System BacktestingPrev Topic Next Topic

New Topic New Poll

Mr-B 216

Ohio
United States
Member #179,973
February 15, 2017
1,333 Posts
Offline

Jul 21, 2022, 11:11 am

I thought I'd share some backtesting results here that some of you may find useful. I myself don't play using any of these systems and I have no bias for or against any of these systems. This just some informative FYI.

These results are from the lakerben Gem system in the state of NM. The results really don't help much for choosing which HOD (see below) to play. The purpose of this is to test various systems to see which ones have the best potential in any given state. What works well in one state may not work well in another.

The system used here produced on average 8 numbers to play per draw. I realize there are up to 504 numbers to play based on the numbers you can create from the patterns available in a 3 x 3 TTT block. I think you'll agree that is not a realistic number to play with, unless you do some filtering. I chose to use the 8 basic TTT patterns, left to right, top to bottom, and diagonal.

Here are the column headers:

HOD = Hits On Draw or how many draws until one our numbers hits.
Cost = Money spent, based on $1 per number played, per draw.
Str = Number of Str8 hits
Str Win = Money won on $1 Str8 wins
Str P/L = Profit or Loss on Str8 plays
Str ROI = ROI for Str8 plays
Box = Number of Box hits
Box Win = Money won on $1 Box wins
Box P/L = Profit or Loss on Box plays
Box ROI = ROI for Box plays

The following results are from lakerben Gem and lakerben Cash Cow systems in NM, OH, PA, and VA. These results are also based on combined play, D>E>D>E....

As you see, Gem is working well in NM, and not too bad in PA. Tis mimics playing your numbers $1 straight on HOD #1, which means playing the system provided numbers for just 1 day following the previous draw used in the system.

What goes around comes around.
lakerben

50

New Mexico
United States
Member #86,094
January 29, 2010
30,561 Posts
Online

Jul 21, 2022, 11:45 am

Nm

221 Last week. The numbers in Nm tend to repeat in a week. I always go back in see what numbers are tied to others. I dont usually use the last draw data for the next. I save it for coming draws depending on the trend I see. So your analysis could be misleading to say the least. Your programming is excellent but the analysis is skewed.

Cash cow.

312

3. 8. 6

1. 6. 8

2. 7. 7

723 on tuesday.

923. Last night.

ONE OF MY OTHER SYSTEMS MIRROR ADJACENT. I analyze using several methods and any lotto player can do the same.

221. LAST WEEK

713

173

026

Alas.

723 on tuesday.

Win!
Mr-B 216

Ohio
United States
Member #179,973
February 15, 2017
1,333 Posts
Offline

Jul 21, 2022, 12:05 pm

Quote: Originally posted by lakerben on Jul 21, 2022
Nm

221 Last week. The numbers in Nm tend to repeat in a week. I always go back in see what numbers are tied to others. I dont usually use the last draw data for the next. I save it for coming draws depending on the trend I see. So your analysis could be misleading to say the least. Your programming is excellent but the analysis is skewed.

Cash cow.

312

3. 8. 6

1. 6. 8

2. 7. 7

723 on tuesday.

923. Last night.

ONE OF MY OTHER SYSTEMS MIRROR ADJACENT. I analyze using several methods and any lotto player can do the same.

221. LAST WEEK

713

173

026

Alas.

723 on tuesday.

I take the numbers produced from the system as you explain it in your first post when you introduce the system. I then test play the numbers produced by you system and test play those numbers for the next 30 draws to see which draw, if any, produces the winner. I don't see where I am misleading with my info.

What goes around comes around.
lakerben

50

New Mexico
United States
Member #86,094
January 29, 2010
30,561 Posts
Online

Jul 21, 2022, 12:19 pm

Quote: Originally posted by Mr-B 216 on Jul 21, 2022
I take the numbers produced from the system as you explain it in your first post when you introduce the system. I then test play the numbers produced by you system and test play those numbers for the next 30 draws to see which draw, if any, produces the winner. I don't see where I am misleading with my info.

Nm

944. July 12th

035

0. 5. 9

3. 8. 6

5. 0. 4

083. Yesterday.

When I see pairs in cash cow repeat from one analysis I go back a week and use them like yesterday. Of course there have been alot of 5's hitting 150.591 etc. Which leads to the 38 pair.

Win!
Mr-B 216

Ohio
United States
Member #179,973
February 15, 2017
1,333 Posts
Offline

Jul 21, 2022, 12:27 pm

My program does show the wins you speak of, they just didn't make the final cut in THIS test, which was only looking at HOD 1 and the main 8 TTT patterns. You see 237 hit as a box on HOD 12 and 923 (TTT pattern 378) is shown as being hit straight as well. You are only seeing a small portion of what my backtest produced. I obviously can't show every little bit of data that pops up in the results. When I backtest, I look for certain patterns or anomalies and go forward from there.

I'm sorry if I rained on your parade. That is not my intention. When you introduce a system, you seldom go on to explain how you use it, so I make the best of what info I have at hand.

What goes around comes around.
Mr-B 216

Ohio
United States
Member #179,973
February 15, 2017
1,333 Posts
Offline

Jul 21, 2022, 12:47 pm

Quote: Originally posted by lakerben on Jul 21, 2022
Nm

944. July 12th

035

0. 5. 9

3. 8. 6

5. 0. 4

083. Yesterday.

When I see pairs in cash cow repeat from one analysis I go back a week and use them like yesterday. Of course there have been alot of 5's hitting 150.591 etc. Which leads to the 38 pair.

Got that one as well, it also did not meet the criteria in THIS test. Just like before, it was found in the TTT pattern hits.

It's all in the data, it just depends on what data we choose to look at in our backtesting. I choose to look at the data that has the best potential for producing winners.The data I looked at in the first test was just a starting point.

What goes around comes around.
lakerben

50

New Mexico
United States
Member #86,094
January 29, 2010
30,561 Posts
Online

Jul 21, 2022, 12:58 pm

I'm at the tail end of covid so you got me rolling!

From the Lakerben toolbox something I use.

719 date

Root 8

Mirror of root 3.

Next day look for the 83 pair in any system!

083 yesterday.

Win!
lakerben

50

New Mexico
United States
Member #86,094
January 29, 2010
30,561 Posts
Online

Jul 21, 2022, 6:08 pm

Nm

6746

7655

7 2 2

6 1 3

5 0 4

5 0 4

Todays result 5730.

Bam!!!

Confucius says backtest that!

Win!
Mr-B 216

Ohio
United States
Member #179,973
February 15, 2017
1,333 Posts
Offline

Jul 21, 2022, 8:04 pm

Quote: Originally posted by lakerben on Jul 21, 2022
Nm

6746

7655

7 2 2

6 1 3

5 0 4

5 0 4

Todays result 5730.

Bam!!!

Confucius says backtest that!

I'm not sure why you are taking such a negative attitude towards me in regards to backtesting. I was under the impression, from some of your posts in the past, that you were a proponent of backtesting.

My intention is to help some others here understand the potential value in backtesting various systems to see which one, if any, will work best for them. I saw a potential in your Gem system working well for those of you in NM, but not so much for those in my state of OH. That is not to say your system is useless in OH, it just means more studying of the data produced in backtests needs to be done. If one wants to use a system, then does it not make sense to use it in a way that yields the best results?

I start this thread, and after one post from me, you decide what I have to offer is a waste of time. That is fine with me as we are all entitled to our opinions. I just feel you formed your opinion a bit prematurely.

I try to be helpful while I'm here. You know I have "Exceled it" for others in the past, as well as released a few free programs.

As for your "Confucius says backtest that!" comment, my response is this... do you play for straight wins or box wins? Do you know how many possible pattern choices there are in that TTT block to choose from for a straight hit, which is how I play. I make better choices by studying my data, just as I'm sure you do. Would you agree that those here who may not be familiar with backtesting might benefit from learning how others gather their data for making better choices when playing the lottery?

What goes around comes around.
AllenB

California
United States
Member #167,120
June 27, 2015
1,777 Posts
Online

Jul 21, 2022, 9:53 pm

Quote: Originally posted by Mr-B 216 on Jul 21, 2022
I'm not sure why you are taking such a negative attitude towards me in regards to backtesting. I was under the impression, from some of your posts in the past, that you were a proponent of backtesting.

My intention is to help some others here understand the potential value in backtesting various systems to see which one, if any, will work best for them. I saw a potential in your Gem system working well for those of you in NM, but not so much for those in my state of OH. That is not to say your system is useless in OH, it just means more studying of the data produced in backtests needs to be done. If one wants to use a system, then does it not make sense to use it in a way that yields the best results?

I start this thread, and after one post from me, you decide what I have to offer is a waste of time. That is fine with me as we are all entitled to our opinions. I just feel you formed your opinion a bit prematurely.

I try to be helpful while I'm here. You know I have "Exceled it" for others in the past, as well as released a few free programs.

As for your "Confucius says backtest that!" comment, my response is this... do you play for straight wins or box wins? Do you know how many possible pattern choices there are in that TTT block to choose from for a straight hit, which is how I play. I make better choices by studying my data, just as I'm sure you do. Would you agree that those here who may not be familiar with backtesting might benefit from learning how others gather their data for making better choices when playing the lottery?

A Few years back I used a Pick 3 File that “Back Tested” several different “Systems” that were active at the Time. I called the File “Off Spring” and it was a breakdown of the components of each System for each Draw Thru a Specified Time Period. Once the formulas of a system were created, they were placed in a Que where they could be recalled just by changing the “System” Name. I think it was doing the same thing that you show.

I also ran a few of lakeben’s “Systems”. It was apparent then that he had a thin skin and would react in the ways used on your Topic. I know you don’t care about the stars but he does. So much so that he gives himself 5 Stars on every one of his topics.

I would like to make a comment about the difference between a Back Test and a Query. No disrespect to lakerben’s many, many “Systems” but one does not really Back Test them. We fill a historical data table and then Query that data to determine if the “System” was successful.

I used to think that in order to do a Back Test, you had to run a Macro in Excel. I found out that just because you run a Macro does not mean that you are running a “Back Test’. Most of the time it was creating more historical data, just on a game by game basis and then logging said data for additional Queries.

Lakerben had a few “Buts” when his System was presented. “I usually…. Do something else…..When”. Practically admitting that the System needs help once in a while. Now if he could explain why he does what he does when he does it, that would be something to “Back Test”
Mr-B 216

Ohio
United States
Member #179,973
February 15, 2017
1,333 Posts
Offline

Jul 22, 2022, 6:16 am

Quote: Originally posted by AllenB on Jul 21, 2022
A Few years back I used a Pick 3 File that “Back Tested” several different “Systems” that were active at the Time. I called the File “Off Spring” and it was a breakdown of the components of each System for each Draw Thru a Specified Time Period. Once the formulas of a system were created, they were placed in a Que where they could be recalled just by changing the “System” Name. I think it was doing the same thing that you show.

I also ran a few of lakeben’s “Systems”. It was apparent then that he had a thin skin and would react in the ways used on your Topic. I know you don’t care about the stars but he does. So much so that he gives himself 5 Stars on every one of his topics.

I would like to make a comment about the difference between a Back Test and a Query. No disrespect to lakerben’s many, many “Systems” but one does not really Back Test them. We fill a historical data table and then Query that data to determine if the “System” was successful.

I used to think that in order to do a Back Test, you had to run a Macro in Excel. I found out that just because you run a Macro does not mean that you are running a “Back Test’. Most of the time it was creating more historical data, just on a game by game basis and then logging said data for additional Queries.

Lakerben had a few “Buts” when his System was presented. “I usually…. Do something else…..When”. Practically admitting that the System needs help once in a while. Now if he could explain why he does what he does when he does it, that would be something to “Back Test”

My long term goal here was to help others decide which system might work best for them in their state. Since lakerben had a large number of systems, I decided to start this venture by looking at a few of his systems. I agree with you Allen, I knew going in that lakerben could be a bit sensitive about his systems, but I never thought he'd see my trying to help as a negative bash against him and his systems. Well needless to say, I'll take his systems off the table and look elsewhere.

Oh well, enough said. I'll continue my journey and post back if and when I find anything interesting.

What goes around comes around.
lottoswe

Sweden
Member #163,014
January 17, 2015
158 Posts
Offline

Jul 22, 2022, 8:34 am

Quote: Originally posted by Mr-B 216 on Jul 22, 2022
My long term goal here was to help others decide which system might work best for them in their state. Since lakerben had a large number of systems, I decided to start this venture by looking at a few of his systems. I agree with you Allen, I knew going in that lakerben could be a bit sensitive about his systems, but I never thought he'd see my trying to help as a negative bash against him and his systems. Well needless to say, I'll take his systems off the table and look elsewhere.

Oh well, enough said. I'll continue my journey and post back if and when I find anything interesting.

Hi,
i have one system, if you can do backtest. I know it work good but i would like to see it on bigger sample.

First, numbers are in 3 groups :

1 - 1 4 7
2 - 2 5 8
3 - 3 6 9 0

You check only groups.

Here is system rules:

- if you have 2 groups in last 4 draws then next play those two groups ( 1 2 2 1 - then play 1 and 2)
- if you have 3 groups in last 4 draws then next plat two groups that don't repeat ( 1 2 3 2 - then play 1 and 3)
- if you have 1 group in last 4 draws then next play other two groups ( 1 1 1 1 - then play 2 and 3)

I don't know how you will do this but maybe help to convert real draws into groups, then do test.

If you have questions about this i'll be glad to help.
Mr-B 216

Ohio
United States
Member #179,973
February 15, 2017
1,333 Posts
Offline

Jul 22, 2022, 9:19 am

Quote: Originally posted by lottoswe on Jul 22, 2022
Hi,
i have one system, if you can do backtest. I know it work good but i would like to see it on bigger sample.

First, numbers are in 3 groups :

1 - 1 4 7
2 - 2 5 8
3 - 3 6 9 0

You check only groups.

Here is system rules:

- if you have 2 groups in last 4 draws then next play those two groups ( 1 2 2 1 - then play 1 and 2)
- if you have 3 groups in last 4 draws then next plat two groups that don't repeat ( 1 2 3 2 - then play 1 and 3)
- if you have 1 group in last 4 draws then next play other two groups ( 1 1 1 1 - then play 2 and 3)

I don't know how you will do this but maybe help to convert real draws into groups, then do test.

If you have questions about this i'll be glad to help.

To make sure we are on the same page, how would you apply your system to these last 4 drawings:

082
536
765
640

What goes around comes around.
lottoswe

Sweden
Member #163,014
January 17, 2015
158 Posts
Offline

Jul 22, 2022, 10:09 am

Quote: Originally posted by Mr-B 216 on Jul 22, 2022
To make sure we are on the same page, how would you apply your system to these last 4 drawings:

082
536
765
640

322 (082)
233 (536)
132 (765)
313 (640)

pos 1 - 3 1 2 3 - play 1 2 (numbers to play 1 4 7 2 5 8)
pos 2 - 1 3 3 2 - play 1 2 (numbers to play 1 4 7 2 5 8)
pos 3 - 3 2 3 2 - play 2 3 (numbers to play 2 5 8 3 6 9 0)
lakerben

50

New Mexico
United States
Member #86,094
January 29, 2010
30,561 Posts
Online

Jul 22, 2022, 11:01 am

Quote: Originally posted by Mr-B 216 on Jul 21, 2022
I thought I'd share some backtesting results here that some of you may find useful. I myself don't play using any of these systems and I have no bias for or against any of these systems. This just some informative FYI.

These results are from the lakerben Gem system in the state of NM. The results really don't help much for choosing which HOD (see below) to play. The purpose of this is to test various systems to see which ones have the best potential in any given state. What works well in one state may not work well in another.

The system used here produced on average 8 numbers to play per draw. I realize there are up to 504 numbers to play based on the numbers you can create from the patterns available in a 3 x 3 TTT block. I think you'll agree that is not a realistic number to play with, unless you do some filtering. I chose to use the 8 basic TTT patterns, left to right, top to bottom, and diagonal.

Here are the column headers:

HOD = Hits On Draw or how many draws until one our numbers hits.
Cost = Money spent, based on $1 per number played, per draw.
Str = Number of Str8 hits
Str Win = Money won on $1 Str8 wins
Str P/L = Profit or Loss on Str8 plays
Str ROI = ROI for Str8 plays
Box = Number of Box hits
Box Win = Money won on $1 Box wins
Box P/L = Profit or Loss on Box plays
Box ROI = ROI for Box plays

The following results are from lakerben Gem and lakerben Cash Cow systems in NM, OH, PA, and VA. These results are also based on combined play, D>E>D>E....

As you see, Gem is working well in NM, and not too bad in PA. Tis mimics playing your numbers $1 straight on HOD #1, which means playing the system provided numbers for just 1 day following the previous draw used in the system.

The hits keep coming with the cash cow:

Nm

375

466

4. 9. 5

6. 1. 3

6. 1. 3

359 lasrt night.

Hmmm another hit??!!

Hmmmmm?!

Win!

New Topic New Poll

Subscribe to this topic

System BacktestingPrev TopicNext Topic

System BacktestingPrev Topic Next Topic