Teams?

[Sonny]

The problem is unless its a time in the shoe when it happens that there is a whole number of decks left you will be underbetting between 1/4 and 3/4 unit for each bet.

You won’t technically be underbetting if your simulator is set up to the same standards as your actual casino play. For example, single-deck estimation (assuming truncating, not rounding) will combine the results of all counts from 1.0 to 1.99999 and use that as the advantage for a TC of +1. You haven’t missed any advantage because it is all summed into that +1 TC. You may not be able to recognize the difference in a +0.12% advantage and a 0.21% advantage, but you will still be sizing your bets based on the overall advantage of a +1 count.

Just be aware that using quarter-deck estimations can sometimes “compress” your TCs, which leads to less accurate results and a lower EV:

http://www.advantageplayer.com/blackjack/forums/bj-main/webbbs.cgi?read=21413 (Archive copy)

I agree with your "work hard to get every edge you can" philosophy, but I think sometimes you can work smarter, not harder. I suppose it comes from my computer programming background. If a piece of code only gets used once during the program’s startup then don’t bother speeding it up too much, but if it gets called hundreds of times per second then you need to spend a lot of time on it even if it is already pretty fast. Optimizing only the most important elements can often be a huge benefit.

-Sonny-

 

[Bojack1]

You won’t technically be underbetting if your simulator is set up to the same standards as your actual casino play. For example, single-deck estimation (assuming truncating, not rounding) will combine the results of all counts from 1.0 to 1.99999 and use that as the advantage for a TC of +1. You haven’t missed any advantage because it is all summed into that +1 TC. You may not be able to recognize the difference in a +0.12% advantage and a 0.21% advantage, but you will still be sizing your bets based on the overall advantage of a +1 count.

Just be aware that using quarter-deck estimations can sometimes “compress” your TCs, which leads to less accurate results and a lower EV:

http://www.advantageplayer.com/blackjack/forums/bj-main/webbbs.cgi?read=21413 (Archive copy)

I agree with your "work hard to get every edge you can" philosophy, but I think sometimes you can work smarter, not harder. I suppose it comes from my computer programming background. If a piece of code only gets used once during the program’s startup then don’t bother speeding it up too much, but if it gets called hundreds of times per second then you need to spend a lot of time on it even if it is already pretty fast. Optimizing only the most important elements can often be a huge benefit.

-Sonny-

Hey Sonny I think you're misunderstanding what I'm saying. If I have a RC of 8 with 1 deck left my TC is 8 and I bet 7 units. We don't compress anything. We calculate the deck estimation to the nearest quarter to find our true count, but if there is 1 full deck left thats what we divide by. Say with a RC of 8 and 2 1/4 decks left our true count would be 3.5 and our bet would be 2.5 units. Its just basically playing at the most accurate and yet practical way we can. Our true counts are as close to exact as you can realistically achieve in live casino conditions. There is no rounding or truncating or compressing, it is what it is. We take in the information given on the table and try to process it into the most accurate form of betting to produce a strong game. As I said before maybe its not for everyone, thats fine, but this is how we have been playing for years with much success. There are some stronger counting systems out there, but in our opinion the error risk out weighed the gain. Believe me we have played this game too much not to weigh as many options as possible but what we do has been the most profitable. So if it ain't broke don't fix it.

 

[Sonny]

We don't compress anything. We calculate the deck estimation to the nearest quarter to find our true count, but if there is 1 full deck left thats what we divide by.

Ah, now I get it. You're using full-deck resolution with quarter-deck accuracy. Basically, you're adding more "notches" to your ruler. Instead of measureing everything in inches your measuring it in quarter-inches. That's pretty dang impressive for shoe games!

-Sonny-

 

[ScottH]

Ah, now I get it. You're using full-deck resolution with quarter-deck accuracy. Basically, you're adding more "notches" to your ruler. Instead of measureing everything in inches your measuring it in quarter-inches. That's pretty dang impressive for shoe games!

-Sonny-

You have to be even more accurate than that for successful shuffletracking.

 

[Sonny]

You have to be even more accurate than that for successful shuffletracking.

No, just the dealer does. The dealer must be consistent and reliable in order for you to get an edge against them. If they make uneven grab sizes or incinsistent moves then your predictions will be off.

-Sonny-

 

[sagefr0g]

Originally Posted by Bojack1
Say with a RC of 8 and 2 1/4 decks left our true count would be 3.5 and our bet would be 2.5 units.

Originally Posted by Bojack1
We don't compress anything. We calculate the deck estimation to the nearest quarter to find our true count, but if there is 1 full deck left thats what we divide by.

Ah, now I get it. You're using full-deck resolution with quarter-deck accuracy. Basically, you're adding more "notches" to your ruler. Instead of measureing everything in inches your measuring it in quarter-inches. That's pretty dang impressive for shoe games!

-Sonny-

i think i'm missing something here or not understanding how this works.
if i'm understanding correctly advantage is not a linear function between true count integer values. the point being how would Bojack justify the bet of 2.5 units for the true count of 3.5 unless he had set his simulator to such values?

best regards,
mr fr0g

 

[Sonny]

if i'm understanding correctly advantage is not a linear function between true count integer values.

That's exactly right. Your advantage does not increase linearly as the count does. If you have a 1% edge at TC+3 and a 1.5% edge at TC+4 then you do not necessarily have an edge of 1.25% at TC+3.5. Similarly, if you normally bet 4 units at +3 and 8 units at +4, you shouldn’t bet 6 units at +3.5. The only reliable way to get your % advantage is by using a simulator.

I remember a similar thread on AP.com:

(Dead link: http://www.advantageplayer.com/blackjack/forums/bj-newbies/webbbs.cgi?read=5230)
(Dead link: http://www.advantageplayer.com/blackjack/forums/bj-newbies/webbbs.cgi?read=5250)

-Sonny-

 

[Bojack1]

Originally Posted by Bojack1
Say with a RC of 8 and 2 1/4 decks left our true count would be 3.5 and our bet would be 2.5 units.

Originally Posted by Bojack1
We don't compress anything. We calculate the deck estimation to the nearest quarter to find our true count, but if there is 1 full deck left thats what we divide by.



i think i'm missing something here or not understanding how this works.
if i'm understanding correctly advantage is not a linear function between true count integer values. the point being how would Bojack justify the bet of 2.5 units for the true count of 3.5 unless he had set his simulator to such values?

best regards,
mr fr0g

Alright first things first, lets get off the simulater and get on the real table for now. When I'm looking at the discard tray I'm estimating how many decks are left to be played. Now I have been trained to be able to determine deck estimation in regular counting to the 1/4 deck. So if I see 3 3/4 decks left in the discard tray that tells me there are 2 1/4 decks left to be played. So just like anyone else using the hi-lo counting method, I divide the remaining decks into the running count. My example earlier had a RC of 8 so lets use that again. I divide 2 1/4 into 8 getting a true count of 3.5. From there I subtract my true offset of 1 to get the proper betting unit of 2.5. I'm not sure where I'm losing you here. Is it the true offset? The true offset represents the initial disadvantage that must be overcome to reach a break even game. On average most figure the house advantage to be .5% with perfect basic strategy. The true offset is always equal to 2 times the initial player disadvantage. With some good games the disadvantage can be less, but as a rule most just round it up to 1.
So basically you find the true count, use that for playing basic strategy indices, then subtract the offset to give yourself the properly sized bet. I'm not sure if this answers your questions but if not let me know exactly what it is you're not following and I will try to answer it.

 

[sagefr0g]

Alright first things first, lets get off the simulater and get on the real table for now. When I'm looking at the discard tray I'm estimating how many decks are left to be played. Now I have been trained to be able to determine deck estimation in regular counting to the 1/4 deck. So if I see 3 3/4 decks left in the discard tray that tells me there are 2 1/4 decks left to be played. So just like anyone else using the hi-lo counting method, I divide the remaining decks into the running count. My example earlier had a RC of 8 so lets use that again. I divide 2 1/4 into 8 getting a true count of 3.5. From there I subtract my true offset of 1 to get the proper betting unit of 2.5. I'm not sure where I'm losing you here. Is it the true offset? The true offset represents the initial disadvantage that must be overcome to reach a break even game. On average most figure the house advantage to be .5% with perfect basic strategy. The true offset is always equal to 2 times the initial player disadvantage. With some good games the disadvantage can be less, but as a rule most just round it up to 1.
So basically you find the true count, use that for playing basic strategy indices, then subtract the offset to give yourself the properly sized bet. I'm not sure if this answers your questions but if not let me know exactly what it is you're not following and I will try to answer it.

the term true offset is not one i've seen in the literature. the term by your description seems very similar to the term true edge as described by Snyder in BlackBelt in Blackjack when he describes the Hi/Lo Lite count. if the two terms are equivalent then i could read Snyder's description more thoroughly and perhaps come to a better understanding of your approach. if the two terms aren't equivalent i guess i would need further explaination of the term true offset or referances so i could look it up.
best regards,
mr fr0g

 

[bluewhale]

Betting With the Count

Hi, i just got a pretty basic question here....
Okay so when you have a positive count of 3 or more (approx.) you have an advantage right? So now what i am saying is why do you need to bet a particular amount (2.5 units or whatever they were talking about). Why don't you just bet as much as you possibly can get away with when you have an advantage. Theorically, wouldn't you just bet the table max everytime the TC gave you the advantage??

 

[sagefr0g]

Hi, i just got a pretty basic question here....
Okay so when you have a positive count of 3 or more (approx.) you have an advantage right? So now what i am saying is why do you need to bet a particular amount (2.5 units or whatever they were talking about). Why don't you just bet as much as you possibly can get away with when you have an advantage. Theorically, wouldn't you just bet the table max everytime the TC gave you the advantage??

wouldn't it be great if that would work .
unfortunately having an advantage even a high advantage doesn't mean that you will definately win a given hand. so the problem is that such unrestrained ie. not making the optimal bet would yield a disasterous ROR (risk of ruin).

best regards,
mr fr0g

 

[bluewhale]

bankroll

oh yeah, did i mention that i'm assuming here that you have an unlimited bankroll....

and don't ask me why i'm playing blackjack with an unlimited bankroll

no but seriously, what do you really mean by "optimal bet"?? because i am sure that if your bankroll was big enough, the right play would be to bet the max on any advantage. So is the optimal bet designed to minimize your ROR?

 

[sagefr0g]

oh yeah, did i mention that i'm assuming here that you have an unlimited bankroll....

and don't ask me why i'm playing blackjack with an unlimited bankroll

no but seriously, what do you really mean by "optimal bet"?? because i am sure that if your bankroll was big enough, the right play would be to bet the max on any advantage. So is the optimal bet designed to minimize your ROR?

:laugh:

on the serious note to a great degree optimal bet is designed to minimize your ROR and provide the most bang for your buck. it takes into account your bankroll, advantage and variance. the optimal bet is derived from an equation that provides what is known as the Kelly wager the bets that provide the optimal logarithmic increase according to bankroll, advantage and variance. i must profess i haven't a full understanding of just exactly how to explain the meaning, mathematics and how the equation is justified. Kelly betting is supposed to be the best way to maximize your profits when making a wager with out ever losing all of your bankroll. most advantage players consider full Kelly betting to be to risky hence adjust their bets to around 1/3 Kelly or lower.

best regards,
mr fr0g

 

[Cass]

oh yeah, did i mention that i'm assuming here that you have an unlimited bankroll....

and don't ask me why i'm playing blackjack with an unlimited bankroll

no but seriously, what do you really mean by "optimal bet"?? because i am sure that if your bankroll was big enough, the right play would be to bet the max on any advantage. So is the optimal bet designed to minimize your ROR?

To answer your question, yes. If your br was big enough you should bet the max possible whenever you have the edge.

 

[ScottH]

To answer your question, yes. If your br was big enough you should bet the max possible whenever you have the edge.

But please, do not try this at home!

 

[sagefr0g]

I will tread lightly here as I do not wish to offend anyone. I do respect anyone who tries to play this game of blackjack at an advantage. However I have seen so many try to do everything possible to try to simplify their game. What it boils down to is really finding ways to weaken it. I have been around a lot of successful advantage players in my life, some of the best I suspect. The common thread between them all is that none of them try to make what works easier, they work hard to make hard work easier. Now I understand it doesn't have to be that way for everybody. If you're a recreational player that plays for the fun more than money than I understand, and there is absolutely no reason to play any other way than what you feel comfortable with. But since all advantages in this game are so minute, its been ingrained in me to squeeze every last bit out that I can. There are many levels of commitment to the craft of advantage play, truth be told you will know which one you are at with the rewards that you reap. Good luck to all who play this game, we all need some degree of it, I just choose not to depend on it.

ditto. too me the 'counter culture' is just about the kewlest thing i've experienced and i'm crossing the threshold into my golden years :violin:
those who sallie forth upon forays into the forboding kingdom of giants in order to do battle and obtain golden eggs and victory have my utmost respect.
one advantage player once quoted Lincoln with regard to the skills of advantage play. circa " the act of taking the time to sharpen the axe gains more firewood than hours of chopping".
hopefully i'm able to restrain the beast within and never offend any of my much respected and more (and/or less) gifted colleagues. it's been said a team is no stronger than it's weakest link. this site in a sense is a team when you think about it.

very best regards,
mr fr0g

 

[Sonny]

So just like anyone else using the hi-lo counting method, I divide the remaining decks into the running count. My example earlier had a RC of 8 so lets use that again. I divide 2 1/4 into 8 getting a true count of 3.5…I'm not sure where I'm losing you here.

Here’s where you’re losing me: What do you do with that extra 0.5 TC? You have a TC of +3.5. You either have to truncate/floor it to +3 or round it up to +4. You can’t bet an extra 0.5 units (or whatever adjustment you would make) just because the TC is 3.5 instead of 3.0. An increase of 0.5 in the TC does not indicate a linear increase in your advantage. So how do you vary your bets based on the extra 0.5 TC? You would have to know the advantage at every TC in increments of 0.5 instead of 1.

-Sonny-

 

[Bojack1]

Here’s where you’re losing me: What do you do with that extra 0.5 TC? You have a TC of +3.5. You either have to truncate/floor it to +3 or round it up to +4. You can’t bet an extra 0.5 units (or whatever adjustment you would make) just because the TC is 3.5 instead of 3.0. An increase of 0.5 in the TC does not indicate a linear increase in your advantage. So how do you vary your bets based on the extra 0.5 TC? You would have to know the advantage at every TC in increments of 0.5 instead of 1.

-Sonny-

Hey Sonny, You are technically correct in the fact we don't get extra advantage playing the way we do. What we do get is the assurance that we will never be off with our bets or become sloppy. By sizing decks to the quarter there is never a chance we will get complacement with throwing what we estimate the correct bet is. Its really just about playing a strong game with very little error. Its like overkill on anything else, just because you don't need it doesn't mean you should'nt have it. Since I do play with a team, and it is a shared bankroll, playing as we do requires a skill level that we can be confident translates well to superior table play. As you know playing on a team requires trust, and its easy trust someones play when you know they have more than enough knowledge to play the game.

Now I have nothing to do with training others aside from those that play on my team. But from what I've learned from people who do train others is if you do only the minimum of what it takes to get the advantage, than thats what you'll get. Even worse is that with lax training methods comes increasingly poor play, usually unknown to the player. Mike Aponte is a friend of mine and is an avid weight lifter aside from being a world class blackjack player. He explains it like this, if you train in the gym and you can lift 400 lbs thats great, and you probably won't ever be called on to lift that much in the real world, but everything else you need to lift becomes much easier, so that extra effort in the gym pays off with out having to max out your capablities in real situations. Thats about all I can say about why we play the way we do, its really the only way we know how.

 

[Automatic Monkey]

Hey Sonny, You are technically correct in the fact we don't get extra advantage playing the way we do. What we do get is the assurance that we will never be off with our bets or become sloppy. By sizing decks to the quarter there is never a chance we will get complacement with throwing what we estimate the correct bet is. Its really just about playing a strong game with very little error. Its like overkill on anything else, just because you don't need it doesn't mean you should'nt have it. Since I do play with a team, and it is a shared bankroll, playing as we do requires a skill level that we can be confident translates well to superior table play. As you know playing on a team requires trust, and its easy trust someones play when you know they have more than enough knowledge to play the game.

Now I have nothing to do with training others aside from those that play on my team. But from what I've learned from people who do train others is if you do only the minimum of what it takes to get the advantage, than thats what you'll get. Even worse is that with lax training methods comes increasingly poor play, usually unknown to the player. Mike Aponte is a friend of mine and is an avid weight lifter aside from being a world class blackjack player. He explains it like this, if you train in the gym and you can lift 400 lbs thats great, and you probably won't ever be called on to lift that much in the real world, but everything else you need to lift becomes much easier, so that extra effort in the gym pays off with out having to max out your capablities in real situations. Thats about all I can say about why we play the way we do, its really the only way we know how.

Fair enough, and if that's what works to develop the discipline needed to play a good game, that's what works.

To me, it's more important to know the difference between what really makes a difference, and what doesn't. So I apply what I call the "lifetime of blackjack test." You don't get any more than 2,000,000 hands in a lifetime, assuming you play full-time for all of your adult life. So any addition to your game that will not produce a greater than one standard deviation chance of being ahead of not doing it (if that sentence makes any sense!) in 2,000,000 hands is not worth it. I haven't calculated them all but I don't think any individual advanced play other than insurance and 16 vs. 10 are above this threshold.

But with team play, I see your point. It will complicate things for a team if players start suspecting each other of weak play or worse, and you'd hate to see a team break up because someone accused someone of not playing DD 9 vs. 7 right or something ridiculous like that, so to keep the peace more accuracy might be helpful, even if it is only psychological.

 

[sagefr0g]

Fair enough, and if that's what works to develop the discipline needed to play a good game, that's what works.

To me, it's more important to know the difference between what really makes a difference, and what doesn't. So I apply what I call the "lifetime of blackjack test." You don't get any more than 2,000,000 hands in a lifetime, assuming you play full-time for all of your adult life. So any addition to your game that will not produce a greater than one standard deviation chance of being ahead of not doing it (if that sentence makes any sense!) in 2,000,000 hands is not worth it. I haven't calculated them all but I don't think any individual advanced play other than insurance and 16 vs. 10 are above this threshold.

But with team play, I see your point. It will complicate things for a team if players start suspecting each other of weak play or worse, and you'd hate to see a team break up because someone accused someone of not playing DD 9 vs. 7 right or something ridiculous like that, so to keep the peace more accuracy might be helpful, even if it is only psychological.

your approach brings to mind an interesting side issue. understanding the idea of Eliott Jacobson's " the blackjack zone" is i believe related to your thinking on your lifetime play test. doubtless such considerations are crucial. perhaps it's because i'm crossing over into the 'golden age zone' and have only just been playing blackjack for less than two years that i have an interest in and am prompted towards considering shorter term issues than the long term. i mean i believe even some one in my circumstance should base ones planning with respect to advantage play upon long term prospects however one such as myself naturally is going to be interested in ones shorter term prospects.
i suppose this is where mathematical approachs such as N0 would be of interest to guys in my circumstances.
additionally i find my self curious as to what degree advantage play yielding say about 1.5% advantage impacts short term prospects. i suppose standard deviation is what can help one gain some cognizance with regard to that issue. any one have any comments or thoughts regarding this?

best regards,
mr fr0g