813 Posts
    Offline
    Posted: September 1, 2007, 8:58 pm - IP Logged

    09/01/07 - 11 23 30 31 32 45

        12 17 25 41 42 44    09 11 22 27 33 37    08 10 30 35 41 43
        13 24 35 36 47 48    21 25 27 35 38 39    12 18 19 27 29 31
        23 31 32 37 39 46    10 11 16 20 25 36    05 12 15 43 46 47
        11 19 28 38 43 45    05 14 19 26 40 41    03 20 27 42 43 48
        08 15 18 33 34 40    06 18 38 44 46 48    02 03 21 32 33 41
        06 09 10 21 26 29    16 21 24 40 45 46    04 12 23 24 33 38
        01 07 14 16 27 30    05 30 33 36 44 45    01 03 05 28 29 35
        20 23 34 41 45 47    16 22 26 32 44 47    04 07 26 34 36 46
        29 32 36 38 40 42    02 24 30 34 37 42    10 13 14 39 42 45
        04 14 15 20 31 35    07 09 19 20 24 39    06 13 15 25 30 49
        02 17 18 26 39 43    01 11 13 17 32 34    07 08 11 29 47 49
        07 12 13 21 22 28    03 06 08 14 17 23

    The winning combinations for Classic Lotto drawing #96 were 11 23 30 31 32 45.   As shown above only one of the 35 combinations matched 3 which means 34 combinations of 6 have gone 96 drawings without matching three.  The ten combinations I picked randomly to post on the prediction board didn't include the one that match three.  Had I just picked the first 10 combinations, they would have.

    Conclusions:
    While the overall odds of 1:54 of getting a hit may be improved by selectively picking combinations with certain parameters, the pay off may not be worth the effort.  Since only one combination matched three, I will continue this experiment using the same group of combinations for another drawing which will be #97 and take place on 09/03/07.  I will re-evaluate the results after the drawing.

     * 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
      19813 Posts
      Offline
      Posted: September 3, 2007, 12:29 pm - IP Logged

        REC.NO.    DATE        NUMBERS          MATCH
        1  96    09/01/07  11 23 30 31 32 45    MIN. 1
      ____________________________________________________
        92  1 08/22/07  05 09 19 22 31 38      1*
        1 ----------------------------------------
        89  2 08/15/07  14 18 23 25 29 44      1*
        2 ----------------------------------------
       10  86  5 08/08/07  11 23 30 34 40 49      3*
        4 ----------------------------------------
       11  85  6 08/06/07  15 21 25 32 40 47      1*
        5 ----------------------------------------
       13  83  7 08/01/07  05 08 19 22 27 30      1
       14  82  8 07/30/07  02 06 07 25 41 45       1*
        6 ----------------------------------------
        1/ 4    2/ 7    4/10    5/10    6/11    8/14  (hit pattern)

      Another way I look at winning numbers is to display their hit patterns.  The hit pattern of the last drawing shows all the numbers hit 8 times in the previous 14 drawings.  When I checked the patterns of the last 50 drawing 25% of the previous drawing had done that and in 80% of them at least 5 numbers hit in the previous 14 drawings so I used my RNG with parameters that covered the most popular hit patterns of combinations that had 6 numbers from the previous 14 drawings to come up with 36 combinations.  I eliminated 9 that didn't match 3 with any of the combinations in the first test and these are the remaining 27:

          09 27 38 40 44 48    15 26 34 42 45 46    02 19 28 40 43 48
          22 28 35 43 45 46    16 22 24 28 44 47    05 08 15 16 18 26
          07 11 18 25 42 47    10 11 23 37 43 46    03 04 08 15 34 49
          04 06 14 23 36 41    02 26 34 40 42 47    07 14 22 35 40 41
          07 24 30 32 40 42    21 29 32 38 41 46    05 16 19 30 36 42
          23 30 34 41 44 46    18 27 28 44 45 48    07 10 14 19 26 49
          16 28 36 38 43 47    09 18 25 27 31 36    06 08 23 31 37 38
          04 08 24 25 31 35    04 07 38 42 46 48    09 11 16 19 29 49
          04 19 26 36 38 40    18 24 36 43 45 48    05 07 14 16 32 49

      I posted the first 10 on the prediction board for today's drawing.  I will continue to evaluate after the drawing.

       * 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
        19813 Posts
        Offline
        Posted: September 3, 2007, 8:23 pm - IP Logged

        09/01/07 - 11 23 30 31 32 45
        09/03/07  08 18 29 41 46 48

            12 17 25 41 42 44    09 11 22 27 33 37    08 10 30 35 41 43
            13 24 35 36 47 48    21 25 27 35 38 39    12 18 19 27 29 31
            23 31 32 37 39 46    10 11 16 20 25 36    05 12 15 43 46 47
            11 19 28 38 43 45    05 14 19 26 40 41    03 20 27 42 43 48
            08 15 18 33 34 40    06 18 38 44 46 48    02 03 21 32 33 41
            06 09 10 21 26 29    16 21 24 40 45 46    04 12 23 24 33 38
            01 07 14 16 27 30    05 30 33 36 44 45    01 03 05 28 29 35
            20 23 34 41 45 47    16 22 26 32 44 47    04 07 26 34 36 46
            29 32 36 38 40 42    02 24 30 34 37 42    10 13 14 39 42 45
            04 14 15 20 31 35    07 09 19 20 24 39    06 13 15 25 30 49
            02 17 18 26 39 43    01 11 13 17 32 34    07 08 11 29 47 49
            07 12 13 21 22 28    03 06 08 14 17 23

        Again the odds of 1:54 of getting a hit was beaten by picking combinations that hadn't match3 in previous drawings but just getting a match3 both times with 35 combinations is not profitable so it's back to the doing more experiments off line.

        The experiment with 27 combinations with the most popular hit patterns of previous winning numbers didn't go much better.  Only one match three.

            09 27 38 40 44 48    15 26 34 42 45 46    02 19 28 40 43 48
            22 28 35 43 45 46     16 22 24 28 44 47    05 08 15 16 18 26
            07 11 18 25 42 47    10 11 23 37 43 46    03 04 08 15 34 49
            04 06 14 23 36 41    02 26 34 40 42 47    07 14 22 35 40 41
            07 24 30 32 40 42    21 29 32 38 41 46    05 16 19 30 36 42
            23 30 34 41 44 46    18 27 28 44 45 48    07 10 14 19 26 49
            16 28 36 38 43 47    09 18 25 27 31 36    06 08 23 31 37 38
            04 08 24 25 31 35    04 07 38 42 46 48    09 11 16 19 29 49
            04 19 26 36 38 40    18 24 36 43 45 48    05 07 14 16 32 49

        This experiment is concluded.  Future experiments will be done off line. 

         * 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
          19813 Posts
          Offline
          Posted: September 3, 2007, 11:25 pm - IP Logged

          Displaying the hit patterns of the MegaMillions numbers may be more useful since a macth3 pays better in this game.

           REC.NO.    DATE        NUMBERS          MATCH
           1 553    08/31/07  08 18 22 40 44    11 MIN. 1
          ___________________________________________________
           1 552  1 08/28/07  37 40 48 53 56    44    1*
           1 ----------------------------------------
           3 550  2 08/21/07  08 38 54 55 56    02    1*
           2 ----------------------------------------
           4 549  3 08/17/07  02 08 33 38 53    12    1
           5 548  4 08/14/07  05 08 09 24 34    17    1
           9 544  5 07/31/07  05 18 37 39 43    42    1*
           3 ----------------------------------------
          11 542  6 07/24/07  08 17 20 23 50    17    1
          17 536  7 07/03/07  21 35 37 39 40    35    1
          18 535  8 06/29/07  01 17 40 51 52    07    1
          27 526  9 05/29/07  02 24 44 51 54    07    1*
           4 ----------------------------------------
          28 525  10 05/25/07  10 14 22 42 43    01    1*
           5 ----------------------------------------
              1/ 1    2/ 3    5/ 9    9/27  10/28

          This could be a project to work on until the MegaMillions jackpot reaches $200M+ again.

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


            United States
            Member #16612
            June 2, 2005
            3493 Posts
            Offline
            Posted: September 4, 2007, 2:33 pm - IP Logged

            Wow that's a great experiment you're testing RJOh.

              RJOh's avatar - chipmunk
              mid-Ohio
              United States
              Member #9
              March 24, 2001
              19813 Posts
              Offline
              Posted: September 4, 2007, 8:20 pm - IP Logged

              Wow that's a great experiment you're testing RJOh.

              Thanks Just6ntlc,

              I'm trying to pick winning numbers by defining their hit patterns which may also expose their repeatability patterns.  It started out with Ohio Classic Lotto but now I'm looking at MegaMillions drawings instead because they may be easier and they have more history.  In the last 229 drawings since its last matrix change, I defined a pattern that covered 34% of the winning combinations of five in past drawings.  Now the task is to find the number of possible combinations of five within that pattern that didn't win and maybe redefine the pattern to cover less combinations without effecting the percent of coverage of the real drawings.

              At first I thought it might be interesting to show how I went about coming up with a theory for picking winning numbers but later decided to continue the experiment off line.  If the experiment is a success it could make winning a jackpot game as easy as winning a pick4 game and as a result there will be a big improvement in my prediction statistics.

               * 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
                19813 Posts
                Offline
                Posted: September 4, 2007, 10:50 pm - IP Logged

                Based on my observations so far, if MegaMillions had rolled last Friday, I would have played these 25 lines tonight. I shall be creating play files and checking them as an experiment until the jackpot is larger and hope to be ready when it is.


                    05 11 26 31 36 +39    16 30 34 35 41 +07    04 27 28 32 52 +35
                    04 23 24 37 54 +40    01 06 14 22 48 +19    02 13 31 33 39 +12
                    07 13 17 44 45 +04    05 06 21 25 39 +32    09 22 31 35 51 +36
                    01 20 25 28 43 +41    02 07 10 19 43 +27    11 13 14 20 49 +11
                    18 38 39 47 48 +23    17 20 29 33 34 +26    05 30 43 44 52 +33
                    02 21 51 52 56 +25    09 12 40 44 54 +46    18 23 26 32 34 +44
                    14 32 42 46 55 +18    12 18 41 49 56 +21    12 21 36 37 42 +17
                    09 10 27 33 50 +42    16 19 23 42 50 +09
                    03 22 29 40 49 +28    03 24 35 36 45 +13

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