Lottery Lotto Gambling Software Systems

[brianpa]

I have a situation that I'm doing something wrong and can't figure out where I'm messing up. Hopefully someone will read this and put me on the correct path.

Here is some history:
I have an ST5 file with just one filter set (Pot_1 = 4-5). We'll call it ST5.1

I also have an ST5 file with many filter settings (but no Ion3 filters). This file also has a filter setting of Pot_1= 4-5 along with the others. We'll call this one ST5.2

Here is the problem:
I will run ST5.1 with the F6 lexico order to get combinations and get 52,975 total combinations. That is with inner filters disabled. Now I will run the F7 Random order and use the ST5.2 file and select the purge option to purge the file created by the ST5.1 lexico step. I have 13 combinations in the new output file. Again the inners are disabled.
Here is where I get perplexed. If I just run the ST5.2 file with the F6 (lexico order) I get 4000 combinations.
I would think that I should still get the 13 combos. Both ST5 files have the same Pot_1 filter setting. The only difference is one has other settings as well. And as previously stated, no Ion3 filters were used and the inner filters were always disabled.

Why is this important to me??
My strategies are from 3 D5 files. I first start with the F6 option then continue using the F7 option with the purging feature to reduce the combos further.
I have different ST5 files which I use with different D5 files.

I know I have all the correct values in the ST5 files. I have checked and re-checked all possibilities of errors on my part. I just dont get it!!! Maybe I'm using the purging feature incorrectly, or maybe I don't understand what the purging feature really does.

THANKS for taking the time!
BrianPA

[brianpa]

I have found the problem!! When using the purge file option within the F7 command, the Pot_1 min. setting is throwing things off.

Here is what I have found:
I use 3 strategies and each has its own corresponding D5 file. When I run each one on its own using the F6(lexico option) there is no problem, the winning ticket is in there.
When I use the first output file from the F6 option and purge that output file with one of the other st5 files, I loose the winning ticket. So I have edited the other st5 files to see which filter is eliminating the winning ticket. I have found that the minimum setting of the Pot_1 filter is the culprit!! Here is something strange though, the winning filter value is 11. Now, when I add a 1 there it still produces the winning ticket. Even when I go up to filter value = 5, the winning ticket is still there. I loose the ticket when I set the min value to 6. So I'm not really sure what causes this, or if it can also be related to the other pot filters. That is a shame cause I use the Pot filter as the MAIN starting point of all my strategies.
You see I had to change all my existing strategies cause the PA Cash5 game has changed from a 5/39 to a 5/43. Not a problem anymore cause I have adjusted and came up with different strategies. Only all my strategies are based off of the Pot_1 filter......UGH!!!

Marcher I hope you have the time to read this and see if it can be corrected. I know you are busy, but if its something easy, it would be greatly appreciated.

THANKS
BrianPA

[marcher]

Matter of fact, this is the worst time for anybody to ask me questions. I am constantly moving between two places. I am still to complete the moving. My computer network is in a state of mess. Keep it koole for a while, will you?

The ‘Purge’ in randomized or wheeling is not recommended. Purge works best in lexicographic.

I don’t know about the filters. It is complicated when working with multiple D? files. Gotta check the same strategy one D? file at a time. Then see if the strategies hit in the same drawings.

That’s why they call them strategies: It is the decision of a human being, or a group of humans who share a contractual interest.

Best of luck!

Ion Saliu

[brianpa]

Thanks for the reply in such a busy time for you. I will try to figure out how to purge in lexico order. Its not an option within Bright5's F6 command. Its probably in another piece of software though. I'll do my searching. Didn't really mean to bother you, I thought that some of the experienced Purgers out there would have lead me in the right direction. I'll find it, I don't intend to be spoon feed just maybe a heading in the right path.

GOOD LUCK with the moving Marcher!!!!
Best Wishes,
Brian

[marcher]

Sorry! My bad!

I confused purge for ranges! Purge is incorporated in randomized (F7), not lexico (F6). Purge, however, works not like randomized, but more like lexico. Purge takes each and every line in an output file and checks it against the filters.

Randomized with very tight filters might not find every possible combination in a while. It is better to get every possible combo in lexico, then randomize (shuffle) the output file. Select combinations from the FFG median zone (around the mid point is OK).

By the way, Brian. All your D5 files should show the same Pot_1, since your Data-5 is atop various SIM-5 files. Otherwise, you might have a case of ‘Fixed’ reporting like in pick-3, but not supported in lotto.

I will be back (to normal) soon! I have to, I need to, and especially I want to!

Best of luck to all!

Parpaluck
PS
I detest doing serious work with laptops. They slow me down to a scream!
IS

[brianpa]

No Problem! You have so many different softwares that it must be a headache in itself to keep tabs on which does what. Let alone moving!! UGH!! Hope you had enough help with the moving.

You are correct though. Each of the ST5 files does have the same Pot_1 filter value. The only time I ran into different values was when I ran W5 and MD5 reports for longer than the lenght of my DATA-5 file.

Here is what I do:
Data-5 = real draws + 300 simulated draws (added sim draws cause not that many real draws yet). Thinking of adding more.
SIM5-01 = all possible combos shuffled
SIM5-02 = same as above except shuffled more
SIM5-03 = same shuffled even more

Combine them (DATA-5 + SIM5)for D5-01, D5-02, and D5-03

Each D5 file has its corresponding ST5 file.....ST5.101, ST5.201, and ST5.301

I ran each ST5 file separatly using the F6 (lexico option). The winning ticket was in each and every output file. Then what I tried was to purge the first output file 041008.101 (named for the date & ST5 file) with the ST5.201 using the appropriate D5-02 file. Not all matches were in the new output file, only some of the matches. Thats what puzzled me. Even though the winning ticket was in both lexico outputs, when I tried to purge, it didn't output all the matches. Now this may seem like a strange way to come up with combos to play, but I can have some very loose strategies and still reduce GREATLY the combos to play.

Marcher, there is no hurry and I don't mean to be a PITA!!! I just thought the more information you have about how I found this, and what I'm doing may be beneficial to you in finding the cause.

Hope all IS WELL!!!
Best Wishes!
Brian

P.S. I know what you mean about laptops. Was thinking about getting one, but when daughter got one for christmas.......WOW!! Took awhile to get used to it. Glad I got her desktop though, was a HUGE upgrade from my dinosaur desktop. HA! HA!

[brianpa]

Marcher,

Is it possible for you to take a look at the purge feature in the Bright5 software package? We discussed this sometime ago, but you were busy with your move.

Hope all is well in Gettysburg!!! Glad to see your still in PA!

Brian

[marcher]

Brian:

I was very busy, indeed. But I didn’t forget about your problem with the purge function in Bright5. That function uses other functions that are common in the entire program. I know I checked the lexico program with all the filters and the result was the winning combination. I checked the random program as well, but the testing is time consuming. It can take a long time for the program to generate the winning combination.

I realize now why I said that ‘purge’ was part of the lexicographic software. I’ll look again and might change the software. I discovered something else. I talked about a rounding feature in Visual Basic (VB). It happened in 2006 when we discovered all kinds of bugs in MDIEditor And Lotto WE. Some filters are dependent on the division operation. I discovered that time that VB rounded 39/2 to 20, but 49/2 to 24! The same thing happens in the PowerBasic compiler. It is called “banker’s rounding”.

I hate that feature: It is very easy to overlook it. The makers of PowerBasic strongly recommend using a data type named LONG (long integer in other compilers). The big problem is that LONG uses the banker’s rounding! I did change the operator in VB to FIX: Take only the integer part of the division, without any rounding. In the case above: 39/2=19; 49/2=24; 43/2=21, etc. I started to make the same change in my character-mode lottery software (compiled with PowerBasic).

Until I release the new versions, you can try one more test with ‘Purge’. You already have a file with all the combinations in the PA Cash-5 game. Make it the file to purge. See the filters in line #1 of your real data file. Delete that first line in Data-5 and save it as Data-5.2. Recreate the D5 file with Data-5.2 on top of Sim-5. You can enable all the filters first. If you don’t get the winning combination, try again disabling the filter you suspect creating the problem. I think I can also improve the performance of purging. The software should count first how many combinations the ‘file to purge’ has, and then load the lotto numbers in arrays.

Thanks in advance.

Ion Saliu

[brianpa]

Marcher,

I will do the testing on the purge as you explained. I will also run some other tests that I thought of and post back in great detail. I will take notes as I'm running the program so I can give you as much detail as possible.

As you stated. testing some things does take time, so please bare with me as I am also working 12 hrs. a day right now........UGH!!!

Please excuse my ignorance when you were talking about Visual Basic and Power Basic Compilier. I am not a programmer and don't have a clue about division operation or changing the operator to fix. GLAD you are there for us Ion!!!!!!

Cash5 tonight at $875K
Gotta get generating and will report back with testing results when completed.

THANKS!
Brian

[brianpa]

[marcher]

Brian:

Thank you so much for your thorough analysis. It helped me a lot. Sorry for that damaging error in Combine5.EXE. The problem is not present in Combine6.EXE.

The problem is with the POT filter, both the minimum and the max levels. First, POT was what I explained it to be. I did explain this filter a few years ago. I considered a “pot” to be a group of consecutive drawings. For example, the last 7 draws. If there are 25 unique numbers in the “pot”, then min_POT eliminates C(25, 5) = 53,130 combinations. Max_POT does the reverse: It forces the software to generate 53,130 combinations.

Supposed you enable both min and max; e.g. 6 and 7. If the “minimum pot” of 6 draws consist of 20 unique lotto numbers, then a number of C(20, 5) = 15,504 combinations are eliminated from the 53,130 “enforced” by the max level of POT.

There are situations when a tight POT like in this example will not generate any combo. There must be at least one more unique lotto number in the “maximum pot” than in the “minimum pot”. In a case like above, the most efficient setting for POT would be: 25 numbers in the “maximum pot” and 24 numbers in the “minimum pot”. C(25, 5) – C(24, 5) = 53,130 – 42,504 = 10,606 lotto-5 combos. Evidently, POT with low values in the W5 reports is the most effective; like 15 or 14 unique numbers in the 2 POTs.

That was then. In 2007, I added the MDI filters to my character mode lottery software. I realized that two of the filters were duplicates. Any1 was identical to ALL; Any5 (or ANY6 in lotto 6) was identical to POT. So, I changed ALL and POT to have only unique filters. Instead of checking equal numbers, ALL/POT in Bright check for ‘equal to Previous_Number+1’ and ‘equal to Previous_Number-1’. That’s how the report generator checks the filters (the F5 function key). The ALL and POT filters are incorporated correctly in Lexico5.EXE and the entire Bright6. Curiously, the ALL filter is correct in Combine5, but not POT!

What I said about min and max POT above is no longer valid in Bright. But it works exactly like that with ANY5 (or ANY6 in lotto-6; the last ANY in the pick lottery software). That answers also a strong question raised by Zulu. The new POT in Bright is more potent for the minimum level of the filter — it eliminates double the amount of combinations eliminated by the old POT (in Pick632.EXE); or double the amount of min_ANY5 (or min_ANY6 for lotto 6).

I am in process of correcting the error. I make some other improvements (performance and appearance). Also, I work seriously on the Least pairings feature. I know now exactly the mathematics of it. It is founded on ‘Ion Saliu’s paradox of N trials’. I’ll upload the software ASAP. I have much more free time now, but I plan a very long trip this summer.

Again, I am very sorry for that headache, even if I didn’t work at all with the ‘Purge’ in Bright5. Time is of the essence. And, again, thanks for your cooperation.

Best of luck!

Ion Saliu

[marcher]

By the way –

I discovered the same error in Wheel5.EXE, Combine3.EXE and Combine4.EXE. I’ll look more closely again, I’ll test more thoroughly the random as well. The problem with testing for random generation: The seed is very, very important. How fast the winning combination is generated depends very much on the randomization seed. Also, repetition is a natural law of random generation. You can get repeats of some combinations, while other qualifying combinations come out very late…

Ion Saliu

[brianpa]

Marcher,

Your Welcome! I'm just glad I could finally contribute something. Although, I wish my contribution was for something else other than finding an error. Its nice to feel like I helped the cause, instead of just being here reading and asking questions. Feels like I'm part of something now. WISH my contribution was winning the jackpot so I could help out by better funding your effort.

Ion, if you ever need testing done on a part of your software, just drop me a line..........I'm always willing to help out!!!!

Brian

P.S. Is there still the issue with the Ion3 filter. I believe it had sonething to do with the max setting with random generation. (could be wrong on that, just remember an issue posted with max Ion3)

[marcher]

[brianpa]

[Advertisement]