Numbers Order in Column

Postby lottoboy » Sat Nov 28, 2009 7:24 am

[quote=Joe]
i check this, thanks for your screen-shots,
[/quote]

Hi, Joe:

Thanks for your sincere reply!!!

I found some status that have confirmed the instability of Number Order Filter_5xx in my Back-tests today. For example, some strange (or incorrect ) results can be eliminated when I click “OKâ€￾ to make a running, then, turned back original setting. But the strange status renewed again in another Column. Please check it. Maybe , there are some bugs in the program.

In addition, I have sent the Error report to you by E-mill. Maybe it's usefull to you. Thanks.

Best,
lb
lottoboy
 
Posts: 738
Joined: Sat Dec 08, 2007 8:37 pm

Postby lottoboy » Mon Nov 30, 2009 5:26 am

Hi, Joe:

How about your checking about the Filter?
I'm looking forward to your result for further Back-tests next.
Thanks.

Best,
lb
lottoboy
 
Posts: 738
Joined: Sat Dec 08, 2007 8:37 pm

Postby lottoboy » Sun Dec 06, 2009 12:27 am

Hi, Joe:

One week has gone now I have not yet gotten any information from you.
Could you give me a reply about your Filter? Thanks.

Best,
lb
lottoboy
 
Posts: 738
Joined: Sat Dec 08, 2007 8:37 pm

Postby Joe » Sun Dec 06, 2009 8:44 am

i couldnt have time to work on filters, hope will be dont by next week-end, sorry for the delay
Joe
 
Posts: 1227
Joined: Mon Sep 18, 2006 6:51 pm
Location: Petaling Jaya, Malaysia

Postby lottoboy » Sun Dec 06, 2009 9:31 am

[quote=Joe]
i couldnt have time to work on filters, hope will be dont by next week-end, sorry for the delay
[/quote]

Hi, Joe:

Thanks for your sincere reply!!
No problem, I can wait until you have free time since your job is first.

Also, I find that the #s order data makes confusion in the Column bar by my Back-tests. For example, the 7 in Column bar of the filter should be
the Col.-7 but its data is same with Col. -6 sometimes. Maybe, you need to change the Column bar like Number Order Filter_6xx from 0 to -10 instead of
the current design from 0 to 11 in Number Order Filter_5xx. Also, could you change the setting of Column Count of Filter_5xx like the Filter_6xx. If so,
we can select the compond settings of Column Count such 0-11 or 3-5, ...etc but now we can make the sigle column only. Thanks.


The below is another question about the Number Order Filter including both 5xx and 6xx:

Hi, Stan and Joe:

Could you add a function show about Distinct Order Count in Columns like the Distinct Diffs Count in the Show Bar of WN History into the current EL 4 now? If so, we can easy to find the latest Numbers Order in Col instead of checking those from 0 to -10 by hand separately. We would appreciate your good job for it.

Best,
lb
lottoboy
 
Posts: 738
Joined: Sat Dec 08, 2007 8:37 pm

Postby stan » Sun Dec 06, 2009 11:09 pm

[quote=lottoboy]
Could you add a function show about Distinct Order Count in Columns like the Distinct Diffs Count in the Show Bar of WN History into the current EL 4 now?
[/quote]

i'm not adding any new features to version 4.x but pls feel free to create a new feature request for version 5.
Expert Lotto Team
User avatar
stan
Site Admin
 
Posts: 6338
Joined: Thu Sep 23, 2004 1:01 pm

Postby lottoboy » Tue Dec 08, 2009 1:57 am

[quote=stan:1260133795]
[quote=lottoboy]
Could you add a function show about Distinct Order Count in Columns like the Distinct Diffs Count in the Show Bar of WN History into the current EL 4 now?
[/quote]

i'm not adding any new features to version 4.x but pls feel free to create a new feature request for version 5.
[/quote]

OK, I will. Thanks.
lottoboy
 
Posts: 738
Joined: Sat Dec 08, 2007 8:37 pm

Postby lottoboy » Wed Dec 09, 2009 4:09 am

[quote=MartinMan:1258648195]
Hello Joe,

I just checked again and the filter is working the way I described above,
andrzejs says it is working so maybe it is just my system. But it will not work. Thanks for your time. I'll do it manually.

Regards,
Martin
[/quote]

Hi, Martin:

Did you check the Filter again recently? How about its working status?
The best way should make an attachment of Picture for your problem with your post, I think.

best,
lb
lottoboy
 
Posts: 738
Joined: Sat Dec 08, 2007 8:37 pm

Postby lottoboy » Wed Dec 09, 2009 8:44 am

Hi, Stan, Joe and Jerzy, Martin …:

I have a question about Number Order Filters (5xx & 6xx) as below:

1) The statistical truth in the Filters

a) The sum of Previews for each row/column and each position in any Columns (0 ~ -10) equals to the full tickets or odds of the game exactly. For example, the full tickets are 575757 in 5/39.

b) The sum of Back-test for each row/column and each position in any Columns (0 ~ -10) equals to 100% exactly.

2) My question is if the sum of Previews for the 5 or 6 winning orders
in any Columns (0 ~ -10) should equal to the full tickets exactly theoretically too?
Also, the sum of Back-test for 5 or 6 winning orders in any Columns (0 ~ -10) should equal to 100% exactly theoretically too? Thanks.

Best,
lb
lottoboy
 
Posts: 738
Joined: Sat Dec 08, 2007 8:37 pm

Postby Jerzy » Wed Dec 09, 2009 2:38 pm

lottoboy,

I am sorry. I don't know what the sum of Previews and sum of Back-test is. Calculating such sums seems to be a complete waste of time. And what is "the 5 or 6 winning order"?

In any filter, the Preview function is used to see how many combination from the package would pass or fail with the filter settings of your choice. The package can contain all possible combinations or only some combinations if it was prefiltered. You just want to see how many tickets can be removed with this filter setting.

The Back-test function is used to see how many previous jackpots that are present in the data base and what % of them would pass or fail with the given filter settings. For example, if 90% of previous jackpots pass with the selected filter setting, the setting is relatively safe but probably it will remove a small or moderate number of tickets. If only 10% of previous jackpots pass with the filter setting, the filter will remove a large proportion of tickets but probably there is only 10 % chance that the next draw jackpot will not be removed from the package.
Jerzy
 
Posts: 811
Joined: Mon Sep 18, 2006 5:26 am

Postby lottoboy » Thu Dec 10, 2009 2:11 am

[quote=Jerzy]
lottoboy,

I am sorry. I don't know what the sum of Previews and sum of Back-test is. Calculating such sums seems to be a complete waste of time. And what is "the 5 or 6 winning order"?

In any filter, the Preview function is used to see how many combination from the package would pass or fail with the filter settings of your choice. The package can contain all possible combinations or only some combinations if it was prefiltered. You just want to see how many tickets can be removed with this filter setting.

The Back-test function is used to see how many previous jackpots that are present in the data base and what % of them would pass or fail with the given filter settings. For example, if 90% of previous jackpots pass with the selected filter setting, the setting is relatively safe but probably it will remove a small or moderate number of tickets. If only 10% of previous jackpots pass with the filter setting, the filter will remove a large proportion of tickets but probably there is only 10 % chance that the next draw jackpot will not be removed from the package.
[/quote]

Jerzy,

You don’t understand my means anymore. Please look at the data in my Back-tests for 5/39 game as below:

P1: 25/87607; 21/109633; 18/117370; 21/124015; 15/137132;
(sum of bt/sum of pre. = 100/575757)
P2: 20/100328; 19/116396; 17/110412; 17/109260; 27/139361;
(sum of bt/sum of pre. = 100/575757)
P3: 16/150417; 25/125862; 28/101074; 12/94544; 19/103860;
(sum of bt/sum of pre. = 100/575757)
P4: 17/153808; 14/125021; 17/103817; 29/96349; 23/96762;
(sum of bt/sum of pre. = 100/575757)
P5: 22/83597; 21/98845; 20/143084; 21/151589; 16/98642;
(sum of bt/sum of pre. = 100/575757)

Also, the (sum of bt/sum of pre. = 100/575757) is still working in the 5 columns above. The bolds are the winning orders in the Col.

Jerzy, please take care with those data above carefully and maybe you can find some interesting things.

Good luck,
lb
lottoboy
 
Posts: 738
Joined: Sat Dec 08, 2007 8:37 pm

Postby Jerzy » Thu Dec 10, 2009 5:17 am

lottoboy,

You run the Preview and Back-test functions for one position in one column. You use only one value for this position from 5 values possible (1, 2, 3, 4, 5). You use the full package of possible combinations.

Then you add 5 results of Preview testing and you see that all combinations have passed.

Next you add the percentages of 5 back-tests and you see that 100% of jacpots have passed.

I cannot see anything surprising in your calculations.

You can also run Preview for all five values in the position at the same time, or Back-test for all five values at the same time, no need to calculate sums manually and the result will be the same.
Jerzy
 
Posts: 811
Joined: Mon Sep 18, 2006 5:26 am

Postby MartinMan » Thu Dec 10, 2009 2:28 pm

Hello lottoboy,
I haven't checked it again. Since it wasn't working right I haven't used it.
Some have said that it is working, but it wasn't on my machine.
So I am sorry, but I can't really tell you.

Regards,
Martin
MartinMan
 
Posts: 141
Joined: Mon Oct 02, 2006 9:38 pm

Postby lottoboy » Thu Dec 10, 2009 9:16 pm

[quote=Jerzy]
lottoboy,

You run the Preview and Back-test functions for one position in one column. You use only one value for this position from 5 values possible (1, 2, 3, 4, 5). You use the full package of possible combinations.

Then you add 5 results of Preview testing and you see that all combinations have passed.

Next you add the percentages of 5 back-tests and you see that 100% of jacpots have passed.

I cannot see anything surprising in your calculations. But you proved that the Number Order in Columns works correctly.

You can also run Preview for all five values in the position at the same time, or Back-test for all five values at the same time, no need to calculate sums manually and the result will be the same.
[/quote]

Jerzy,

You are still not clear my Back-test for the Filter.

According to the statistical law, the sum of Previews for the 5 winning orders in any Columns (0 ~ -10) should equal to the full tickets exactly theoretically too. Also, the sum of Back-test for 5 winning orders in any Columns (0 ~ -10) should equal to 100% exactly theoretically too.

In addition, the values of BT should match the values of Preview proportionally in the Filter. In other words, the big value of BT should match big value of Previewin in any Positions and Columns. For example, the Pre value of the bt=29 should greater than the Pre value of the bt=17 (The situation about the low value is same) If not, or the differences between the theoretical values and the practical values for the Winning Orders are too big. It is abnormal or wrong.

We can find some useful strategies from the correct values of Preview and BT in the Winning Orders. Keep your eyes open please: ONLY ONE WINNING ORDER IS IN THE FIVE POSITIONS. But “run Preview for all five values in the position at the same time, or Back-test for all five values at the same timeâ€￾ is meaningless completely for the Winning Orders.

I do hope my explanation above is usefull to you or gives you some idea for fiding a new strategy by using the Filter.
lottoboy
 
Posts: 738
Joined: Sat Dec 08, 2007 8:37 pm

Postby Jerzy » Fri Dec 11, 2009 3:19 am

lottoboy,

I aggree with you about the significance of the Preview and Back-test. Thank you for discussing this. I feel you were trying to tell me more.

Of course, finding the exact winning numbers order is almost impossible, but we can find a way around this problem.

It would be good if the Preview function was able to display the percentage of the tickets that passed. The Back-test function can do it.

If you place in the package all the winning combinations from your game history, the results of Preview and Back-test should be the same. Unfortunately, in my experience with Numbers Order in Columns it doesn't happen.

For example:

I placed in the package 1275 combinations of my game history. I run Preview and Backtest for the column 0, position 1, value 1 in the position.

The results:

PREVIEW
passed: 426 (33.41% - calculated by myself)
failed: 849
total: 1275

BACK-TEST
passed: 301 (23.61% - calculated by the filter)
failed: 974
total: 1275

Other filters, like O/E filter or Sum filter give the same results for Preview and Backtest when the package contains all the winning combinations of the game history. So there must be something wrong with the Numbers Order in Columns filter.
Jerzy
 
Posts: 811
Joined: Mon Sep 18, 2006 5:26 am

PreviousNext

Return to Plugin Development

Who is online

Users browsing this forum: No registered users and 2 guests