CV Data Simulations - problem!

[matt21]

hi everyone, i am encountering a strange problem when I ran simulations using CV Data yesterday, and am wondering whether anyone else has encountered this. My gut feeling is that this relates to my defined multi-hand betting strategy (the simulations seem to work fine for single-hand betting stratgies).

I run the simulation and then look at the 'True Count distribution' table and/or the 'True Count Won/Lost' table. The data is broken down into rows by true count. This is all looking good except that there is a signficant amount of data for 'TC <- -39' - meaning situations where the True Count is -39 or lower. This must of course be incorrect. The bell-shaped true count distribution on my tables seem to range between -16 and +20. The tables are saying that 24.3 million out of the 230 million hands occurred at true count of -39 or lower. Additionally, this row of data contributes a good 25% of the 'Dollars/Hour Win rate'.

I am reluctant to use this data for decision-making because i don't know understand what has gone on here.

Can anyone shed light on this situation?

 

[SleightOfHand]

hi everyone, i am encountering a strange problem when I ran simulations using CV Data yesterday, and am wondering whether anyone else has encountered this. My gut feeling is that this relates to my defined multi-hand betting strategy (the simulations seem to work fine for single-hand betting stratgies).

I run the simulation and then look at the 'True Count distribution' table and/or the 'True Count Won/Lost' table. The data is broken down into rows by true count. This is all looking good except that there is a signficant amount of data for 'TC <- -39' - meaning situations where the True Count is -39 or lower. This must of course be incorrect. The bell-shaped true count distribution on my tables seem to range between -16 and +20. The tables are saying that 24.3 million out of the 230 million hands occurred at true count of -39 or lower. Additionally, this row of data contributes a good 25% of the 'Dollars/Hour Win rate'.

I am reluctant to use this data for decision-making because i don't know understand what has gone on here.

Can anyone shed light on this situation?

I'm assuming you mean running count, not true count, implying an unbalanced count system. While I have minimal knowledge of such systems, I can give you some of my "common sense" advice. With an unbalanced system like KO, you have an IRC of -24 in an 6D game. Therefore, you will experience counts lower than -24 50% of the time. If you are using a balanced system, then just disregard this lol

(This part is something that I just thought of with little to no proof, but seems to answer your question)
In a 6D game (IRC -24), a RC of -39 is the same (or at least similar) as a RC of -15 in a balanced system. And although this won't ever happen off the top of the shoe, an RC of -15 off the top is a TC of -2.5. Assuming TCs of -2.5 and less are as frequent as +2.5 and more, with some mental adding from the charts in CVCX for Hi-Lo, it appears that TCs of -2.5 and less occur about 10% of the time, which would correlate to your 24.3 / 230 million hands. Since Hi-Lo and KO are different counts, the RC frequencies are going to be different, but hey, the numbers are close lol.:grin::grin::grin:

PS: Can anyone verify my theory or tell me where I'm wrong?

 

[matt21]

hi sleight of hand!
no i do in fact mean TRUE COUNT rather than RUNNING COUNT. I use Hi-Low as my counting system.

The output from the simulation does actually make sense, except for the row classified as 'TC <- -39', i.e. all the other rows give reasonable data - for example, the % distributions for other true counts are 42.4% (for 0), 13.0% (for +1), 5.9% (for +2), 2.9% (for +3), 1.6% (for +4) etc.

it really seems to point to a software error.

 

[SleightOfHand]

hi sleight of hand!
no i do in fact mean TRUE COUNT rather than RUNNING COUNT. I use Hi-Low as my counting system.

The output from the simulation does actually make sense, except for the row classified as 'TC <- -39', i.e. all the other rows give reasonable data - for example, the % distributions for other true counts are 42.4% (for 0), 13.0% (for +1), 5.9% (for +2), 2.9% (for +3), 1.6% (for +4) etc.

it really seems to point to a software error.

Well then, I'm baffled (haha thats a funny word. "I was baffled when I saw a boogle of weasels jump out of a wagon and masticate my wenis")

 

[QFIT]

hi everyone, i am encountering a strange problem when I ran simulations using CV Data yesterday, and am wondering whether anyone else has encountered this. My gut feeling is that this relates to my defined multi-hand betting strategy (the simulations seem to work fine for single-hand betting stratgies).

I run the simulation and then look at the 'True Count distribution' table and/or the 'True Count Won/Lost' table. The data is broken down into rows by true count. This is all looking good except that there is a signficant amount of data for 'TC <- -39' - meaning situations where the True Count is -39 or lower. This must of course be incorrect. The bell-shaped true count distribution on my tables seem to range between -16 and +20. The tables are saying that 24.3 million out of the 230 million hands occurred at true count of -39 or lower. Additionally, this row of data contributes a good 25% of the 'Dollars/Hour Win rate'.

I am reluctant to use this data for decision-making because i don't know understand what has gone on here.

Can anyone shed light on this situation?

Are you using the MRI function? This excludes selected counts. The excluded counts are displayed in that row.

 

[jack.jackson]

Qfit, how do you pause the index generation??

 

[QFIT]

Qfit, how do you pause the index generation??

Hmm. Never thought of doing that. I'll put it on the list.

 

[jack.jackson]

Hmm. Never thought of doing that. I'll put it on the list.

Unless mistaken I found a temporary way to pause it.

The other day I went to run a Virus scan(I know you should'nt do this when generating indices) with the Index generator running. My Virus scan just froze and would not scan. By right clicking, on top of the Blue trim of CVdata(little window pops up) my Virus scan ran at full speed. My conclusion is, is that if I want to enable a program, I will first engage the program or software, then right click on CVdata. This seems to work as a temporarily solution. Or at least partially. Note, that if you click on ANYTHING else it re-enabl'es CVdata back to its full state.

 

[QFIT]

Unless mistaken I found a temporary way to pause it.

The other day I went to run a Virus scan(I know you should'nt do this when generating indices) with the Index generator running. My Virus scan just froze and would not scan. By right clicking, on top of the Blue trim of CVdata(little window pops up) my Virus scan ran at full speed. My conclusion is, is that if I want to enable a program, I will first engage the program or software, then right click on CVdata. This seems to work as a temporarily solution. Or at least partially. Note, that if you click on ANYTHING else it re-enabl'es CVdata back to its full state.

Interesting. That does work. You can also hit the Abort button. It will pause after a bit. Then hit Continue. I forgot that I added a Continue option on the Abort screen.

 

[jack.jackson]

Interesting. That does work. You can also hit the Abort button. It will pause after a bit. Then hit Continue. I forgot that I added a Continue option on the Abort screen.

I just verified it, and it definitely works for me. I also tried right-clicking CVdata from the TaskBar, but to no Avail. You must right-click on the Blue trim from the program. Sometimes it takes a few to several seconds to engage. Your Registry may become temporarily distorted. Once the little window, pop's up, is when the Red light on my computer starts running continously and the Virus Scan starts to run at full speed. This is how I first noticed. Out of curiousity, to eliminate the possibilty, that it was a windows feature I also right clicked on the Blue Trim of my scan to see if it would pause as well. It did not!

On a seperate note, what puzzles me, is I know both my Virus scan and CVdata uses alot of resources, and if you try to open to many windows or run to many applications at once your Registry will get distorted until Windows catches up or has to close.

I use the little Red light as a tool, for my PC, such as unknown process'es running in the background. And when I run a Virus scan, the Red light runs continously, but when im running CVdata, it just stays completely dimmed out.

What Resources or Process'es could CVdata, possibly be using differently than my Virus scans, to cause my Red light not to run continously???

Ive been trying to determine the source, of random re-boots, and the only thing I havent tried yet, is new Memory Sticks, HardDrive or a new processor. MeMtest86 showed my Memory was OK tho. Im thinkin this could be a clue to my problem.

 

[QFIT]

Depends on what the red light means. Often it means disk activity. Virus scans constantly access disk. CVData is CPU bound. CVData also hands control over to Windows periodically. This slows it slightly, but keeps from locking out the mouse and other processes.

 

[matt21]

Are you using the MRI function? This excludes selected counts. The excluded counts are displayed in that row.

hi QFIT and thanks for your reply!

i am not sure what you mean by MRI function. I had a look around and under the Nuances/Complex Settings/Data tab i found an area called 'MRI Slice'. In my simulation the 'MRI off' option is ticked. So presumably I am not using the MRI function?

Also, what are 'selected counts' that you mentioned.

Maybe this is easier to resolve if i call you during the day to discuss. Presumably you are part of the company selling and supporting the software?

 

[QFIT]

hi QFIT and thanks for your reply!

i am not sure what you mean by MRI function. I had a look around and under the Nuances/Complex Settings/Data tab i found an area called 'MRI Slice'. In my simulation the 'MRI off' option is ticked. So presumably I am not using the MRI function?

Also, what are 'selected counts' that you mentioned.

Maybe this is easier to resolve if i call you during the day to discuss. Presumably you are part of the company selling and supporting the software?

From the main menu, select Tools then Export and send me the export file. [email protected]

 

[matt21]

From the main menu, select Tools then Export and send me the export file. [email protected]

done!

 

[jack.jackson]

Depends on what the red light means. Often it means disk activity. Virus scans constantly access disk. CVData is CPU bound. CVData also hands control over to Windows periodically. This slows it slightly, but keeps from locking out the mouse and other processes.

Thats interesting. Ill try to run another Hard DisK scan. I just wanted to add that CVData has never caused a BSOD(Blue screen Of Death)It's Mainly, but not limited too, Video and Audio apps. Yes I've tried updating these Drivers.

I was just wondering if there was anyway to resume where I left off w/CVData, should a Random-Reboot occur in the middle of a Index Generation?

 

[QFIT]

Thats interesting. Ill try to run another Hard DisK scan. I just wanted to add that CVData has never caused a BSOD(Blue screen Of Death)It's Mainly, but not limited too, Video and Audio apps. Yes I've tried updating these Drivers.

I was just wondering if there was anyway to resume where I left off w/CVData, should a Random-Reboot occur in the middle of a Index Generation?

No, I used to have a restart capability. But it became to messy with multi-threading.