When Does Bad Counting Become Worse Than No Counting?

[Canceler]

Making things too complicated...

qfit's two posts are right on - if you make "minimal" mistakes it wont affect you very much. if youre making anything more than "minimal" mistakes, then you shouldnt be playing in a real casino yet anyway (but hey, it's your money to lose i guess).

I'm with rukus. Either you've practiced enough so that you're confident in your ability to count, or you haven't. If you're confident, try the casino, otherwise stay home and practice some more.

 

[blackjack avenger]

A Little Rhyme!

Don't go to the casino
while counting makes you say oh no

can count it down at home
then to the casino you may roam:devil:

 

[vonQuux]

I'm with rukus. Either you've practiced enough so that you're confident in your ability to count, or you haven't. If you're confident, try the casino, otherwise stay home and practice some more.

While I see your general point, I can't say I agree.

This is a bit like going to your teacher and asking "what's a passing grade?" and they reply "Look, either you know the material or you don't and if you don't, study more."

I think what's happening is that you're assuming that I think the number I'm seeking is the "good enough" point. I'm not suggesting that at all.

Benchmarks are certainly understood to be useful on these forums. For example, I often see "30 seconds" as the target for counting down a deck of cards. If you haven't gotten this fast, you should probably keep going but I don't think anyone is suggesting that once you can, you should stop practicing.

And while I certainly can't (and don't) demand that someone give me the answer, I can (and will) point out that the question still isn't answered.

And given that this seems to be a pretty important benchmark, the fact that it keeps getting shrugged off is awful disconcerting. Nobody is perfect and if you don't know the answer to the break-even point for errata, you cannot possibly know if your play is advantage, can you?

vQ

 

[Automatic Monkey]

I sincerely appreciate everyone's input and there's certainly some good advice contained therein. And so I don't mean to be obnoxious and I DEFINITELY don't want to seem like I'm demanding that someone do the math for me, but I do want to point out that the question remains unanswered.

That is, precisely how far off does my count have to become before the errors cause more harm than if I wasn't counting at all?...

There is no such point. Even "ploppy counting" (looking at the table for an estimation of high and low cards dealt that round) provides information that will decrease the house edge slightly from the house edge against a Basic Strategy player. Thus unless you are discarding BS and/or intentionally betting contrary to the count, random errors cannot make your results worse than not counting at all.

Mr. Renzey is here and can explain it better, but Renzey's Front Count is a method where you backcount until you get an advantage, and then you just sit down, stop counting and play. Sounds bad, but you are playing with an advantage, and not an insignificant one! It's not something I would recommend if you really can count, but it illustrates just how much of a wild-ass approximation anything having to do with counting is.

Advise you not fret about this. Assume you never, ever make a counting error and just bet the count. Otherwise if you are afraid you have made errors, you'll be hesitant to get your big bets out, and that will cost you more than any error you can realistically make will.

 

[blackjack avenger]

Perhaps You Don't Like the Answer

While I see your general point, I can't say I agree.

This is a bit like going to your teacher and asking "what's a passing grade?" and they reply "Look, either you know the material or you don't and if you don't, study more."

I think what's happening is that you're assuming that I think the number I'm seeking is the "good enough" point. I'm not suggesting that at all.

Benchmarks are certainly understood to be useful on these forums. For example, I often see "30 seconds" as the target for counting down a deck of cards. If you haven't gotten this fast, you should probably keep going but I don't think anyone is suggesting that once you can, you should stop practicing.

And while I certainly can't (and don't) demand that someone give me the answer, I can (and will) point out that the question still isn't answered.

And given that this seems to be a pretty important benchmark, the fact that it keeps getting shrugged off is awful disconcerting. Nobody is perfect and if you don't know the answer to the break-even point for errata, you cannot possibly know if your play is advantage, can you?

vQ

The question has been answered a couple times. Don't go to the casino until you can do it correctly at home. Qfit showed you the cost of various types of errors. A minor error here and there is not to costly. It depends so much on the style of play and the games you play that there is no one numerical answser. Also, you cannot measure your success rate at a casino! So you have to over learn at home. You have to be razor sharp at home in order to handle the casino distractions.

To answer your last question. Yes, we can know if we are playing at an advantage because most of us are not concerned with just good enough.

If you are a new player you are not off to a good start. You are thinking of betting large sums of money while striving to be just good enough? Does any parent send their kids to the store with money when they sort of know how to count money?

Don't waste any more of your time on this thread or with this thinking. Practice, if hi low is to hard; have you worked on TC?, there are easier counts.

 

[vonQuux]

There is no such point.

Of course there is. Blackjack is a game of calculations. A simple example is the indices which say that one should insure any hand against a dealer's Ace at a TC of >= 3. Inversely, taking insurance vs. a dealer's Ace when the TC is below three will provide more harm than benefit.

I can also demonstrate that such a number exists by a simple example.

Suppose I sit down at a table and start counting. I'm at +5 when the pit boss asks me a question and when I return to my hand I think to myself "What was the count again? Oh, right, negative twenty!" I think we can probably agree that my expected profit is harmed by this error which means, in turn, that there is some point at which my error went from being "suboptimal but better than BS" to "worse than BS alone."

And all I'm trying to do here is figure out where this point lay. That's all. It isn't heretical to want to know the envelope in which I'm operating...

Even "ploppy counting" (looking at the table for an estimation of high and low cards dealt that round) provides information that will decrease the house edge slightly from the house edge against a Basic Strategy player. Thus unless you are discarding BS and/or intentionally betting contrary to the count, random errors cannot make your results worse than not counting at all.

You hit the nail right on the head without realizing it. You're acknowledging that intentionally betting contrary to the count is harmful which supposes that the game cares about your intentions. =)

Mr. Renzey is here and can explain it better, but Renzey's Front Count is a method where you backcount until you get an advantage, and then you just sit down, stop counting and play. Sounds bad, but you are playing with an advantage, and not an insignificant one! It's not something I would recommend if you really can count, but it illustrates just how much of a wild-ass approximation anything having to do with counting is.

Great, perhaps you're right on this but what I'm asking is a very simple question that I seem to be having a hell of a time getting a straight answer to -- how "off" does one need to be before the effort is worse than nothing at all?

The question has been answered a couple times. Don't go to the casino until you can do it correctly at home.

But ...that isn't the question I'm asking. My question is "at what point does bad counting become worse to one's bottom line than if one did not count at all."

Qfit showed you the cost of various types of errors. A minor error here and there is not to costly.

Excellent, now we're getting a bit warmer! At what point are errors "too costly?"

It depends so much on the style of play and the games you play that there is no one numerical answser.

Just a moment ago you said the question had been answered and now you're saying that it's not answerable?

Further, this is not a credible answer since I've specified the style of play (perfect BS) and the precise game rules. There is a mathematical answer to this question. I'm not demanding that anyone do that calculation for me but to say that there is no answer is absurd.

To answer your last question. Yes, we can know if we are playing at an advantage because most of us are not concerned with just good enough.

Unless you play a perfect game, in order to know for certain that you are playing AP, one must know at what point one's play ISN'T AP.

If you are a new player you are not off to a good start. You are thinking of betting large sums of money while striving to be just good enough? Does any parent send their kids to the store with money when they sort of know how to count money?

This is drifting into ad hominem silliness and has nothing to do with the simple math question I've posed.

Don't waste any more of your time on this thread or with this thinking. Practice, if hi low is to hard; have you worked on TC?, there are easier counts.

Thanks for the advice but with all due respect I'm going to ignore it.

Yeeesh.

vQ

 

[Sonny]

This question is not as simple as it seems. There are several things to consider when it comes to mistakes and profitability, especially when trying to make a “grade” to measure them.

As AutoMoney said, it takes a ridiculous amount of errors (and big ones at that) to negate the effects of card counting. I mean, you would have to be a compete idiot to count so badly that you did not have some sort of advantage. There is no doubt about that. However, there is much more to Advantage Playing than card counting.

If someone is making several mistakes while counting then I would tend to believe that they might not be betting properly either. We have been getting countless posts from new members who claim to be good counters but don’t know how to spread their bets. Either they don’t know when to raise their bets, or by how much, or what kind of spread they need to use, or what their max bet should be, or what kind of risk they are taking, etc. If someone is making mistakes with the counting part then they are probably not playing a winning game in the first place.

Then there is the issue of variance. There is a very good reason why Qfit made his charts using SCORE instead of win rate. Looking only at your win rate is not an accurate way to compare mistakes. For example, if your TC is always +1 higher than it should be then you will always be betting too much. This would actually increase your win rate despite the fact that you are making a pretty big mistake. If you were to look at the variance of this mistake you would see that it is significant. You are winning more money but you are taking much more risk while doing so. If someone has calculated their bet sizes, risk of ruin, and N0 based on perfect play then they will be overbetting, playing with more risk than they anticipated and not getting to the long run as fast. These would all have to be factored into the grading system even though the played may never even know about them.

As we can see, the cost of mistakes should be measured by much more than just EV. If making 1 mistake per shoe decreases your SCORE by 5, what grade do you get? If making 2 mistakes increases your RoR by 10%, how does that change your grade? If it adds another 200 hours to your N0, is it worth it? What is your criteria for determining when someone is ready to play for real in a casino? Obviously everybody is going to make mistakes because we are all human, but at what point is someone not good enough to play? All of these things would have to be considered in the “grade” you give.

This brings to mind a comment Qfit made about the Speed Count system. He showed that the system is so weak that the risk involved makes it a bad strategy for most people. Even though it does give the player a small advantage, the player will go broke so often (because of the variance) that the system is not an effective winning strategy:

http://www.blackjackinfo.com/bb/showthread.php?p=21603

I would imagine that making several mistakes would have a similar effect as using a weaker system. Even though you may always have an advantage, your skills can easily make the gamble not worthwhile.

-Sonny-

 

[vonQuux]

Thanks Sonny.

I think you're saying that one would have to define many tolerances before an error in one particular aspect of advanced play could result in the whole system being less optimal than omitting that AP technique altogether.

In other words, when using the phrase "worse than nothing," how does one define "worse?"

So it would appear that I could get an answer if I knew enough to ask a better question.

But I don't. So I'll go back to practicing counting. :grin:

Your success in articulating the deficiencies of my logic is appreciated.

vQ

 

[Sonny]

I think you're saying that one would have to define many tolerances before an error in one particular aspect of advanced play could result in the whole system being less optimal than omitting that AP technique altogether.

Mainly I wanted to point out that there are several ways that a mistake can hurt you. Just because a particular mistake doesn’t change your EV much doesn’t mean that it isn’t hurting you pretty badly some other way. There are some mistakes that seem almost invisible or insignificant but can affect your real-world results significantly. Many people don't think about that when they consider the effects of mistakes, but a good grading system should.

-Sonny-

 

[Bojack1]

Thanks Sonny.

I think you're saying that one would have to define many tolerances before an error in one particular aspect of advanced play could result in the whole system being less optimal than omitting that AP technique altogether.

In other words, when using the phrase "worse than nothing," how does one define "worse?"

So it would appear that I could get an answer if I knew enough to ask a better question.

But I don't. So I'll go back to practicing counting. :grin:

Your success in articulating the deficiencies of my logic is appreciated.

vQ

Yeah Sonny's good at getting a point across.

My opinion here is don't worry whats the least amount of skill it will take to play a winning game. I believe anybody who asks the question of how bad can I be and still win, has the wrong outlook on the game to begin with. How many people would hire someone who walked in looking for a job announcing, whats the least amount of work I can do and not get fired? If you are asking these type of questions I think you already gave yourself the answer in your last post, keep practicing.

 

[vonQuux]

My opinion here is don't worry whats the least amount of skill it will take to play a winning game. I believe anybody who asks the question of how bad can I be and still win, has the wrong outlook on the game to begin with. How many people would hire someone who walked in looking for a job announcing, whats the least amount of work I can do and not get fired?

This answer illustrates the core problem here and it's a problem I notice is seriously rampant on these forums.

Yes, I was asking from a practical perspective but once it was made clear that the counting would have to be far more deviated from what I would go into a casino being able to do, I was still interested in the answer from a purely mathematical perspective.

Sheer curiosity. That's it.

But still, everyone was hell-bent on tacking their motives unto my query and I find that really, really annoying, not to mention condescending. If a person doesn't know the answer, OK. If someone else doesn't want to answer, so be it. But trying to protect me from my own motives (which I don't even actually have) is well past paternalistic.

If I did intend to half-ass my learning, whose business is that but mine?

vQ

 

[blackjack avenger]

No Respect for the Elders LOL

Of course there is. Blackjack is a game of calculations. A simple example is the indices which say that one should insure any hand against a dealer's Ace at a TC of >= 3. Inversely, taking insurance vs. a dealer's Ace when the TC is below three will provide more harm than benefit.

blackjack avenger
>>>>>>>>>Oh really? It depends on the count used. Are you using an RA indice or CD indice. There is an example of you thinking there is an exact number when it isn't quite so.

I can also demonstrate that such a number exists by a simple example.

Suppose I sit down at a table and start counting. I'm at +5 when the pit boss asks me a question and when I return to my hand I think to myself "What was the count again? Oh, right, negative twenty!" I think we can probably agree that my expected profit is harmed by this error which means, in turn, that there is some point at which my error went from being "suboptimal but better than BS" to "worse than BS alone."

>>>>>>>>>>You did not demonstrate that there is an exact number. That is your question right?

And all I'm trying to do here is figure out where this point lay. That's all. It isn't heretical to want to know the envelope in which I'm operating...

>>>>>>>>>>Your system should tell you the envelope you are operating in.

You hit the nail right on the head without realizing it. You're acknowledging that intentionally betting contrary to the count is harmful which supposes that the game cares about your intentions. =)

Great, perhaps you're right on this but what I'm asking is a very simple question that I seem to be having a hell of a time getting a straight answer to -- how "off" does one need to be before the effort is worse than nothing at all?

>>>>>>>>>>Many issues in blackjack are case dependent. Often there is no simple answer or one answer. If there was one easy answer you would have gotten it by now. Several heavy hitters on this site have answered your question.

But ...that isn't the question I'm asking. My question is "at what point does bad counting become worse to one's bottom line than if one did not count at all."



Excellent, now we're getting a bit warmer! At what point are errors "too costly?"



Just a moment ago you said the question had been answered and now you're saying that it's not answerable?

>>>>>>>>>>The question was answered without the exact number you are looking for.

Further, this is not a credible answer since I've specified the style of play (perfect BS) and the precise game rules. There is a mathematical answer to this question. I'm not demanding that anyone do that calculation for me but to say that there is no answer is absurd.

>>>>>>>>>>Are you so sure? Being off by one has more weight at the end of the shoe then at the beginning. If you are off in the positive direction under certain circumstances it can improve your play. In other instances it can be a detriment. There is no one answer! If your inexperience wants to disagree with my experience; and the others, then that is your choice.

Unless you play a perfect game, in order to know for certain that you are playing AP, one must know at what point one's play ISN'T AP.

>>>>>>>>>>No player living plays a perfect game. What do you think I mean when I say that? A counting system is a best estimate of the cards value, not the exact value. BS is not exact. The indices we use are not exact.

This is drifting into ad hominem silliness and has nothing to do with the simple math question I've posed.

>>>>>>>>>>My analogy was farily sound. Didn't you bring up the teacher analogy? The student asks the minimum to pass? So you want to risk real money on minimal qualifications? Well, minimal qualifications may maximize your chances of going broke! I imagine I have more experience then you. It also seems my comments are very similar to the other experienced players here. My analogy was meant to question what appears to be your desire to bet real money as soon as your are just barely over some razor thin line of acceptable error rate! What if we give you some number but we are off by X. Are you perhaps starting to get the idea of the folly of your thinking?



Thanks for the advice but with all due respect I'm going to ignore it.

Yeeesh.

>>>>>>>>>>Wait, so I am clear. Several experienced players have told you to work on your proficiency, yet you are not going to take that advice? You know better then we do? ok!

>>>>>>>>>>I will throw you a curveball that you may not realize. The hi low count all by itself is in error. It is an estimate of the value of each card. It is a very good estimate, especially when considering ease of use. It is one of the standard counts. However, compared to higher level counts it is as much as 5% less effective. So there you go, inaccuracy can win.

>>>>>>>>>>Perhaps a little Yin and Yang logic. Errors will be made, don't make errors.

vQ

>>>>>>>>>>Yeeesh times 2 LOL
>>>>>>>>>>Sonny if you got through, congrats

>>>>>>>>>>P.S. If anytime in the future someone takes time to try to answer your question don't think of it as confrontational.

 

[Automatic Monkey]

Of course there is. Blackjack is a game of calculations. A simple example is the indices which say that one should insure any hand against a dealer's Ace at a TC of >= 3. Inversely, taking insurance vs. a dealer's Ace when the TC is below three will provide more harm than benefit.

Actually, no! Not if you also take it when it is above TC=+3. Snyder demonstrated in an article about insurance cover that you can simply insure everything larger than your minimum bet and you will receive most of the benefit of taking insurance at the ideal point.

Even when you take insurance at the ideal point, sometimes your insurance bets will be -EV. Suppose the TC=+3 because the shoe is ace-heavy. Being you are not sidecounting aces you're not going to know this and you are going to take insurance, even though it is a losing bet. Big deal. It doesn't mean you should forget about taking insurance or start sidecounting aces. So you have to think of your insurance bets as one big condensate of all your insurance bets, rather than just the advantage or disadvantage of making one individual bet.

I can also demonstrate that such a number exists by a simple example.

Suppose I sit down at a table and start counting. I'm at +5 when the pit boss asks me a question and when I return to my hand I think to myself "What was the count again? Oh, right, negative twenty!" I think we can probably agree that my expected profit is harmed by this error which means, in turn, that there is some point at which my error went from being "suboptimal but better than BS" to "worse than BS alone."

And all I'm trying to do here is figure out where this point lay. That's all. It isn't heretical to want to know the envelope in which I'm operating...

So what you are saying is that you are replacing the count with a random number and betting it. If you were simply a BS player who made betting decisions based on a random number, the result would be exactly like any other BS player, no better and no worse. If you are deviating from Basic Strategy based on this random number it would indeed be harmful, just like any other illogical deviation from Basic Strategy. Being taking insurance and deviations from BS aren't that common even for a counter, this kind of error wouldn't have that much of an effect on a counter either. It doesn't seem like a valuable sort of thing to calculate.

But do you know I make this kind of error all the time myself? I reverse the count, switching from positive to negative and negative to positive. I do it every time I take a drink at the table, not only once, but TWICE! Do you have any idea how much this must cost me in terms of EV?

You hit the nail right on the head without realizing it. You're acknowledging that intentionally betting contrary to the count is harmful which supposes that the game cares about your intentions. =)

In terms of counting and it's effect on results, yes, because that calls in the difference between random and systematic error. Suppose I have a bias or a superstition where I count every black queen as a low card. That's going to introduce a systematic error; it's not going to correct itself and it's going to be present for every measurement in every shoe and get worse as I progress through the shoe. Now suppose 50% of the time, at random, I count an 8 as a high card and 50% of the time as a low card. That's a random error, and the average sum of that error at any time will be zero. The effect of that error on my EV will be related to the average difference between the actual count and my count with the errors and it will be a much smaller number.

Thus the results of our individual hands don't care about our intentions, but the results of our endeavor in counting do. That's cool, I like a good metaphysical discussion!

 

[vonQuux]

No Respect for the Elders LOL

I got this far into your reply and stopped.

 

[vonQuux]

So what you are saying is that you are replacing the count with a random number and betting it. If you were simply a BS player who made betting decisions based on a random number, the result would be exactly like any other BS player, no better and no worse.

Suppose I have a bias or a superstition where I count every black queen as a low card. That's going to introduce a systematic error; it's not going to correct itself and it's going to be present for every measurement in every shoe and get worse as I progress through the shoe. Now suppose 50% of the time, at random, I count an 8 as a high card and 50% of the time as a low card. That's a random error, and the average sum of that error at any time will be zero.

I thought the same thing and it was the basis for a cover strategy but unless I'm missing something (entirely possible), QFIT laid this one to rest here.

I'm kind of jacking my own thread here but it's pretty much dead anyway, so... QFIT, can you comment on this one?

vQ

 

[Automatic Monkey]

I thought the same thing and it was the basis for a cover strategy but unless I'm missing something (entirely possible), QFIT laid this one to rest here.

I'm kind of jacking my own thread here but it's pretty much dead anyway, so... QFIT, can you comment on this one?

vQ

You missed an important part- although the average sum of the errors will be zero the average effect on your advantage will not be. It is still a very small effect, much smaller than that of a systematic error but it also doesn't give you very much cover.

 

[blackjack avenger]

Why? No Sense of Humor Or?

I got this far into your reply and stopped.

Afraid you may learn something? I and others tried to answer your question and point out the error of your thinking. So someone gives their time to you and you think it is not worthy of your attention? How rude and arrogent. LOL

However, I am not offended. I wish you luck in being on the edge of the envelope.

 

[vonQuux]

Afraid you may learn something?

No, it's a Pavlovian thing.

vQ

PS: The above quote is how far I got into your most recent post. Learning yet? :grin:

 

[bluewhale]

this is actually a good question to know. obviously its not like any counter should say "okay as long as i can beat the game i might as well start playing". But with the movie 21 out theres bound to be a slew of new "counters". Is it possible for some of these chumps to be actually beating the game (marginally). I.e. they are estimating the count trying hard to keep up and using a BS chart they bought at the gift store. that might actually work.

the answer of course i imagine would be too weird. it would depend on what kind of mistakes a person makes. how often he makes them. conditions etc.

 

[blackjack avenger]

Dried Bones on the Highway!

this is actually a good question to know. obviously its not like any counter should say "okay as long as i can beat the game i might as well start playing". But with the movie 21 out theres bound to be a slew of new "counters". Is it possible for some of these chumps to be actually beating the game (marginally). I.e. they are estimating the count trying hard to keep up and using a BS chart they bought at the gift store. that might actually work.

the answer of course i imagine would be too weird. it would depend on what kind of mistakes a person makes. how often he makes them. conditions etc.

If you read the posts on these topics it does appear they are asking how much to just get by!

You mention the variables, there would be many to consider on how costly being off by X would be. The game, bet ramp, spread, playing style. I can easily think of scenarios where it would not matter much and others where it could be devastating. There is no one answer X.

However, having said that. There is somewhat an answer. Level one counts are "off" when compared to higher level counts but they are still effective. Higher level counts are "off" when considering the value of the actual cards removed. Unless you are going to use exact value of cards removed then all counts are off.

Just the term "marginal" tells you all you need to know. Many would go broke because their win cannot overcome the SD. Many would not make enough even if winning. Many would quit because the SD would be unsettling.

Isn't the Ace Five count a winning count but the advantage is so small that it's in practical terms worthless?

I can see where the question can have some interest, but there is just not much here. Is it better to take time and effort on figuring out how to win more or spending time on figuring how to just get by?

The big hitters have provided answers and I would think if this was an important question more time and effort would have been provided.

It seems to me the term marginal and advantage player seem contradictory. Many marginal's will end up as dried bones sucked dry by the casinos!