a "tested version" for pick 3Previous TopicNext Topic

Dear All, i found a tested version for pick 3

here are the download link: http://home.tiscali.nl/lsystems/Pick3.html

data structure in lot.txt shown

5
2000

471
...

it mean

NO    : descending
D1    : start from row 2003
D2    : start from row 4003
D3     : start from row 6003
DATE : start from row 8003

to make your own data use excel or other spreadsheet software.

i hope this tested version increase your winning change more frequently.

regards
sysp34

If you wish to know more about the program: Review the following thread:

http://www.lotterypost.com/thread/122022

Another Lottery Program (LS2004) can be found on the same site at:

http://home.tiscali.nl/lsystems/

Also see the following thread:

http://www.lotterypost.com/thread/117341

Dear All Pick3 hobbiest

i apologize to re-post that program again.

just review that the pattern filter & overall/statistics work as design. and to make more frequently win all you had to do is make wheeling system from 1 Digit Filter: just pick number that not appears from last 5 draw will increase winning.

GREAT JOB stoopendaal

regards
sysp34

Yes, Harmen (Stoopendaal) did a great job with it, not exactly as I wanted, but a great job anyway, it was sort of a joint effort, but he did all the programming, since I can't, but did give him plenty of filters ideas, he also implemented his ideas and some of Pick 4 Master's.

It was a triple effort, kind of, he being the only programmer. Too bad that he is a very very busy person and farther development is in check. Kind of like LottoSync.

Hi Lantern

  I have been using the Pick 3 program in Illinois eve .

the two digit filter I set at 27 for both the left and right pairs for elination.

Now I will try the digit by postion set at 5 along with sums in range of from 8 to 19 should be about 20 %

success rate if my first look is correct.

BigJer
 

How many draws do I add to get  a good results?

BigJer

2 digit filters are of course Pair(s) filters, The program is like that, because I wanted a "Graphical Interface" or GUI for the 1 and the 2 digit filters, as they would be easier to understand by more people this way than if they were such as they are on the Crunch 2, which most or a lot of people have trouble understanding and using the filters there.

I wanted also to be able to "Pre-Select and or Program the 2 Digit filters such as the 1 digit filters are, but Harmen didn't have the time to properly work on that part of those filters, so that particular feature of the 2 digit filters was "Dropped", so now they have to be manually selected, it takes very much time this way, but can be done and the program is more flexible in this way.

There should had been a feature or option on the filters to "Include Selected Filter Patterns and or pairs" and also to "Exclude Selected Filter Patterns and or pairs" for all the included filters, not just the 2 digit filters. But what is done is done and it would had taken very much more work for Harmen to do and made the program a little more complicated and Harmen has very very little free time as it is.

Yes sums from 8 to 19 or 8 to 20 are fine, sums' statistics for your state should tell what is best.

In reallity a good stats program might show that most, but not all of the 8 to 20 sums should be selected and also a few of the from 0 to 7 and or 21 to 27 should also be selected, but not all of them either.

As to pairs, a good pairs' stats program that would also include SKIPS and Pairs'  Appearances should tell which pairs to include and or which to exclude, both by position and also Boxed by any-position.

Also the last 50 or so past draws will show some pairs that you might be able to exclude, mostly the last 1 to 10 draws, also carefully check Pairs' Patterns (High-Low, Even-Odd and whatever else) for the last few draws.

Good luck.

Tim

How many draws should you add to the program's database ?

It all depends on your filters' technique, anything from 60 to 1,500 would be O.K.

For any order or boxed filtering from 40 to 400 past draws might be enough for most boxed filtering techniques, 60 would be just about right.

For straight numbers filtering from 60 to about 1,500, 1500 would be just about right for "Deep" filtering and for light filtering 60 past draws would be about just right.

Good luck.

By the way, Boxed Filtering can and should be used along with straight filtering as Box filters are also straight filters, but "Only Straight" filters are only for straight filtering of course.

Here might be some info on possible stats.

If you want to sell your program, maybe you should post it's stats results, here at the lottery post, maybe mostly on the pick 3 forum for people to see and use, if they see that it is really usefull to them and might help them get the winning number then they might buy it, it is a little big in size, if it can't be somehow compressed a lot more, them maybe it can be broken down into a few portions and after the downloads put together, there are many programs that can do this, but for many people this would not be a problem if they have a high speed connection, you can also submit the program to download sites from where it can then be distributed out.

For deep filtering of combos for great and accurate reduction good stats are a most have, but the right stats and in the right form are best, it all depends on what the user needs for his particular system(s) and or method(s), but any good stats in any form(s) can be used.

More stats are better tban fewer, a person can always overlook those that he does not need.

The more accurate and complete that the stats are and in a use-able form, the more accurate that a prediction can be made.

You might want to include your site on your user's profile and maybe also on your signature.

Maybe just on your users profile and should be on the bottom of your posts as your site.

Good luck.

Paurths (But also Stoopendaal for his program). 

I took a better look at the screens of the stats program and the program look real good, the skips could use, the following info:

Average skips for a particular section (Selectable) of past draws, anywhere during the history of the draws.

50% average skips info, The range of skips where 50% of the draws fall in, continuos range and also sectioned partitions and  50% "Individual Skip #" (The particular skip #s that make 50% of the draws (Average)).

Also "NOT" skips, but a number of particular pattern's hits during a given part or section (selectable) of the past draws' history.

Also average hits per week and biweekly and monthly of particular patterns for a given section (Selectable range of past draws) of past draws such as for the last 2 weeks, 1 month and  3 months and 6 months and 1 year.

Those are just some ideas for deeper stats.

I didn't remember to say this, but "Serious" daily games players, should have and use PAIRS info also, just as your program has 3 digits numbers (Whole pick 3 combos) info, a pairs' expert could greatly increase his or her chances of winning with the proper pairs' info, anybody could to some extent.

Detailed pairs info both straight and boxed can help anybody a lot, both appearances and skips, in any form of stats, but better in proper stats' form(s).

As to filters, a person should be able to filter out a given number of past draws from a source of combos, straight draws and boxed draws individually.

The stats program should also provide the past draws's history in boxed form, besides straight, just as if  794 came out it should provide both like this:

794, 479. either with-out spaces or with them or with - or , or whatever.

Also a person should be able to filter out a number of pairs and also range of past draws' pairs, both by position (straight) and boxed by any position separatelly or individually, also any number of any individual pairs that a person would input into the program and also any pick 3 combos a person puts in or pastes into the program it should be able to filter out from the source of combos.