Algorithms for number selection?

RJ

Most of the time I play it's a rush to get tickets before they shut down for the draw.  The way I count progress

has nothing to do with tickets but my setup.  The predictions board has no real value to me and I can't see where

posting tickets I don't intend to buy does anything to improve my play.  I think it's more of a bragging contest than

anything else.   Many post more tickets in a week than I would play in a year, I think many would be shocked at how

little I play the lottery.   I accepted a challange from jimbo the clown and started playing 10 to 15 sets every couple

days.  I think I won somewhere around $500 bucks over a 9 day time frame with around $80 invested and then he

sent me a PM asking me not to post them.  Go figure.   My software is as dumb as a rock but if the person at the wheel

understands a few things and can make some good guesses then it will produce some good returns.  Several that

have used it have told me they have won more using it then they did with anything else.   I love to program and the

lottery is just a hobby,  the new release will most likely be the last lottery program I will ever write.   When Jimbo

first came to LP he kept sending me messages wanting me to write stock software for him.  When I finially told him

to bug off that's when he started attacking my digit system post to the point that  todd closed it.   I just looked at the

predictions board and not much there to bragg about, I will be releasing the new version in a week of so and people

can make up their own minds if it's any good.   I would like to see someone hit a JP using it in the next couple months.

Maybe a few people will form a pool and give it a good workout.

RL

I have done much study on lotteries over the years and have found some interesting algorithms on my way.

One hypnotises I came up with was Pick 6 draws tend to bias to the outer individual position distributions over time

Analysis showed this to be true on past history on a large number of draws I tested.

I analysed all possible outcomes which showed why this would be true, it was simply due to no lower numbers than 1 or higher numbers than 49 (in this instance) meaning the field of play on the outer numbers is reduced, hence the higher hit ratios findings in these two outer zones.

A good method of picking numbers in a 6/49 based on the above findings would be to take a sample of the past 10 draws.

Split the 6 numbers in to three groups of 2 (3 groups of 2)

01 02                 03 04                 05 06

---------------------------------------------------

12 16                 23 25                 31 42

20 27                 33 41                 44 46

01 05                 15 22                 26 42

07 08                 16 22                 24 28

10 23                 25 34                 41 47

08 13                 18 21                 40 44

01 03                 09 23                 37 46

18 22                 25 35                 39 41

04 13                 17 26                 27 40

02 07                 08 10                 22 23

Add each vertical column up and divide by 10 to get the average

08 14                 19 25                 33 40

Now pick numbers with the bias found so

B1 No higher than 8

B2 No higher than 14

B3 No higher than 19

B4 No lower than 25

B5 No lower than 33

B6 No lower than 40

After much analysis it can be seen that the two lower and two higher predictions hit far more often than the two inner predictions.

Usually opening kickoff at the super bowl...

Developer

I am working on very much the same thing.  For my 5-39 I use 10 groups, 9 have 4 and the 10th has 3.

The group from which the 1st and 5th numbers will come seems to be the easiest to hit and is very much

a product of what you suggest above.   It may not make us winners overnight but I think it's a step in the

right direction.  I try and select the groups for the outside numbers and many times block a couple on the

inside as this works well for me most days.   the highest number of draws tend to come from 4 groups with

5 coming in second and then 3.  On the days that all 5 numbers come from 4 groups blocking a couple gives

a little better odds over selecting groups to put into play.  I am working on a suggestion algorithm which I hope

will give me the 2 or 3 to block and 2 or 3 of the best to put into play. 

RL

Hello, rl randomlogic, your lottery is 39/5, I give this idea to make the suit or the central trio   Example = 02 12 16 25 37      12,16,25 suit is central, then taking the number 01 and number 39, but because the number can only be one of the 39 th and last position,   So we have 37 numbers in central suit total = 7770 suit in 100% of any lottery, so then number of points is placed after, but these 7770 can create filters to decrease to 5% of this value would be in three bands the tip, the central strip (suits 7770) and and the final track. To close the bet. Now a study of 7770 terminations of digits to see the filters.

dr san

That's all well and good if a person can lock in on the first and last number,  I find it difficult enough to

hit the first and last group and would hate to try and pin it down to one number each.  If I choose 5

groups to play then my total sets is equal to 4X4X4X4X4=1024.  Since I have 4 numbers in each group

except group 10 then I have 16 possible combinations of first and last number pairs.  I can sometimes

remove one number from some of the groups which reduces the total sets before I begin to filter.  I also

have a multi-string wheel that can accept up to 15 numbers in each string.  This works just like the groups

but is not tied to specific numbers and is a little more flexible.   There are many ways to build a subset of

numbers that makes them easier to handle and I guess one method is as good as any other but the main

goal should always focus on it's predictability.  I don't believe in predicting numbers but I do believe that

certain secondary data is very predictable.   My games first number can range from 1 into the 30's so even

the best thought out methods will still be subject to randomness.  Once you have your subset then one can

or should use as much math as possible to reach a playable amount of sets.  I am writting a paper on random

sampling and if people would look at the lottery drawing as taking random samples from the whole matrix

and consider a few things in the process it would change the way they play.  I was never big on using past

data to predict what will happen next but have updated my views in this area.   Much information can be

gleemed if you know what to look for.  The next draw will come from the whole universe of sets but the past

data can give you some expectation as to what it might be.  Just like taking random samples from a production

line one can gain some valuable information.  The method the balls or numbers are drawn is not important, it

does not matter how they are mixed just as long as they are mixed well.  Using random sampling has been

around for a long time and that's because it works.  Most assembly lines for mass produced products count on

it for quality control. 

RL

Hello rl perfect, then to see the pairs of points   The pairs are (the 1st and 5th)   Limit of the 1st 01-14 and 5th in the limit of (29-39) it is possible 300 pairs   Unclear whether the pair hit the first number to be within the limits 01-14 and 5th number in the limit from 29 to 39 = 02 37 the pair example, if points are the pairs of the suit is between the pair   02xxx37 = xxx (central suit), it is very clear that filter, the three frequencies, larger, medium and cold, etc. .. let a former emplo lottery in its 39/5, a couple of tips that late 5:33, then goes inside the suit 05 xxx33, or if the suit in its lottery central late 12,19,34 is clear that the 1st position is 12 down and the 5th position is 34 up, rl rondo, is often hit the track or suit means

I may be close to developing something like an algorithm but it's not as simple as A x B x C and it can involve up to 75%or more of the numbers in the regular lottery pool and it has to be recalculated after each drawing. 

The real advantage is in knowing the numbers and the range assigned to each position of the combinations.  The winning combination will usually be in a group generated using these parameters which can be as large as one to two thousands combinations but picking a few combinations from this group will have better odds of a win than the same numbers of combinations from the possible combinations in general since the combinations in this group will more often have some of the winning numbers than those generated randomly.

If successful, I should see some improvements in my prediction statistics.

I may be closer than you RJoh...

I have a question for you...on your Matrix Walk Chart for Rolling Cash 5/39...do the numbers go from left to right or right to left?

I am sure you are aware of all the same parameters that I am aware of because the Matrix is equivalent.

And i strongly doubt your spreadsheets are any less sophisticated than mine.  Although  i dont know how Basic appears on screen but i guess it has all the same bells and whistles as excel...

so which way does it go?

yes the points are key...and that is why i use fib integers represented as skips to filter my set.

One hypnotises I came up with was Pick 6 draws tend to bias to the outer individual position distributions over time

Analysis showed this to be true on past history on a large number of draws I tested.

So are you saying i would do better to observe distribution pattern movement better if i was put under hypnosis?

You could be right, however not being familiar with Matrix Walk Charts and not using spread sheets I can't respond to your statements and questions properly.

I'm writing all the routines I'm using just for my own use so there's no need for bells and whistles if they can't contribute to the results I'm hoping for.

I'll be testing those results by posting them on the prediction board so anyone can check them there if they wish.

Hello, RL, rondo, if I share your lottery 39/5 in 4 parts or 4 sectors   In 100% of the sector will draw a zero or a number and another sector will have at least   Two numbers (of course there could be more) in 100% 100% this standard is to use to filter? Sure you do not know which sector but the two features mentioned above are at 100% of the draws so do not care how this filter is always losing, you can play with three groups, but because in all three groups have 4-5 at 100%. this would be a great leap basic filter

my original intention in this comment was to say spreads.

I googled algorithm and found that there are listed ordered kinds of algorithms.
There is a page full of those on Wiki. I don't have time now to read all of that.
It might be smarter to start with probability and sequences, meanwhile analyse
the drawings.

My first question is, what kind of algorithms are there? Are there a few main
groups that can be covered? What are that main ways to crunch numbers,
leaving no gaps? It probably takes a motivated dr. prof. in mathematics to
summarize all this.

A good model for learning is roulette, and I am not the only one thinking so.
There are roulette based programs out there. Odd and even are possible bets
in roulette. Number crunching can be learned by playing it.
A program that picks numbers only executes a code, with only one algorithm,
you couldn't possibly say that it thinks or uses a selective system.

"You have to play something!"- the roulette player
"There are no professional roulette players!" - the professional
"You can't win at roulette, unless you steal from it!" - Albern Einstein
But, if you can't win at roulette, how could you possibly win in the lottery?