Welcome Guest
You last visited August 22, 2017, 8:39 am
All times shown are
Eastern Time (GMT-5:00)

# Statistics around the balance of even/odd and small/big numbers

Topic closed. 142 replies. Last post 7 years ago by RL-RANDOMLOGIC.

 Page 3 of 10

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: January 15, 2011, 8:00 pm - IP Logged

In a sense ,it's likr those post that I read that in a pick 6 game 1'2'3'4'5'6 can be a winning combination...Sure It can..Though when and if it happens good luck and congratulation to the winner..I sure not going to wait around,,Just because (evens) hold a advantage be it big or small  at any particular time am I going to wait around choosng 4 odd numbers because all things tend to even out or their are more odd numbers in a particular game..Just look at Mega for the  last few drawings.. numbers between(1-9)

8-6-4-1/4-2 thats 5 to 1

and their are 5 odd/4 even numbers

So go ahead and choose all odds between (1-9)   for Mega over the next few drawings and let me know how it turns out.

MAYDAY,

I did a scan of the PA Cash-5 results over the life of the game.  There have been 5,460 Draws between 04/23/1992 and tonight, 01/15/2011.  Here are the dates and numbers drawn in two installments: 1st all the ODDS, then all the EVENs.  Note the difference between the total hit counts.

--Jimmy4164

All ODD Number Sets Drawn
Pennsylvania CASH-5

04/23/1992  Thru  01/15/2011

Date        Occur  (1) (2) (3) (4) (5)

7/30/1992       1   9  13  15  31  33
5/3/1993        2   1  11  17  27  37
5/6/1993        3   1   5   9  15  27
8/2/1993        4  13  29  33  35  39
10/11/1993      5   3   7  15  21  23
9/5/1994        6   9  21  23  29  39
12/5/1994       7  11  13  19  23  33
4/6/1995        8   1  15  29  31  39
12/4/1995       9   1  27  29  31  33
4/4/1996       10   1   7   9  17  19
7/15/1996      11   1   3  11  23  39
8/28/1996      12   1   5  21  37  39
10/16/1996     13   7  13  15  29  33
10/27/1996     14   1  25  29  37  39
12/5/1996      15   1   3  11  33  37
8/21/1997      16   1   7  15  25  33
11/10/1997     17  15  19  21  29  35
12/18/1997     18   5   9  13  31  33
1/29/1998      19   3  11  25  27  35
3/16/1998      20  11  15  21  23  39
3/19/1998      21   1  23  29  37  39
3/25/1998      22  11  13  15  27  37
4/16/1998      23   5   9  21  31  39
5/7/1998       24   7  11  25  27  39
6/13/1998      25   7  11  17  21  37
6/27/1998      26   7  13  17  21  33
8/1/1998       27   5  25  27  31  33
8/6/1998       28   9  19  23  35  37
8/29/1998      29   7   9  17  33  39
10/6/1998      30   5  21  23  25  39
10/15/1998     31   1  11  17  21  23
11/2/1998      32  11  21  27  37  39
12/20/1998     33   1   3   5  25  27
5/4/1999       34   9  13  19  23  35
5/16/1999      35   1   5  17  21  27
6/15/1999      36   1  15  17  31  35
9/1/1999       37   9  11  13  21  31
10/8/1999      38   3   7  23  31  33
11/19/1999     39  15  17  21  23  39
12/2/1999      40   3   5  17  31  35
3/13/2000      41   3  11  13  29  39
3/16/2000      42   7   9  17  29  31
4/13/2000      43   9  13  15  17  35
5/1/2000       44   1   3   7  11  27
5/3/2000       45   3   5  23  27  31
6/29/2000      46  13  19  23  25  29
7/11/2000      47   5   7   9  23  37
8/4/2000       48  11  19  21  33  39
10/9/2000      49  21  23  25  29  33
2/10/2001      50  19  21  33  37  39
2/14/2001      51   5  11  21  23  39
5/12/2001      52   1   3  23  27  37
6/11/2001      53   1   9  15  25  33
6/16/2001      54   3   5  21  27  33
9/21/2001      55  15  17  27  35  37
10/5/2001      56   5  15  27  29  39
11/11/2001     57   7  17  31  35  37
4/18/2002      58  13  15  17  21  29
5/9/2002       59   3   5  17  23  39
6/18/2002      60   5   9  17  21  23
6/29/2002      61   7  11  21  25  29
8/20/2002      62   5  15  19  31  37
9/2/2002       63  11  13  23  25  29
11/25/2002     64   3   5  13  17  21
12/4/2002      65   3   9  11  13  35
12/15/2002     66   5   7  11  21  37
12/26/2002     67   1  19  25  33  35
1/16/2003      68   7  17  33  37  39
1/23/2003      69  13  21  29  35  37
1/30/2003      70   7  11  27  29  33
2/19/2003      71   1   3   5  17  29
3/21/2003      72   5  21  25  29  35
3/29/2003      73   3  11  17  21  35
5/16/2003      74   3   5   7  21  25
7/1/2003       75  19  21  27  29  35
8/4/2003       76  19  25  31  35  37
8/11/2003      77   3   7  27  35  37
9/1/2003       78  15  17  25  37  39
12/16/2003     79   1   5   7  13  17
1/8/2004       80   7  11  13  17  27
3/5/2004       81  17  19  21  23  31
4/1/2004       82   3  21  29  31  37
4/2/2004       83   1  23  27  31  33
5/2/2004       84   3   7  17  23  31
5/29/2004      85   5   7  13  23  37
8/12/2004      86   1   5   9  11  29
8/17/2004      87   5   9  17  25  27
9/12/2004      88   1  17  25  31  35
9/16/2004      89   3  13  15  27  31
10/1/2004      90   1   5   9  19  37
10/20/2004     91   7  11  13  25  39
11/27/2004     92  15  17  21  33  35
1/23/2005      93   1   7   9  13  23
2/5/2005       94   3   5  21  23  25
2/26/2005      95   5  11  13  21  31
2/27/2005      96  17  21  23  27  39
3/26/2005      97   5  11  15  31  39
4/5/2005       98   3  15  23  31  35
4/24/2005      99  15  17  19  25  31
4/25/2005     100  11  15  27  33  35
5/17/2005     101   9  13  19  25  35
5/22/2005     102   1  19  33  37  39
8/25/2005     103   5  21  23  27  35
9/8/2005      104   3  11  15  21  25
10/30/2005    105   5  21  25  27  37
1/2/2006      106  19  23  25  31  33
1/17/2006     107   1   3  19  21  33
1/24/2006     108  17  25  29  37  39
4/19/2006     109   9  11  17  19  33
4/21/2006     110  17  21  25  29  39
4/24/2006     111   5   9  15  29  35
5/21/2006     112   7  17  21  31  37
6/4/2006      113   1   5   9  17  29
7/1/2006      114  13  19  21  31  37
7/18/2006     115   1   3   5  23  39
7/23/2006     116   3  11  15  31  39
7/26/2006     117   3  19  23  25  35
8/17/2006     118   1   9  29  31  35
9/30/2006     119  11  17  23  31  35
11/5/2006     120   7  23  25  33  37
12/22/2006    121   7   9  17  25  31
1/5/2007      122   1  19  21  25  35
2/15/2007     123  11  17  23  25  35
3/9/2007      124   1  19  25  27  33
3/15/2007     125   5   7  15  27  35
3/16/2007     126   3  13  31  33  37
4/16/2007     127  13  17  19  31  33
4/25/2007     128  11  19  23  29  33
5/13/2007     129   9  15  19  37  39
5/29/2007     130  19  23  33  35  37
6/11/2007     131   5  11  31  35  39
6/13/2007     132   5   7  15  33  37
8/20/2007     133   1   5  17  19  29
8/27/2007     134  21  23  29  37  39
9/5/2007      135   1   3  15  17  21
10/15/2007    136   1   9  29  31  33
11/24/2007    137   5   9  11  19  27
1/7/2008      138   1  19  23  35  39
1/10/2008     139   9  13  17  31  35
1/16/2008     140   5   7  15  25  27
2/11/2008     141   3  11  19  33  43
5/9/2008      142   7  13  19  31  37
8/25/2008     143   5  23  31  39  41
11/30/2008    144   1  11  13  23  27
12/11/2008    145  11  21  33  37  39
12/22/2008    146   5   9  19  35  39
2/15/2009     147   5  13  21  35  37
5/24/2009     148   1  21  29  35  37
10/29/2009    149   1   9  27  33  41
11/13/2009    150  17  27  33  35  41
12/25/2009    151   3  11  13  21  29
12/28/2009    152   7  15  27  33  41
1/3/2010      153  17  23  25  35  39
2/21/2010     154   7  11  23  29  39
4/7/2010      155  19  21  23  27  37
5/18/2010     156   1   5  13  27  39
5/30/2010     157   9  17  25  33  39
6/2/2010      158   1  19  21  31  35
6/4/2010      159   1   3  23  35  37
8/14/2010     160   3   7  13  31  39
8/17/2010     161   3  11  13  21  27
9/3/2010      162   7  25  27  31  35
9/4/2010      163   7  11  13  21  25
10/7/2010     164   5  27  29  33  39
10/14/2010    165   1  21  27  33  43
12/3/2010     166   1   7  21  33  41
12/24/2010    167   1  15  17  23  27
1/9/2011      168   5  17  25  31  37

All EVEN Number Sets Drawn
Pennsylvania CASH-5

04/23/1992  Thru  01/15/2011

Date        Occur  (1) (2) (3) (4) (5)

8/20/1992       1   4   8  14  20  32
6/14/1993       2   4  18  22  24  26
9/23/1993       3  14  20  24  28  32
11/15/1993      4   6  12  14  24  32
4/14/1994       5   4   8  10  18  22
10/27/1994      6   8  20  26  32  34
6/26/1995       7  12  22  30  32  34
11/13/1995      8  10  12  30  32  36
12/22/1995      9   2   6   8  18  22
4/17/1996      10  10  12  16  20  36
2/12/1997      11   2  16  20  32  36
3/5/1997       12  24  26  28  34  36
6/30/1997      13   6   8  24  32  36
8/4/1997       14   2   8  12  14  16
2/8/1998       15   4  14  22  36  38
2/28/1998      16   6  20  24  34  36
4/5/1998       17   4   8  12  14  24
4/30/1998      18   2   6  14  26  34
5/14/1998      19   2  20  22  34  36
10/29/1998     20  10  18  24  28  30
11/22/1998     21   4   8  10  24  34
12/28/1998     22   6  10  14  36  38
2/13/1999      23   4   8  10  14  26
2/23/1999      24   2  24  28  32  38
3/6/1999       25   2  12  26  28  32
5/11/1999      26   6  12  18  26  36
7/3/1999       27   8  12  24  26  36
10/15/1999     28   8  10  26  30  38
12/26/1999     29   6  16  24  30  34
12/29/1999     30   4  24  28  32  36
9/1/2000       31  14  18  26  34  38
9/28/2000      32  10  18  22  28  36
11/12/2000     33   6  10  16  34  38
12/3/2000      34   4   8  10  28  36
1/21/2001      35   8  14  22  26  30
1/25/2001      36   2   6  16  34  38
4/10/2001      37   6  12  16  28  32
7/16/2001      38   2  16  26  28  34
7/27/2001      39   6  10  18  20  30
8/29/2001      40   2   6  16  20  24
12/10/2001     41   2   4   8  22  32
12/21/2001     42   4  14  16  26  34
3/30/2002      43   6  12  16  26  28
4/22/2002      44   4  14  24  34  38
4/28/2002      45   2  20  24  28  34
9/15/2002      46  12  14  22  26  36
9/23/2002      47   2  14  16  26  36
10/1/2002      48   4  18  20  24  28
12/31/2002     49   6   8  14  18  20
1/19/2003      50   4  10  20  26  30
6/5/2003       51   2  14  20  22  30
9/17/2003      52   2   6  12  22  38
10/1/2003      53   4   8  24  30  32
11/14/2003     54   2  12  20  22  36
1/4/2004       55   6   8  22  24  38
2/8/2004       56   6  14  24  26  28
3/31/2004      57   4  10  24  28  36
4/23/2004      58   2  16  20  26  28
5/5/2004       59   4  20  32  36  38
5/15/2004      60   6  30  32  34  38
7/31/2004      61  10  22  26  36  38
11/5/2004      62   2   4   6  10  32
12/21/2004     63  10  12  26  28  36
1/25/2005      64   4  14  32  34  36
3/28/2005      65   2   4  10  24  32
5/7/2005       66   6  14  16  18  22
7/27/2005      67  10  12  22  30  32
8/29/2005      68   4   8  16  32  38
10/17/2005     69   2  14  16  22  32
4/12/2006      70   8  10  12  26  36
4/13/2006      71   2   4  18  22  32
7/6/2006       72   2   8  20  28  30
8/22/2006      73   2   6  18  20  26
9/11/2006      74   2  10  14  24  36
10/3/2006      75   2   4  10  26  28
3/28/2007      76   2   8  18  22  38
5/22/2007      77  10  20  26  28  30
8/10/2007      78   6  10  12  32  36
8/21/2007      79  12  26  28  30  32
8/25/2007      80   4  18  20  26  38
9/8/2007       81  12  14  16  32  34
10/6/2007      82   6  20  26  36  38
1/23/2008      83  12  14  16  32  38
3/2/2008       84   4  24  26  38  42
5/12/2008      85   2  24  28  30  36
5/23/2008      86   6  10  14  30  32
6/13/2008      87   8  26  28  30  38
9/14/2008      88  10  14  20  26  40
9/15/2008      89   2  12  18  34  40
9/27/2008      90   6   8  10  32  42
9/30/2008      91  10  12  26  34  40
11/4/2008      92   4  12  16  26  32
12/6/2008      93   2  14  18  20  36
12/10/2008     94   2   6   8  20  32
2/28/2009      95  22  28  30  34  40
3/15/2009      96  10  14  28  40  42
3/16/2009      97   8  12  14  16  34
5/2/2009       98   8  10  14  24  26
6/21/2009      99   2   4  16  18  42
2/11/2010     100   8  34  36  40  42
4/9/2010      101   2  10  16  30  34
5/13/2010     102   2   8  10  26  42
7/7/2010      103   8  10  18  36  42
8/4/2010      104  10  18  24  32  36
11/19/2010    105  12  16  22  40  42
12/13/2010    106  10  14  16  26  42
12/28/2010    107   4   6  18  22  30

------------EOF--------------

United States
Member #59354
March 13, 2008
4322 Posts
Offline
 Posted: January 15, 2011, 8:38 pm - IP Logged

5-43 matrix

Totals

26334  * 5 odd number sets

20349  * 5 even number sets

Overall

26334 + 20349 = 46683

26334 / 46683 = 56%   5 odd expected +/- SD

20349 / 46683 = 44%   5 even expected +/- SD

Actual draws

168 * 5 odd numbers

107 * 5 even numbers

Actual draws

107 + 168 = 275

107 / 275 = 39%

168 / 275 = 61%

all even 5% lower then expected  +/- SD

all odd 5% higher then expected  +/- SD

Nice Bias!

Which would prove that if stef played more odd the even then his chances of winning more

smaller prizes is confirmed while still not decreasing his chances for a jackpot.

RL

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: January 16, 2011, 12:35 am - IP Logged

5-43 matrix

Totals

26334  * 5 odd number sets

20349  * 5 even number sets

Overall

26334 + 20349 = 46683

26334 / 46683 = 56%   5 odd expected +/- SD

20349 / 46683 = 44%   5 even expected +/- SD

Actual draws

168 * 5 odd numbers

107 * 5 even numbers

Actual draws

107 + 168 = 275

107 / 275 = 39%

168 / 275 = 61%

all even 5% lower then expected  +/- SD

all odd 5% higher then expected  +/- SD

Nice Bias!

Which would prove that if stef played more odd the even then his chances of winning more

smaller prizes is confirmed while still not decreasing his chances for a jackpot.

RL

I knew I wouldn't have to make those calculations!

"Which would prove that if stef played more odd the even then his chances of winning more smaller prizes is confirmed while still not decreasing his chances for a jackpot."

Wrong!

The Expected Values are 149 ODD, and 115 EVEN.

Versus  the             168      and 107 Actual values.

Anyone who thinks the statistical significance of these deviations is sufficient to claim an "edge" over 5460 Draws is delusional!

http://www.staff.city.ac.uk/~sj361/p1369.pdf

This study documents how humans tend to make these kinds of judgement errors when making probability estimations.

United States
Member #59354
March 13, 2008
4322 Posts
Offline
 Posted: January 16, 2011, 1:10 am - IP Logged

Right!

"Anyone who thinks the statistical significance of these deviations is sufficient to claim an "edge" over 5460 Draws is delusional!"

Define "statistical significance"

You prove my point again that statistics has no significance here at

all, Not now not ever. The 5-43 matrix is biased by the one extra

odd number and playing more odd then even will win more often then

more even then odd. Just a fact!

RL

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: January 16, 2011, 2:39 am - IP Logged

The observed distributions of the difference in frequency counts of ODD numbers versus EVEN in a SET, the SUMS of the five numbers, OR the number of unique occurences of (0-9) digits in a set, etc, etc, etc, are of NO VALUE in trying to determine an "EDGE" in selecting numbers in a lottery.

Let's PRETEND for a minute.  Let's pretend that special paint has been developed that allows one to spray the [let's say] 39 ping pong balls with colors that allow their underlying numbers be read when necessary.  OK?

And let's let some very young children, who can't yet count beyond the number of fingers on one hand, play with the numbered balls and separate 1/3 of them (13) from the other 26.  In other words, let's RANDOMLY select 13 balls of the 39.  OK?

Now, let's spray paint the 13 balls RED, and the 26 balls BLUE.  OK?  (Now each ball has TWO identifying characteristics, NUMBER and COLOR.)  OK?

If we turn on the ball machine and RECORD the ball COLORS and NUMBERS resulting from repeatedly selecting 5 balls WITHOUT REPLACEMENT, only returning the 5 balls after each set has been drawn, we will end up with a record something like:

05(B)  15(B)  23(R)  32(B)  36(R)

10(R)  12(B)  24(R)  29(B)  30(B)

15(B)  25(R)  26(B)  36(R)  39(B)

etc

etc

You get the idea, RIGHT?

Now, I would be shocked, if after several hundred or thousand drawings of sets of 5 balls, that SOME HUMAN BEING would claim that there would NOT be a preponderance of BLUE balls!  In fact, it should not take long before the number of BLUE balls drawn [divided by] the number of RED balls drawn would approach a value of TWO(2).  OK?  ARE YOU SURE?

BUT WAIT!

These sets of five balls have numbers imprinted on them underneath the RED and BLUE paint, the NUMBERS which will be used to determine the winning tickets!

AND GUESS WHAT?  A lottery commission has agreed to pay people holding tickets with the results of these drawings REAL MONEY, BUT, based on the numbers (1-39) printed on them, WITH NO REGARD FOR THE PAINT COLORS!

So, if you can think of the RED and BLUE colors as identifiers for ODD / EVEN NUMBERS, or some other attribute, it should be CLEAR that the winnings you earn will have ABSOLUTELY NOTHING to do with the assignmemt of the COLORS to the balls, which could have been made by children playing, or some other RANDOM PROCESS.

In fact, each of the balls COULD have been assigned a unique hieroglyphic, with NO reference to ANY NUMBER SYSTEM!!!

I KNOW, you're wondering!  But the Colors in our ODD/EVEN case were not assigned randomly, but based on the REMAINDER resulting from dividing the ball number by 2.  When that remainder was ZERO, it was called EVEN; when it was ONE, it was called ODD.  However, whether a ball is painted RED because it has an EVEN number on it, or because a child threw it out of a box, or because its hieroglyphic refers to the male gender, it:

1)  Will not FEEL any different to the ball machine!

2)  Does NOT weigh any more or less than any other ball

3)  Has no more resistance to the air currents in the machine than any other ball

4)  Does NOT bounce any higher than any other ball

..... ETC., ETC.,,,,,,,

(To repeat the opening line above:)

The observed distributions of the difference in frequency counts of ODD numbers versus EVEN in a SET, the SUMS of the five numbers, the number of unique occurences of (0-9) digits in a set, etc, etc, etc, are of NO VALUE in trying to determine an "EDGE" in selecting numbers in a lottery...

BECAUSE...

THE ASSIGNMENT OF ODD / EVEN, DIGIT COUNT, or SUM of a set, etc., HAS NOTHING TO DO WITH ANYTHING CONNECTED TO THE FORCES WHICH CAUSE A PARTICULAR BALL TO EMERGE FROM THE MACHINE, BECAUSE THESE ASSIGNMENTS ARE NOTHING MORE THAN...

LABELS,

LABELS,

LABELS,

LABELS,

--Jimmy4164

It is principally at games of chance that a multitude of illusions
support hope and sustain it against unfavourable chances.
(Laplace, 1796)

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: January 16, 2011, 2:51 am - IP Logged

Right!

"Anyone who thinks the statistical significance of these deviations is sufficient to claim an "edge" over 5460 Draws is delusional!"

Define "statistical significance"

You prove my point again that statistics has no significance here at

all, Not now not ever. The 5-43 matrix is biased by the one extra

odd number and playing more odd then even will win more often then

more even then odd. Just a fact!

RL

A result has Statistical Significance if it is beyond what would be expected by chance.

"The 5-43 matrix is biased by the one extra

odd number and playing more odd then even will win more often then

more even then odd. Just a fact!"

This is the fatal flaw in your reasoning.  Please read the analysis in my post immediately above, and then study to thoroughly understand the work pointed to here, before posting again, PLEASE!

http://www.staff.city.ac.uk/~sj361/p1369.pdf

United States
Member #59354
March 13, 2008
4322 Posts
Offline
 Posted: January 16, 2011, 6:55 am - IP Logged

jimmy

My state uses RNG's to pick numbers so nothing you said about the balls makes any difference

to me and I don't follow that logic anyway.   I don't fault the math or even your resoning.  Lets

say that the drawing used small particles that were encoded with a 5 number code.

There are 575757 particles in the drum from which only one would be selected.   There are 29,241

particles with more odd then even numbers.  This would show no real statical advantage that

could be used except that these patricles will be drawn more often, about 5% more on average.

I checked the (FY-2010) stats for show me 5 and found that the total ticket sales were \$26,700,000

for the game.  MO. runs it's games 7 days a week 365 days a year so they sell around \$73,150 tickets

each day on average. The probability for a 4 of 5 match is 1 in 3386.8 so on average 21.5  match 4

tickets are sold for each draw.  Some days only have 7 or 8  while others go as high as 25 to 30.   The

problem I have with  statictics is this.

The only way to make it work is to add everything up and average it out over a long period of time.

On a day to day analysis as I have showed the results fluctuate wildly.  One would think this is attributed

to ticket sales but that is not always the case.

When better then expected results are thrown into the pot and averaged over the course of a year then

what happends, "NO Statistical Significance."   If no advantage could be found then I would have quit

programming lottery software long ago.  I do have bad days and sometimes bad months but I still beat

the odds overall for what I play doing better then statictics would suggest possible.

Breakdown of 5-39 matrix

sets with more odd then even = 302,499     52.5%

sets with more even then odd = 273, 258    47.5%

P.S.  I still think that no amount of Statistical Analysis will ever help me win a jackpot and so it has

nothing useful to offer me when picking my sets.

rl

Kentucky
United States
Member #32652
February 14, 2006
7546 Posts
Offline
 Posted: January 16, 2011, 11:27 am - IP Logged

We all had to start somewhere.

Several years ago I found what I thought was a neat little pattern in the numbers and set about writing code that would do the grunt work without making mistakes. It turned out that the little pattern worked just fine, but produced an unplayable total number of combos. That is -- the "winner" fit the pattern, but so did oodles of "losers".

Years ago I downloaded a free wheeling program and had lots of fun trying to filter all the possible combos in a 5/39 game down to a playable size. when I got it down to about 9000 combos and add one more filter, I'd get "0" combs and an error code saying I used conflicting filters. The lowest I ever got was about 3500 combos which is not my idea of a playable size.

As for systems that actually work, a co-worker of mine keyed on three numbers and played them with ten other numbers for \$5. He had showed me his tickets and about I week later I saw where all three of his numbers were drawn along with the number "1" that I had remembered being on his tickets. I called him and asked if he'd played and said "yes" but hadn't checked what was drawn. After I told him what numbers were drawn, there was a moment of silence followed by some screaming by his wife; he just won \$100,000!

Tiggs would probably say "it was pure luck" and maybe he's right, but there are methods of play where jackpots can be won for small wagers.

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: January 16, 2011, 12:20 pm - IP Logged

jimmy

My state uses RNG's to pick numbers so nothing you said about the balls makes any difference

to me and I don't follow that logic anyway.   I don't fault the math or even your resoning.  Lets

say that the drawing used small particles that were encoded with a 5 number code.

There are 575757 particles in the drum from which only one would be selected.   There are 29,241

particles with more odd then even numbers.  This would show no real statical advantage that

could be used except that these patricles will be drawn more often, about 5% more on average.

I checked the (FY-2010) stats for show me 5 and found that the total ticket sales were \$26,700,000

for the game.  MO. runs it's games 7 days a week 365 days a year so they sell around \$73,150 tickets

each day on average. The probability for a 4 of 5 match is 1 in 3386.8 so on average 21.5  match 4

tickets are sold for each draw.  Some days only have 7 or 8  while others go as high as 25 to 30.   The

problem I have with  statictics is this.

The only way to make it work is to add everything up and average it out over a long period of time.

On a day to day analysis as I have showed the results fluctuate wildly.  One would think this is attributed

to ticket sales but that is not always the case.

When better then expected results are thrown into the pot and averaged over the course of a year then

what happends, "NO Statistical Significance."   If no advantage could be found then I would have quit

programming lottery software long ago.  I do have bad days and sometimes bad months but I still beat

the odds overall for what I play doing better then statictics would suggest possible.

Breakdown of 5-39 matrix

sets with more odd then even = 302,499     52.5%

sets with more even then odd = 273, 258    47.5%

P.S.  I still think that no amount of Statistical Analysis will ever help me win a jackpot and so it has

nothing useful to offer me when picking my sets.

rl

RL-RANDOMLOGIC,

Your observation of the difference in the frequency of sets with more or less odd or even numbers is analogous to observing that the sum of the 5 numbers in a majority of sets falls into a minority of the total range of possible sums.  The same is true for observing the number of unique digits in a set, or the average "spread" between the numbers.  None of these patterns have anything to do with the probability of any particular set emerging.

They just ARE!

I thought my pointing out that the balls could be labeled with hieroglyphs, thus eliminating any reference to number systems at all, would make the above clear.  Apparently not.  Let's label all of your 575757 particles with unique hieroglyphs, where each "particle" refers to 5 hieroglyphs from a different set of 39.  If this doesn't help you to see that the ARBITRARY assignment of numbers from an ARBITRARILY [BASED] number system is of NO CONSEQUENCE in this situation, I have only one other suggestion:

As a programmer, you should be able to see what I'm telling you is true if you write a simulation which generates random [1...39] sets, using your criteria for what you think are more likely combinations.  E.G., you could generate a random set, and reject it if its number of ODD numbers falls below a certain threshhold.  Once you've chosen a set you like, read the next ACTUAL DRAW result from a file, and score the resulting comparison.  You will have to program it to recognize and record all the small prizes because there are not yet enough lottery results available to expect many jackpots!  Do this for all the draw data you've got, and see what happens.  I think you will be disappointed.

With the program thus written, you could then follow up by disabling the rejection procedure, accepting any and all RNG generated sets, letting the program function otherwise in the same way.  If you use all the data from [say] Missouri, I think you will find the totals will look much the same as those obtained above selecting more ODDs.  With a little tinkering, you could also test your other hypotheses, the digit frequencies and so on.  You could also ignore the Actual Draw dataset and generate Random Draws.  This way you could simulate centuries of results!

I'm serious.  There were PhD mathematicians who could not be convinced to "Switch" doors in the Monty Hall Problem until they wrote their own [simple] computer simulation and observed the results.

I believe an ANALOGY can be drawn between the problem featured in this video, and our problem at hand.  Besides, it's a fun video to watch.  Let's try to be friends.

--JIMMY4164

bgonÃ§alves
Brasil
Member #92564
June 9, 2010
2264 Posts
Offline
 Posted: January 16, 2011, 4:40 pm - IP Logged

The central limit theorem (CLT)

bgonÃ§alves
Brasil
Member #92564
June 9, 2010
2264 Posts
Offline
 Posted: January 16, 2011, 4:46 pm - IP Logged

olá RL =

For your purpose, you are far from dealing with so big numbers, so you
must take the difference between impossible and improbable much more
seriously. E.g., the chances for number 3 to be drawn in 10 consecutive
drawings may be very small, but they are not exactly zero. And if it
happens, it has no impact on whether the 11-th time number 3 will be
drawn again. But of course, the 11-th time, just like any time before,
drawing of number 3 is not certain. And that is the only reason why
getting number 3 now 11 times in a row is less probable then getting it
10 times. And getting it 12 times is even less probable and so on. But
there is no sharp limit, beyond which it is impossible.

There is no absolute limit, it’s all about probabilities and these
change as you get more information. Once you know that number 3 was
already drawn 9 times in a row, it makes no sense to say that it is
“almost impossible” to get a line of 10. You are only one drawing
away from it and that drawing knows nothing of the other 9.
Probability is always about what you don’t know, not about what you
know. Calculating a probability for something that involves already
known parameters as if all values for those parameters were still
possible is simply based on a wrong assumption.

The central limit theorem (CLT) makes no promise that the drawings in
the future will compensate for any improbable events in the past, they
don’t care for the past. They will lead to normal distribution all
right, but not by compensating for the past, rather by dwarfing the past
(if there is enough of them, of course), and that is a difference. It
still means that the past gives you no useful information about the
future, as you would like.

Using complicated math like trigonometry etc. can’t change this basic
fact. It’s a classic mistake in science, especially statistics, when
people think that if you use complicated math and do the numbers right,
you can’t be wrong. Unfortunately, with oversimplified interpretation
of results, it can all be useless.

United States
Member #59354
March 13, 2008
4322 Posts
Offline
 Posted: January 16, 2011, 8:54 pm - IP Logged

Jimmy

Your observation of the difference in the frequency of sets with more or less odd or even numbers is analogous to observing that the sum of the 5 numbers in a majority of sets falls into a minority of the total range of possible sums.  The same is true for observing the number of unique digits in a set, or the average "spread" between the numbers.  None of these patterns have anything to do with the probability of any particular set emerging.

They just ARE!

No problems here, exactly how I see it.

I thought my pointing out that the balls could be labeled with hieroglyphs, thus eliminating any reference to number systems at all, would make the above clear.  Apparently not.  Let's label all of your 575757 particles with unique hieroglyphs, where each "particle" refers to 5 hieroglyphs from a different set of 39.  If this doesn't help you to see that the ARBITRARY assignment of numbers from an ARBITRARILY [BASED] number system is of NO CONSEQUENCE in this situation, I have only one other suggestion:

Here is a bump in the road.  Nothing in my way of thinking has anything to do with what can or

cannot be drawn.  The lottery drawing process in my opinion can never be calculated with any

certainty and that's that.  However,  when building a set of numbers to play it only makes since to

incorporate information that best mimics the the largest  sample of possibilities within the matrix

and then play from that pool.   I have to make choices even if it is a choice to purchase a QP.

My system of play requires me to make on most days 5 or 6 main choices out of 10.  This

gives me far fewer choices then making 5 of 39.  Three of these choices are no brainers most

days and so I only need to make 2 or 3 choices from the remaining 7.   1 in 21 or 1 in 35 for the

final 2 or 3 selections.  There is no certainty that these choices will be correct and the final 2 or 3

are nothing more then my best guess that is made from viewing the bias for different time frames

within the past drawing which are then compared to the entire matrix.

Here is another example that I think will allow you to understand my thinking as to the selection

process.  Forget the balls, numbers, colors, ect, ect, ect.  What if the lottery required  picking one

number from 000001 to 575757.  This is really no different then the way the lottery picks its

numbers.   Pick-3 and pick-4 games are also the same thing. P3 =000 to 999 and P4 = 0000 to

9999.  The drums are for intertainment only and without them many would not play and that is why

the state lotteries that use RNG's build software replicas of mechanical drawing machines

so that people can watch their numbers being drawn.  How exiciting is watching a random number

generator draw one set?  Lets say that I select  "193456" to play, which  = set 03 19 24 35 36 for

those interested.  I could just selected my number at random but this also would be boaring and I

would not play.  I once posted a reply to you about going to a ball game to watch my local team

and spent over 100 bucks for a little fun.  I would have gave \$500.00 to get to play 2nd base for

a couple innings just to particapate.   Back to the lottery, If the lottery ever decides to convert to

selecting the winning ticket by the method above then you can bet I would be breaking the number

down into 6 digits and make my best effort to pick them.  My efforts have no effect on what is

drawn but from my own experience I make better choices then random chance would account for

using as much data as I can.

As a programmer, you should be able to see what I'm telling you is true if you write a simulation which generates random [1...39] sets, using your criteria for what you think are more likely combinations.  E.G., you could generate a random set, and reject it if its number of ODD numbers falls below a certain threshhold.  Once you've chosen a set you like, read the next ACTUAL DRAW result from a file, and score the resulting comparison.  You will have to program it to recognize and record all the small prizes because there are not yet enough lottery results available to expect many jackpots!  Do this for all the draw data you've got, and see what happens.  I think you will be disappointed.

All of the test you mentioned above I have already made and I run them very often as I make

changes to my program.  I don't ever try to match number for number but look at the results

from the point of sets vs matrix.

I have a program that builds a random database of the same amount of draws with all filters that

are included in my software.

When compared to the actual database there is very little difference and many times filter value

totals are less then +/-  .06% overall.  This leads me to believe that I am correct in my assumptions

that what happends most is a result of the draws following the matrix and is unrelated to the device

used to pick the numbers.  While the sets picked are suffency random in nature the information

contained within them is somewhat predictable.  Not so much day to day but within blocks of

drawings which leads me to the watch and wait method of playing.  I have tried many times to

backtest this system and it cannot be done because it requires the human element.  If I run a

static backtest using random generated variables then the program becomes random in nature

and gives expected results.   My software is a tool and even the bayesian tool returns a suggested

set of values that are ranged from best to worst gotten from two sorces, the matrix and many blocks

of actual data that when compared is evaluated for it's consistency.  Like using the DNA from one

person to link another.  How close the match the higher it goes on the list.

There is nothing that predicts anything as it only accepts inputs from the user to build the sets.

The generator goes through every possible set and only accepts sets that pass the users inputs.

I posted a video of my software to show how few choices are needed to reduce over 5 million

to less then 20 sets.  Some settings will produce many more but on those days I don't play.  I

cannot explain the results I get using any math known to me and thus the anomaly must be the

human element.

RL

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: January 17, 2011, 2:27 am - IP Logged

Jimmy

Your observation of the difference in the frequency of sets with more or less odd or even numbers is analogous to observing that the sum of the 5 numbers in a majority of sets falls into a minority of the total range of possible sums.  The same is true for observing the number of unique digits in a set, or the average "spread" between the numbers.  None of these patterns have anything to do with the probability of any particular set emerging.

They just ARE!

No problems here, exactly how I see it.

I thought my pointing out that the balls could be labeled with hieroglyphs, thus eliminating any reference to number systems at all, would make the above clear.  Apparently not.  Let's label all of your 575757 particles with unique hieroglyphs, where each "particle" refers to 5 hieroglyphs from a different set of 39.  If this doesn't help you to see that the ARBITRARY assignment of numbers from an ARBITRARILY [BASED] number system is of NO CONSEQUENCE in this situation, I have only one other suggestion:

Here is a bump in the road.  Nothing in my way of thinking has anything to do with what can or

cannot be drawn.  The lottery drawing process in my opinion can never be calculated with any

certainty and that's that.  However,  when building a set of numbers to play it only makes since to

incorporate information that best mimics the the largest  sample of possibilities within the matrix

and then play from that pool.   I have to make choices even if it is a choice to purchase a QP.

My system of play requires me to make on most days 5 or 6 main choices out of 10.  This

gives me far fewer choices then making 5 of 39.  Three of these choices are no brainers most

days and so I only need to make 2 or 3 choices from the remaining 7.   1 in 21 or 1 in 35 for the

final 2 or 3 selections.  There is no certainty that these choices will be correct and the final 2 or 3

are nothing more then my best guess that is made from viewing the bias for different time frames

within the past drawing which are then compared to the entire matrix.

Here is another example that I think will allow you to understand my thinking as to the selection

process.  Forget the balls, numbers, colors, ect, ect, ect.  What if the lottery required  picking one

number from 000001 to 575757.  This is really no different then the way the lottery picks its

numbers.   Pick-3 and pick-4 games are also the same thing. P3 =000 to 999 and P4 = 0000 to

9999.  The drums are for intertainment only and without them many would not play and that is why

the state lotteries that use RNG's build software replicas of mechanical drawing machines

so that people can watch their numbers being drawn.  How exiciting is watching a random number

generator draw one set?  Lets say that I select  "193456" to play, which  = set 03 19 24 35 36 for

those interested.  I could just selected my number at random but this also would be boaring and I

would not play.  I once posted a reply to you about going to a ball game to watch my local team

and spent over 100 bucks for a little fun.  I would have gave \$500.00 to get to play 2nd base for

a couple innings just to particapate.   Back to the lottery, If the lottery ever decides to convert to

selecting the winning ticket by the method above then you can bet I would be breaking the number

down into 6 digits and make my best effort to pick them.  My efforts have no effect on what is

drawn but from my own experience I make better choices then random chance would account for

using as much data as I can.

As a programmer, you should be able to see what I'm telling you is true if you write a simulation which generates random [1...39] sets, using your criteria for what you think are more likely combinations.  E.G., you could generate a random set, and reject it if its number of ODD numbers falls below a certain threshhold.  Once you've chosen a set you like, read the next ACTUAL DRAW result from a file, and score the resulting comparison.  You will have to program it to recognize and record all the small prizes because there are not yet enough lottery results available to expect many jackpots!  Do this for all the draw data you've got, and see what happens.  I think you will be disappointed.

All of the test you mentioned above I have already made and I run them very often as I make

changes to my program.  I don't ever try to match number for number but look at the results

from the point of sets vs matrix.

I have a program that builds a random database of the same amount of draws with all filters that

are included in my software.

When compared to the actual database there is very little difference and many times filter value

totals are less then +/-  .06% overall.  This leads me to believe that I am correct in my assumptions

that what happends most is a result of the draws following the matrix and is unrelated to the device

used to pick the numbers.  While the sets picked are suffency random in nature the information

contained within them is somewhat predictable.  Not so much day to day but within blocks of

drawings which leads me to the watch and wait method of playing.  I have tried many times to

backtest this system and it cannot be done because it requires the human element.  If I run a

static backtest using random generated variables then the program becomes random in nature

and gives expected results.   My software is a tool and even the bayesian tool returns a suggested

set of values that are ranged from best to worst gotten from two sorces, the matrix and many blocks

of actual data that when compared is evaluated for it's consistency.  Like using the DNA from one

person to link another.  How close the match the higher it goes on the list.

There is nothing that predicts anything as it only accepts inputs from the user to build the sets.

The generator goes through every possible set and only accepts sets that pass the users inputs.

I posted a video of my software to show how few choices are needed to reduce over 5 million

to less then 20 sets.  Some settings will produce many more but on those days I don't play.  I

cannot explain the results I get using any math known to me and thus the anomaly must be the

human element.

RL

RL-RANDOMLOGIC,

I was considering commenting on Dr San's discussion of the Central Limit Theorem (Nice Summary Dr San), to say that your methods might not be ignoring it and that you might not be committing the Gambler's or Hot Hand fallacies, but coming to false conclusions through some other as yet undefined fallacy.  However, after reading this, your latest post, I must return to my earlier suspicions.  Based on your own words, you clearly are making decisions based on previous draw history, even if only with a few days worth.

I say this because of statements like:

"My system of play requires me to make on most days 5 or 6 main choices out of 10."

"There is nothing that predicts anything as it only accepts inputs from the user to build the sets.  The generator goes through every possible set and only accepts sets that pass the users inputs.  I posted a video of my software to show how few choices are needed to reduce over 5 million to less then 20 sets.  Some settings will produce many more but on those days I don't play.  I cannot explain the results I get using any math known to me and thus the anomaly must be the human element."  (And a little LUCK?)

The statements above, especially the underlined phrases, reveal that your methods definitely require knowlege of the past history of the game's draws.  Otherwise, the "users inputs" would not need to vary from one day to the next.  The only other reasons that could require you to choose different sets of 5 numbers on one day, versus another, would have to be in the realm of future forecasting, or some sort of voodoo!  I don't remember you ever mentioning anything in that regard, so I must assume you are looking back, and consequently, you are committing either the Gambler's Fallacy, or the Hot Hand Fallacy, or both, further complicated by your erroneos assumptions about the significance of the frequencies of various numerological labeling phenomena.

"I have tried many times to backtest this system and it cannot be done because it requires the human element."

The heuristics required to simulate the types of systems we're talking about here are elementary compared to what's being done in the fields of Robotics, Systems Engineering, Operations Research, and Artificial Intelligence in general today.  You must know this.  I'm sorry RL, but your claim is what is usually referred to as a "cop out."  As long as you hide behind the claim that your system can't be backtested, you will continue to delude yourself into thinking that you have a better chance than others to win a jackpot.  Take another look at my two 2010 summaries of MadDog's Powerball Challenge, one broken out by Player, the other by Date.

There is one thing we can probably agree on -- your methods will NOT cause you to LOSE any more than QP players while you wait for your ship to come in.  Our binoculars are just as good as yours as we scan that horizon for our "Cargo!"

Sorry, but that's the way I see it.

--Jimmy4164

United States
Member #59354
March 13, 2008
4322 Posts
Offline
 Posted: January 17, 2011, 9:37 am - IP Logged

Jimmy

Should you ever grow a brain of your own then maybe you will see through your misconceptions.

The fallcy you mention stems from a belief that you think, I think that past draws have something

to do with the future draws,  Not so.  I do not think you understand what I mean when I say that

the draws will follow the matrix.  The Bayesian program could be given those attributes but you

must first consider that it is not looking at the numbers but data that is gotten from the entire matrix.

You have proven that you do not understand the fallacy you speak of.  If you were trying to make a

stock choice based on any gained knowledge then you would be guilty of the fallacy yourself according

to your reply and any wager or any bet by any other person would fall under the same fallacy that

was not selected at random.

Lets go back to the odd/even lession,  Lets say that the you have six stocks in your portfolio labeled

as

zeroodd

oneodd

twoodd

threeodd

fourodd

fiveodd

Each time the set drawn has zero odd numbers then the price for zeroodd remains unchanged meaning

that you money is doing nothing,  and each time the zero odd hits then the stock rises one point. You

evaluate your portfolio and see that twoodd and  threeodd has out performed all the others by at least

2 to 1.  My question is which stocks would you keep and which would you sell.  What is the chance that

zeroodd, oneodd, fourodd and fiveodd will ever perform at the same level as twoodd and threeodd.

Again lets say that you find 6 more stocks named prime0, prime1, prime2, prime3, prime4 and prime5.

prime1 has out performed prime5 by 276 to 1.  You do a little research and you find out why prime5 has

performed so badly compared to prime1.

Now a little math.

for a 5-39 matrix

there are  792 sets with 5 prime numbers

there are 210600 sets with 1 prime number

5 primes has been drawn one time

1 primes has been drawn 276 times

1 / 729 = .001262626

276 / 210600 = .001310541

NO real Statical Difference! so I guess that you would buy the prime5 stock anyway.  I don't play what

will happen based on what has happened but what can and will happen most.

Backtest

Lets say that I play 2 to 3 odd most days because of the matrix, I make a run and because of the

digits and other setting I have used I get too many numbers to play.  I decide to play 3 odd instead

of 2 to 3 so that I can reduce my sets. again lets say that I still have too many numbers to play, I

may change the digits or other filters I have selected and rerun.  Write me a program that can make

a logical selection based on the best information that can be gotten and I will be glad to run and test

it for you.  The fact that I win more then the odds would suggest is proof enough and requires no other

testing in my opinion.

If this falls into the fallacy then so be it.  And If playing digits 1-2-3 along with my best choice of 2 or 3

more is a fallacy then so be it.  The way you see it every selection that anyone could make regarding

anything falls into the fallacy and that includes the markets, horse races, blackjack, poker,  driving to work,

eating at a restaurant, flying on a plane, going to the doctor, taking a prescription,  jogging, walking down

stairs, kissing, and everything else from birth to the grave.  I have to make a decision to get out of bed

in the morning, I guess I should just flip a coin and take the work out of it kind of like buying a QP.

Luck does not exist but Chance however does.  When I

win more often then Chance can account for then it is

no longer chance so what is it.  That should send your

brain into an endless loop because it does not compute,

does not compute, does not compute, does not compute,

does not compute, does not compute, does not compute.....

RL

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: January 17, 2011, 1:17 pm - IP Logged

Jimmy

Should you ever grow a brain of your own then maybe you will see through your misconceptions.

The fallcy you mention stems from a belief that you think, I think that past draws have something

to do with the future draws,  Not so.  I do not think you understand what I mean when I say that

the draws will follow the matrix.  The Bayesian program could be given those attributes but you

must first consider that it is not looking at the numbers but data that is gotten from the entire matrix.

You have proven that you do not understand the fallacy you speak of.  If you were trying to make a

stock choice based on any gained knowledge then you would be guilty of the fallacy yourself according

to your reply and any wager or any bet by any other person would fall under the same fallacy that

was not selected at random.

Lets go back to the odd/even lession,  Lets say that the you have six stocks in your portfolio labeled

as

zeroodd

oneodd

twoodd

threeodd

fourodd

fiveodd

Each time the set drawn has zero odd numbers then the price for zeroodd remains unchanged meaning

that you money is doing nothing,  and each time the zero odd hits then the stock rises one point. You

evaluate your portfolio and see that twoodd and  threeodd has out performed all the others by at least

2 to 1.  My question is which stocks would you keep and which would you sell.  What is the chance that

zeroodd, oneodd, fourodd and fiveodd will ever perform at the same level as twoodd and threeodd.

Again lets say that you find 6 more stocks named prime0, prime1, prime2, prime3, prime4 and prime5.

prime1 has out performed prime5 by 276 to 1.  You do a little research and you find out why prime5 has

performed so badly compared to prime1.

Now a little math.

for a 5-39 matrix

there are  792 sets with 5 prime numbers

there are 210600 sets with 1 prime number

5 primes has been drawn one time

1 primes has been drawn 276 times

1 / 729 = .001262626

276 / 210600 = .001310541

NO real Statical Difference! so I guess that you would buy the prime5 stock anyway.  I don't play what

will happen based on what has happened but what can and will happen most.

Backtest

Lets say that I play 2 to 3 odd most days because of the matrix, I make a run and because of the

digits and other setting I have used I get too many numbers to play.  I decide to play 3 odd instead

of 2 to 3 so that I can reduce my sets. again lets say that I still have too many numbers to play, I

may change the digits or other filters I have selected and rerun.  Write me a program that can make

a logical selection based on the best information that can be gotten and I will be glad to run and test

it for you.  The fact that I win more then the odds would suggest is proof enough and requires no other

testing in my opinion.

If this falls into the fallacy then so be it.  And If playing digits 1-2-3 along with my best choice of 2 or 3

more is a fallacy then so be it.  The way you see it every selection that anyone could make regarding

anything falls into the fallacy and that includes the markets, horse races, blackjack, poker,  driving to work,

eating at a restaurant, flying on a plane, going to the doctor, taking a prescription,  jogging, walking down

stairs, kissing, and everything else from birth to the grave.  I have to make a decision to get out of bed

in the morning, I guess I should just flip a coin and take the work out of it kind of like buying a QP.

Luck does not exist but Chance however does.  When I

win more often then Chance can account for then it is

no longer chance so what is it.  That should send your

brain into an endless loop because it does not compute,

does not compute, does not compute, does not compute,

does not compute, does not compute, does not compute.....

RL

RL-RANDOMLOGIC,

You said, "Lets say that I play 2 to 3 odd most days because of the matrix, I make a run and because of the digits and other setting I have used I get too many numbers to play."

What sorts of events or observations prompt you to make one setting over another in your various parameters?  And what would compel you to use different settings today, from those you used yesterday?

--Jimmy4164

 Page 3 of 10