Basic card counting question.

[LuckyStars]

Hello to all,
I‘ve recently started learning to play BJ seriously. While saving my initial BR, I’m practicing my counting and Basic Strategy by playing for 1-2 hours an evening. I deal out 3 hands and record the outcomes. This allows me to see the win/loss swings that I read about, as well as practice CC and BS.
I never deal through the entire deck because the casinos I’ve played at always had a cut card within the shoe so I’m simulating that. But what I’m noticing is there’s often an abundance of high or low cards within the bottom 1-1.5 decks that seems to me is skewing the entire count.
Maybe I’m looking at it wrong or overthinking it, but I get shoes where I’m betting high, or the minimum on the negative counts the entire shoe, only to find the count stayed positive or negative because of a group of cards that the dealer will never get to.
Thanks for any replies and I apologize if I didn’t use the proper terminology.

 

[DSchles]

While dealing to yourself in this manner can provide a minimal amount of practice, please understand that you can't really draw any valid statistical conclusions from a few hundred, or even thousand, such hands. Eventually, you will need simulation software to provide that kind of information.

Fluctuations in the count of the kind you describe are perfectly normal and are an inherent part of the game. But, the biggest mistake newcomers make is to assume that anything that happens while practicing on their kitchen table can be extrapolated to represent "reality." Have fun practicing, but just don't worry about the results.

Don

 

[LuckyStars]

Thanks for the reply. That’s what I needed to hear.

 

[LuckyStars]

While dealing to yourself in this manner can provide a minimal amount of practice, please understand that you can't really draw any valid statistical conclusions from a few hundred, or even thousand, such hands. Eventually, you will need simulation software to provide that kind of information.

Fluctuations in the count of the kind you describe are perfectly normal and are an inherent part of the game. But, the biggest mistake newcomers make is to assume that anything that happens while practicing on their kitchen table can be extrapolated to represent "reality." Have fun practicing, but just don't worry about the results.

Don

I hope I can reply without any comments from trolls. I spent almost 10 years as an industrial engineer, where I learned a lot about statistics. So I want to discuss a few of your points,
it doesn’t matter if I’m at a kitchen table, as you described, or a class room, or a table. A true test is measured by the facts, not where it was performed. ex. Garbage in/Garbage out.....
You agreed that the occurrence I explained was normal. Either it be a general occurrence or an anomaly, it should be explained. I’ll never know because you tell me it’s built into the calculationof models that I must buy but then depend on being correct. not saying your wrong, but I’ve built several models that can be manipulated by a simple error.
Without selling me some program, please explain how I can disprove my concern before I spend thousands of dollars.

 

[kcchiefsfan1982]

Thanks for the reply. That’s what I needed to hear.

Basically, if you take 5 card counters that count perfect, play perfect, with a perfect bet spread...in the first 500 hours...the 5 players will have results that differ greatly from each other.

But, by 2,000 hours...their results should be pretty tight and start to look similar.

 

[xengrifter]

Basically, if you take 5 card counters that count perfect, play perfect .... by 2,000 hours...their results should be pretty tight and start to look similar.

'Start' being the operative word...
... even at 100k hands there will be noticeable variance in results recorded.

 

[DSchles]

I hope I can reply without any comments from trolls. I spent almost 10 years as an industrial engineer, where I learned a lot about statistics. So I want to discuss a few of your points,
it doesn’t matter if I’m at a kitchen table, as you described, or a class room, or a table. A true test is measured by the facts, not where it was performed. ex. Garbage in/Garbage out.....
You agreed that the occurrence I explained was normal. Either it be a general occurrence or an anomaly, it should be explained. I’ll never know because you tell me it’s built into the calculationof models that I must buy but then depend on being correct. not saying your wrong, but I’ve built several models that can be manipulated by a simple error.
Without selling me some program, please explain how I can disprove my concern before I spend thousands of dollars.

Quick answer: you don't need to spend thousands for simulation software. You just go to qfit.com and buy what you need, for hundreds, not thousands.

The hourly standard deviation of a counter's results is anywhere from 15 to 25 times the hourly expectation. At the upper end, since you understand the math, for e.v. to equal just one s.d., giving you an approximately 84% chance to be ahead at the conclusion of play, at 100 hands per hour, you'd need to play 625 hours, which might take you several years of actual casino play. We call this statistic N0 (N-zero).

To be more certain of the results, out to two standard deviations (~98% certainty), you now need four times that value, or 2,500 hours of play!

You're going to be at your kitchen table a very long time!! Or, you can buy the software that performs billions of hands in seconds, and gives you the most reliable results known to mankind. Obviously, every serious player knows what the right choice is, but that is entirely up to you.

Don

 

[Dog Hand]

Quick answer: you don't need to spend thousands for simulation software. You just go to qfit.com and buy what you need, for hundreds, not thousands.

The hourly standard deviation of a counter's results is anywhere from 15 to 25 times the hourly standard deviation. At the upper end, since you understand the math, for e.v. to equal just one s.d., giving you an approximately 84% chance to be ahead at the conclusion of play, at 100 hands per hour, you'd need to play 625 hours, which might take you several years of actual casino play. We call this statistic N0 (N-zero).

To be more certain of the results, out to two standard deviations (~98% certainty), you now need four times that value, or 2,500 hours of play!

You're going to be at your kitchen table a very long time!! Or, you can buy the software that performs billions of hands in seconds, and gives you the most reliable results known to mankind. Obviously, every serious player knows what the right choice is, but that is entirely up to you.

Don

Don,

A small but significant typo in your answer: you wrote "The hourly standard deviation of a counter's results is anywhere from 15 to 25 times the hourly standard deviation." The end of your sentence should have been "...from 15 to 25 times the hourly expected value."

Hope this helps!

Dog Hand

 

[LuckyStars]

Basically, if you take 5 card counters that count perfect, play perfect, with a perfect bet spread...in the first 500 hours...the 5 players will have results that differ greatly from each other.

But, by 2,000 hours...their results should be pretty tight and start to look similar.

Thanks for the reply. That makes sense.

 

[LuckyStars]

Quick answer: you don't need to spend thousands for simulation software. You just go to qfit.com and buy what you need, for hundreds, not thousands.

The hourly standard deviation of a counter's results is anywhere from 15 to 25 times the hourly expectation. At the upper end, since you understand the math, for e.v. to equal just one s.d., giving you an approximately 84% chance to be ahead at the conclusion of play, at 100 hands per hour, you'd need to play 625 hours, which might take you several years of actual casino play. We call this statistic N0 (N-zero).

To be more certain of the results, out to two standard deviations (~98% certainty), you now need four times that value, or 2,500 hours of play!

You're going to be at your kitchen table a very long time!! Or, you can buy the software that performs billions of hands in seconds, and gives you the most reliable results known to mankind. Obviously, every serious player knows what the right choice is, but that is entirely up to you.

Don

Thanks for the explanation. While I normally practice while I’m on the road for work, I definitely want get better sooner then later.

 

[kcchiefsfan1982]

Thanks for the explanation. While I normally practice while I’m on the road for work, I definitely want get better sooner then later.

I hope your initial start goes well. It will go one way or the other. Mine went south...it is hard to mentally recover from putting hours in, only to lose money.

 

[LuckyStars]

I hope your initial start goes well. It will go one way or the other. Mine went south...it is hard to mentally recover from putting hours in, only to lose money.

I appreciate that. I’ve always liked the game but usually just take a few hundred for a night and hope for luck lol. The math behind it is really interesting so I’m practicing and saving at least a small bankroll before going in.
I‘ll admit that some of these stories are pretty scary. I’d never thought a loosing streak could go on like what I’m learning.

 

[xengrifter]

I‘ll admit that some of these stories are pretty scary. I’d never thought a loosing streak could go on like what I’m learning.

Gambling with a positive advantage is still gambling.