Welcome Guest
You last visited January 20, 2017, 8:51 am
All times shown are
Eastern Time (GMT-5:00)

# Working on algorithm

Topic closed. 30 replies. Last post 8 months ago by RJOh.

 Page 1 of 3
mid-Ohio
United States
Member #9
March 24, 2001
19900 Posts
Offline
 Posted: April 24, 2016, 7:28 pm - IP Logged

Working on an algorithm for predicting combinations for jackpot games.  Besides avoiding combinations that have already won, does anyone have any suggestions?

I've come up with a way to classify previous drawing using letters such that 04/19/2016 MM results were ABCDE/ACDEF.  Checking all the MM drawings since its last matrix change, I found that 13% of the MM drawings were ABCDE/ and ABCDF/.  So far there are 133 classifications with those two repeating 31 times.

Using what I've come up with so far I've been posting ten MM lines on the prediction board and the best I've done is a 3+0 but mostly 0+1 and 1+1.

I don't expect to get a decent win every time but do better the 13% of the time the winning combinations have the most popular classification.

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

United States
Member #172415
January 23, 2016
3795 Posts
Offline
 Posted: April 24, 2016, 7:30 pm - IP Logged

Working on an algorithm for predicting combinations for jackpot games.  Besides avoiding combinations that have already won, does anyone have any suggestions?

I've come up with a way to classify previous drawing using letters such that 04/19/2016 MM results were ABCDE/ACDEF.  Checking all the MM drawings since its last matrix change, I found that 13% of the MM drawings were ABCDE/ and ABCDF/.  So far there are 133 classifications with those two repeating 31 times.

Using what I've come up with so far I've been posting ten MM lines on the prediction board and the best I've done is a 3+0 but mostly 0+1 and 1+1.

I don't expect to get a decent win every time but do better the 13% of the time the winning combinations have the most popular classification.

What do u mean by ABCDE

mid-Ohio
United States
Member #9
March 24, 2001
19900 Posts
Offline
 Posted: April 24, 2016, 7:37 pm - IP Logged

What do u mean by ABCDE

I replace the numbers with letters.

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

United States
Member #172415
January 23, 2016
3795 Posts
Offline
 Posted: April 24, 2016, 8:24 pm - IP Logged

I replace the numbers with letters.

But which numbers?

New Jersey
United States
Member #17843
June 28, 2005
51100 Posts
Online
 Posted: April 24, 2016, 10:00 pm - IP Logged

But which numbers?

The Secret Numbers...

A mind once stretched by a new idea never returns to its original dimensions!

Texas
United States
Member #55889
October 23, 2007
5760 Posts
Offline
 Posted: April 25, 2016, 9:40 am - IP Logged

But which numbers?

Probably dividing 75 numbers into 5 groups of 15 numbers.

Years ago, when the matrix was 56 numbers, I had a chart that listed how many times each number was drawn in each position. I could also use that to determine what groups such as 1-9, 10-19, 20-29, etc played in each position the most. Unfortunately, my computer got a virus, and I ended up losing my info. In the end, that info didn't really help me win anything significant.

CAN'T WIN IF YOU'RE NOT IN

A DOLLAR AND A DREAM (OR \$2)

Florida
United States
Member #135615
November 27, 2012
405 Posts
Offline
 Posted: April 25, 2016, 11:54 am - IP Logged

My last attempt at at Powerball (I tend to stick to Cash 3 but occasionally I branch out). I started trying to apply the same metrics I use on Cash 3. It didn't take long to realize that since there's one pool of many balls instead of three small pools of balls, I'd need to take a different approach. What lane the balls come in is false. The balls are all drawn out of order and then put in order on the website. So basing anything on the order seemed fruitless.

From there, I started working on pairs. First, I figured out what numbers appear the most often. With the first five numbers that appear in that list, I made five different sets. I took that first number of each set and found the next two numbers for each set by finding what numbers hit most often with it. Once with both the top number and once without. That gave me three numbers for each of the five sets. I did some work with Sums to try and figure out the last three numbers, making sure the final numbers where not popular matches (need to account for entropy). However, I never found anything worth playing. It seemed too random so I went back to my 1:1000 odds of Cash 3. :)

San Angelo, Texas
United States
Member #1097
January 31, 2003
1405 Posts
Offline
 Posted: April 25, 2016, 1:47 pm - IP Logged

Working on an algorithm for predicting combinations for jackpot games.  Besides avoiding combinations that have already won, does anyone have any suggestions?

I've come up with a way to classify previous drawing using letters such that 04/19/2016 MM results were ABCDE/ACDEF.  Checking all the MM drawings since its last matrix change, I found that 13% of the MM drawings were ABCDE/ and ABCDF/.  So far there are 133 classifications with those two repeating 31 times.

Using what I've come up with so far I've been posting ten MM lines on the prediction board and the best I've done is a 3+0 but mostly 0+1 and 1+1.

I don't expect to get a decent win every time but do better the 13% of the time the winning combinations have the most popular classification.

I've been using alpha-numeric substitution for many years, plus some important add-ons.
It's probably one of the best methods for TREND analysis, which we need to make good educated guesses.
I assume you are making the substitutions by Decades.
In a 75-integer game, there are 8 Decades - A, B, C, D, E, F, G, H.
A, of course, has 9 integers, H has 6 integers.
You may be including more integers per Decade, but, I've never been able to understand the
rationale for the modifications. Seems its an artificial method that is in opposition to the normal pathway. But, that's my opinion.
One of the reasons I quit playing the Jackpot games is that there are just too many alpha signatures, or arrangements, to track.
The tracking charts just don't provide any useful clues, as they do for games with lesser
integers, because there is insufficient data.
In other words, not enough repetition.
A method I use for TX Cash 5, which has 4 Decades, is break the arrangements
into smaller groups.
First 3, Last 3; First 2, Middle 1 and Last 2.
I maintain running counts, of course.
As result, the final signatures are formatted in a way to include all of the  higher counts.
I play 5 sets per drawing using 5 different Alpha arrangements.
Thus, in TX C5, ABCCD shows up more often.
Again, I employ some add-on methods that break the games down into much smaller portions.
Too much to explain here.
I tried breaking Mega Millions down into smaller pieces, but, again, not enough repetition to create any trends.
I'm not a computer guy and I know nothing about creating algorithms.
All I can offer is that you will be probably be dealing with  a lot of heartburn. But, stick with it. You just might have a good day and
choose the winning integers.
Have a good one!
Bobby623
bgonÃ§alves
Brasil
Member #92564
June 9, 2010
2134 Posts
Offline
 Posted: April 25, 2016, 4:38 pm - IP Logged

Standards in the Cartesian plane of the numbers themselves and separate initial and final digit

mid-Ohio
United States
Member #9
March 24, 2001
19900 Posts
Offline
 Posted: April 26, 2016, 11:53 am - IP Logged

The Secret Numbers...

The numbers aren't secret, a few years ago I posted a thread explaining them. I stop posting to it when several members said it was becoming lengthy and boring. I tried locating it so I could post a link to it but couldn't.

Seems others have tried somthing similar without any success with jackpot games. Since jackpot games are my favorates I'll continue working on it privately and maybe mention it again when I have somethig worth mentioning.

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

Kentucky
United States
Member #32652
February 14, 2006
7344 Posts
Offline
 Posted: April 27, 2016, 12:32 am - IP Logged

Working on an algorithm for predicting combinations for jackpot games.  Besides avoiding combinations that have already won, does anyone have any suggestions?

I've come up with a way to classify previous drawing using letters such that 04/19/2016 MM results were ABCDE/ACDEF.  Checking all the MM drawings since its last matrix change, I found that 13% of the MM drawings were ABCDE/ and ABCDF/.  So far there are 133 classifications with those two repeating 31 times.

Using what I've come up with so far I've been posting ten MM lines on the prediction board and the best I've done is a 3+0 but mostly 0+1 and 1+1.

I don't expect to get a decent win every time but do better the 13% of the time the winning combinations have the most popular classification.

Before the matrix changes, I always thought using "recently drawn numbers" was an interesting strategy, but only if the field can be reduced to "5 out 20". And with limited drawing results, it might be a better way to go.

How many numbers are in each classification?

mid-Ohio
United States
Member #9
March 24, 2001
19900 Posts
Offline
 Posted: April 29, 2016, 10:04 am - IP Logged

Before the matrix changes, I always thought using "recently drawn numbers" was an interesting strategy, but only if the field can be reduced to "5 out 20". And with limited drawing results, it might be a better way to go.

How many numbers are in each classification?

Did the "5 out 20" strategy ever work for you even before the matrix change?

There are several classifications that include all seventy-five numbers which reduces the active numbers to a group of 57 for tonight's drawing.  Including the 15 bonus numbers, the odds of winning a jackpot are still five times harder than that of a 6/49 game.

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

United States
Member #59354
March 13, 2008
4083 Posts
Offline
 Posted: April 30, 2016, 10:57 pm - IP Logged

RJ

What I have found is that by combining my different systems only using the parts which were easiest to predict

seems to be working well.  Over the years I was able to trap the winning set in maybe 1K lines often enough but

always lost it in the final filtering stages.  Using the main cores of several different systems as filters rather than

cores almost seems magical.  Today I made a "after the fact" setup using the old DMP-Pro  similar to what I used

in the past then took those lines and passed them through newly adapted filtering system and in my first run

reduced that list to 2 lines.  Even 25 to 50K lines can be reduced with very little effort using setting that I consider

easy to hit.  While I still have to make some good choices I only use the stuff that in the past I had very high hit

rates.   The old lexie program could reduce my 5-39 to 100 lines with 4 correct digit selections but it's very hard to

hit those 4 digits on a regular bases.  The lexe steps and sc data tools are now ran like filters where only a few of

the easiest values are set.   In the pic below I only used the Steps to reduce the DMP lines to 2 but could have

spread the process to any or all of the 6 game engines making just a couple choices in each one.   I am sure you

have may old systems that were built around some sort of main core,  try taking these main cores and convert

them so that they function as filters.

Rl

Working on my Ph.D.  "University of hard Knocks"

I will consider the opinion that my winnings are a product of chance if you are willing to consider

they are not.  Many great discoveries come while searching for something else

Trump / 2016 & 2020

mid-Ohio
United States
Member #9
March 24, 2001
19900 Posts
Offline
 Posted: May 2, 2016, 3:53 pm - IP Logged

Did the "5 out 20" strategy ever work for you even before the matrix change?

There are several classifications that include all seventy-five numbers which reduces the active numbers to a group of 57 for tonight's drawing.  Including the 15 bonus numbers, the odds of winning a jackpot are still five times harder than that of a 6/49 game.

Inspite of filtering out two of the winning numbers to reduce the pool to 57, the 10 lines I posted on the prediction board matched 2+1 for \$5 which would have covered half the price of playing them.  I would like to get the price of playing even lower.

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

mid-Ohio
United States
Member #9
March 24, 2001
19900 Posts
Offline
 Posted: May 2, 2016, 10:25 pm - IP Logged

RJ

What I have found is that by combining my different systems only using the parts which were easiest to predict

seems to be working well.  Over the years I was able to trap the winning set in maybe 1K lines often enough but

always lost it in the final filtering stages.  Using the main cores of several different systems as filters rather than

cores almost seems magical.  Today I made a "after the fact" setup using the old DMP-Pro  similar to what I used

in the past then took those lines and passed them through newly adapted filtering system and in my first run

reduced that list to 2 lines.  Even 25 to 50K lines can be reduced with very little effort using setting that I consider

easy to hit.  While I still have to make some good choices I only use the stuff that in the past I had very high hit

rates.   The old lexie program could reduce my 5-39 to 100 lines with 4 correct digit selections but it's very hard to

hit those 4 digits on a regular bases.  The lexe steps and sc data tools are now ran like filters where only a few of

the easiest values are set.   In the pic below I only used the Steps to reduce the DMP lines to 2 but could have

spread the process to any or all of the 6 game engines making just a couple choices in each one.   I am sure you

have may old systems that were built around some sort of main core,  try taking these main cores and convert

them so that they function as filters.

Rl

An interesting strategy, looks like you have a program already in the works.

Right now I have to run several programs to get the information I'm using but hopefully once I have some success I can combine the best of them into one program.

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

 Page 1 of 3