Here are all the comments posted on the site, with the most recent discussions listed first.
To participate in any of these discussions, you can reply on the article page.
I could see this being a one-time advantage play for a mid-sized high-roller who is prepared to lose upwards of $250,000 in a single night of well-feigned intoxication, and then utilize a pre-negotiated/flat-rate lawyer to litigate the marker loss. Similar to a loss-rebate strategy.
Your question asks specifics that probably aren’t relevant to what you really want to know…
Are the random number generators in my trainer and those in a casino video blackjack machine “similar”?
Well, yes and no. They are probably very different in the actual programming, because there are many ways to code an RNG. But yes, they both provide reasonably high quality streams of random numbers.
What you likely really want to know is how your results will compare when playing in a casino compared to on my site’s trainer.
No one can answer that ahead of time, because it comes down to the random results.
If you are luckier on my site, you’ll do better there. If you are luckier in the casino, your results will be better in the casino.
Both my trainer and the casino’s games use the same rules and conditions. What happens after that is luck, good or bad.
The short term results in either place will not give you ANY information about what will happen next time you play.
There is one frequent difference between my trainer and video blackjack in a casino. Many casino video blackjack games pay only even money for blackjack. If that is the case, the screen will likely say “Blackjack pays 2 for 1”. That means they take your 1 unit bet and pay you back 2 units if you get a blackjack. That’s the same as even money, or a “1 to 1” payout. This rule is terrible, and will substantially increase your losses.
Usually, the large multi-player video blackjack games do pay the normal 3:2 for blackjack, but be sure to check the rules on the machine you want to play. Most of the single player vbj machines will have the worse rule.
Some major casinos in Macau don’t allow surrender only when dealer has an Ace. There is no hole card. And they often remind player to take even money. What is the strategy for such rules if you have 8-8 or 7-7 when dealer has a 10? Thanks.
Since the Strategy Engine does not have an option for “Early Surrender, but not against Ace” (yet!), you have to mix and match the charts.
Here is the strategy for 6D, H17, ES, No Hole Card: https://www.blackjackinfo.com/blackjack-basic-strategy-engine/?numdecks=6&soft17=h17&dbl=all&das=yes&surr=es&peek=no
Ignore the advice to surrender against an Ace, since it is not allowed. (Or you can create a second chart with No Surrender and use the Ace column from that.)
For your specific questions: The dealer has a 10 card, there is no hole card, and you are allowed to surrender…
You should surrender the following hands (as seen on the linked chart above):
(7,7) vs Ten
(8,8) vs Ten
Hard 14 vs Ten
Hard 15 vs Ten
Hard 16 vs Ten
Also, Hard 16 vs 9.
first i want to apologyze about the thing concerning my security, it was just bullshit. i was trolling a bit, my country is not dangerous, i love colombia. but some people get angry at you when you play BS. i have another question, here in my city there is a side bet which you gamble to get a pair. if you get a mix pair(red and black) like a 6 of diamonds and a 6 of spikes they pay 6 times your side bet. if you get a color pair(red and red or black and black) like a 6 of diamonds and a 6 of hearts they play 12 times your side bet and if your ge a perfect pair(the same card) like a 6 of diamonds and a 6 of diamonds they pay you 25 times your side bet. my question is it worth it?
Like most side bets, it’s quite expensive. The house edge in six decks on the paytable you describe is 5.47%.
My advice about side bets is very simple: Ignore them.
Are there further lessons that deal with alternative counting strategies? Im going to check anyway, just wanted to say thanks for systematizing what i already knew, looking forward to reading more
Je viens de voir des informations et un extrait de UTube sur le "Speed Count", où l'on ne compte que les cartes plus (2-7) jouées pendant la main, puis on soustrait le nombre de mains jouées pour obtenir un compte courant. Bien entendu, ils recommandent vivement leur système, revendiquant un avantage de 1% pour les joueurs qui l'utilisent. Quels sont les niveaux d'augmentation des mises ? (Je suppose que puisqu'ils utilisent 2-7 comme KO et REKO, on commence à -2 x [nombre de mains]).
J'ai essayé d'utiliser REKO. Je parviens très bien à tenir le compte lors des parties à deux jeux, mais j'ai du mal à le faire lors des parties à six jeux, car je n'ai pas envie de m'asseoir et de regarder les cartes pendant toute la durée de la partie.
Je ne suis pas un fan de Speed Count, à moins que vous ne vous sentiez vraiment incapable de gérer quelque chose de plus puissant. Le compte rapide est un peu plus faible que l'OC ou le Hi-Lo, et lorsque vous jouez une partie à six jeux, il augmente vraiment le nombre de mains que vous devez jouer avant d'atteindre le légendaire "long run". Ma recommandation est de continuer à travailler pour maintenir le compte en utilisant KO. J'ai du mal à croire l'avantage de 1% revendiqué par les joueurs utilisant le Speed Count, sauf dans de très bonnes conditions et avec un grand écart de mise. KO et Hi-Lo peuvent rarement atteindre ce niveau en pratique, avec 0.5% à 0.7% étant un avantage plus typique dans un jeu à six jeux. Le Speed Count fonctionne-t-il ? Oui, mais votre avantage sera plus petit et il vous faudra plus de mains pour atteindre une probabilité élevée d'être en avance sur la variance du jeu.
I have your Advanced Card Set (which are very helpful and I would highly recommend to everyone). I am building a custom playing strategy in CVBJ for DD as I’ve been wanting to learn correct BS and Indices instead of the composite. Comparing the indices on your 2D, H17 card to the matrix you’ve linked above, I’m seeing several differences, are there certain rule assumptions in the two sets of indices that I’m overlooking?
Just looking at the top of hard total section for example (>> card index, link index)
8 vs 5 >> 3, 6
8 vs 6 >> 2, 3
9 vs 2 >> -1, 2
9 vs 3 >> 1, 0
9 vs 4 >> -3, -3 (same)
9 vs 5 >> -5, -4
10 vs 9 >> -2, -1
I have struggled with how to deal with the discrepancies between the GameMaster’s published index numbers and the ones I spent so many hours devising for the strategy cards. GM himself noted that he used risk-adverse index numbers, and that explains most of the differences, although not all. My choice to not risk-adjust my indexes is based on the fact that exceptionally few players bet near a Kelly fraction of their true bankroll. Even players who do make bet sizing calculations based on Kelly tend to use the amount of money they have on hand for gambling right now, with no consideration of future income streams. Of course, that does reduce risk of ruin, and I am definitely not saying that is a bad idea. But it does taint any risk-averse calculations that are based on the lower bankroll estimates. If you choose to use risk-averse numbers, then the appropriate index number literally depends on the size of your bet on that specific hand at the time, which seems ridiculous to have to consider. And it is not like risk-averse indexes can magically make the game far less risky. They have a very small effect, despite the considerable complexities they cause. I prefer (and use) straight index numbers, and comfortably know that I’m doing the right thing because I am not risking anywhere near a Kelly fraction of my bankroll anyway. (In that case, risk-averse and non-adjusted indexes are identical anyway!)
I mention the differences between his indexes and mine inside the lesson at https://www.blackjackinfo.com/blackjack-school/lesson-14-advanced-course-part-2/ .
But the discrepancies do make me uncomfortable, and they require a lot of explanation to readers. Witness this lengthy reply! Since I cannot get in touch with Maître du jeu to discuss a resolution, I have left things as they are so far. But, as time passes, I become more convinced that I should simply edit all of his index numbers so that they are consistent with mine. I spent hundreds of hours optimizing my numbers, and have a great deal of confidence in their utility. Generating index numbers is much more involved than most people realize. Each number is also dependent on the choice of index numbers further down the decision chain. That is, determining an accurate index for 14vT also depends on the values chosen for 15vT and 16vT, since those are among the possible outcomes for hitting 15vT. It is a lengthy recursive process. I don’t know what method GameMaster used to generate his numbers, but I doubt that his methodology was as careful as mine. For that reason, I completely recommend my values. If there is a difference, I say use my number.
The good news is that when the index differs by more than a point or two, that indicates a situation where the edge likely changes slowly as the count changes. That means the most contested numbers are likely those that matter the least, where a point or three won’t affect your results much at all. The truth is that if you were to use all of his numbers vs all of my numbers, the difference in expected results is small.
Thanks, Ken, for the very detailed response. Your last point is especially helpful in explaining the larger discrepancies which were particularly concerning/confusing to me.
So for 9 vs 2 >> – 1, 2 which has a large difference and crosses the +/- threshold, the edge changes slower and the basic strategy decision to ‘double’ (index -1) is close enough that after factoring in risk-aversion it would change the BS decision to ‘hit’ (as the index is >=2)?
On a side note, though I know the incremental expected value is small, are you able to provide or can I purchase your full index sets beyond -5 to 5 for the six sets of rules if you have them?
During the iterative process of generating the indexes, I truncated the list to only -5 to +5 at each step of the process, so I do not have the outlying indexes. Wong’s Blackjack professionnel has a wider range for shoe games in H17 and S17 if I recall correctly.
Hi, I’m somewhat confused about the „european“ strategy (6 decks, S17, D9, DAS, NoSurr, NoPeek(!), no refund(!!) of split or double on dealers BJ) calculated by your engine. Although I understand that 11 vs. ten might run into a dealers BlackJack, it is still 15 ten’s for me to get 21 and only 4 aces for the dealer to reach a BJ (counted per deck). So why not DD on 11 vs. ten?
Next point is not splitting 8s against ten or ace. I can easily follow not to split 8s against A due to „a lot of ten’s around there“. But again, why not splitting that ugly 16 points against a dealers ten, who needs a „seldom“ ace to knock me out? Even more confusing: if 8s should not be splitted against ten, why should aces be splitted against ten?
The short answer is: That’s just the way the numbers work out. Just trust the strategy.
But I’m sure that is not a very satisfying answer for you. So, I’ll try to give you some more insight. What about doubling 11vT? You can’t just look at the likelihood of you drawing a ten versus the dealer drawing an Ace. Most of the time neither of those things will happen. The tricky part is comparing all the other possible outcomes. You’ll make money when you double 11vT, but it won’t be as much as you will make on average by just hitting. For one thing, when you double you give up the ability to draw more than one card. If you draw an Ace,2,3,4, or 5, you would like to be able to hit again, but you are not allowed to. All of those factors go into the calculations, and in a no hole card game you should not double 11vT. (In fact, in the S17 game you describe, you shouldn’t double 11vT even if the dealer did take a hole card!)
The other decisions have similarly complicated explanations. There are rarely easy answers for the “why” of basic strategy. The calculations take all the possible outcomes into account, and show you the most profitable (or more frequently, the least unprofitable!) way to play each decision.
Not true, Black Jack is a game of 2 no more no less. It’s you and the dealer. Whatever anyone else does, shouldn’t affect you. Unless you were counting.
I could see this being a one-time advantage play for a mid-sized high-roller who is prepared to lose upwards of $250,000 in a single night of well-feigned intoxication, and then utilize a pre-negotiated/flat-rate lawyer to litigate the marker loss. Similar to a loss-rebate strategy.
Post script –
Johnson was indicted for check fraud by Vegas DA after grand jury found against him … and then he settled for 50%, I am told –
http://lasvegassun.com/news/2015/jun/12/gambler-indicted-wanted-owing-casino-500000-losses/
<>
How do casino BJ machines compare to your site – similar random # generators, speeds, etc?
Your question asks specifics that probably aren’t relevant to what you really want to know…
Are the random number generators in my trainer and those in a casino video blackjack machine “similar”?
Well, yes and no. They are probably very different in the actual programming, because there are many ways to code an RNG. But yes, they both provide reasonably high quality streams of random numbers.
What you likely really want to know is how your results will compare when playing in a casino compared to on my site’s trainer.
No one can answer that ahead of time, because it comes down to the random results.
If you are luckier on my site, you’ll do better there. If you are luckier in the casino, your results will be better in the casino.
Both my trainer and the casino’s games use the same rules and conditions. What happens after that is luck, good or bad.
The short term results in either place will not give you ANY information about what will happen next time you play.
There is one frequent difference between my trainer and video blackjack in a casino. Many casino video blackjack games pay only even money for blackjack. If that is the case, the screen will likely say “Blackjack pays 2 for 1”. That means they take your 1 unit bet and pay you back 2 units if you get a blackjack. That’s the same as even money, or a “1 to 1” payout. This rule is terrible, and will substantially increase your losses.
Usually, the large multi-player video blackjack games do pay the normal 3:2 for blackjack, but be sure to check the rules on the machine you want to play. Most of the single player vbj machines will have the worse rule.
Some major casinos in Macau don’t allow surrender only when dealer has an Ace. There is no hole card. And they often remind player to take even money. What is the strategy for such rules if you have 8-8 or 7-7 when dealer has a 10? Thanks.
Since the Strategy Engine does not have an option for “Early Surrender, but not against Ace” (yet!), you have to mix and match the charts.
Here is the strategy for 6D, H17, ES, No Hole Card: https://www.blackjackinfo.com/blackjack-basic-strategy-engine/?numdecks=6&soft17=h17&dbl=all&das=yes&surr=es&peek=no
Ignore the advice to surrender against an Ace, since it is not allowed. (Or you can create a second chart with No Surrender and use the Ace column from that.)
For your specific questions: The dealer has a 10 card, there is no hole card, and you are allowed to surrender…
You should surrender the following hands (as seen on the linked chart above):
(7,7) vs Ten
(8,8) vs Ten
Hard 14 vs Ten
Hard 15 vs Ten
Hard 16 vs Ten
Also, Hard 16 vs 9.
first i want to apologyze about the thing concerning my security, it was just bullshit. i was trolling a bit, my country is not dangerous, i love colombia. but some people get angry at you when you play BS. i have another question, here in my city there is a side bet which you gamble to get a pair. if you get a mix pair(red and black) like a 6 of diamonds and a 6 of spikes they pay 6 times your side bet. if you get a color pair(red and red or black and black) like a 6 of diamonds and a 6 of hearts they play 12 times your side bet and if your ge a perfect pair(the same card) like a 6 of diamonds and a 6 of diamonds they pay you 25 times your side bet. my question is it worth it?
Like most side bets, it’s quite expensive. The house edge in six decks on the paytable you describe is 5.47%.
My advice about side bets is very simple: Ignore them.
Are there further lessons that deal with alternative counting strategies? Im going to check anyway, just wanted to say thanks for systematizing what i already knew, looking forward to reading more
Je viens de voir des informations et un extrait de UTube sur le "Speed Count", où l'on ne compte que les cartes plus (2-7) jouées pendant la main, puis on soustrait le nombre de mains jouées pour obtenir un compte courant. Bien entendu, ils recommandent vivement leur système, revendiquant un avantage de 1% pour les joueurs qui l'utilisent. Quels sont les niveaux d'augmentation des mises ? (Je suppose que puisqu'ils utilisent 2-7 comme KO et REKO, on commence à -2 x [nombre de mains]).
J'ai essayé d'utiliser REKO. Je parviens très bien à tenir le compte lors des parties à deux jeux, mais j'ai du mal à le faire lors des parties à six jeux, car je n'ai pas envie de m'asseoir et de regarder les cartes pendant toute la durée de la partie.
Quel est le "bilan" de SpeedCount.
Je ne suis pas un fan de Speed Count, à moins que vous ne vous sentiez vraiment incapable de gérer quelque chose de plus puissant. Le compte rapide est un peu plus faible que l'OC ou le Hi-Lo, et lorsque vous jouez une partie à six jeux, il augmente vraiment le nombre de mains que vous devez jouer avant d'atteindre le légendaire "long run". Ma recommandation est de continuer à travailler pour maintenir le compte en utilisant KO. J'ai du mal à croire l'avantage de 1% revendiqué par les joueurs utilisant le Speed Count, sauf dans de très bonnes conditions et avec un grand écart de mise. KO et Hi-Lo peuvent rarement atteindre ce niveau en pratique, avec 0.5% à 0.7% étant un avantage plus typique dans un jeu à six jeux. Le Speed Count fonctionne-t-il ? Oui, mais votre avantage sera plus petit et il vous faudra plus de mains pour atteindre une probabilité élevée d'être en avance sur la variance du jeu.
Bonjour Ken,
I have your Advanced Card Set (which are very helpful and I would highly recommend to everyone). I am building a custom playing strategy in CVBJ for DD as I’ve been wanting to learn correct BS and Indices instead of the composite. Comparing the indices on your 2D, H17 card to the matrix you’ve linked above, I’m seeing several differences, are there certain rule assumptions in the two sets of indices that I’m overlooking?
Just looking at the top of hard total section for example (>> card index, link index)
8 vs 5 >> 3, 6
8 vs 6 >> 2, 3
9 vs 2 >> -1, 2
9 vs 3 >> 1, 0
9 vs 4 >> -3, -3 (same)
9 vs 5 >> -5, -4
10 vs 9 >> -2, -1
Merci !
I have struggled with how to deal with the discrepancies between the GameMaster’s published index numbers and the ones I spent so many hours devising for the strategy cards. GM himself noted that he used risk-adverse index numbers, and that explains most of the differences, although not all. My choice to not risk-adjust my indexes is based on the fact that exceptionally few players bet near a Kelly fraction of their true bankroll. Even players who do make bet sizing calculations based on Kelly tend to use the amount of money they have on hand for gambling right now, with no consideration of future income streams. Of course, that does reduce risk of ruin, and I am definitely not saying that is a bad idea. But it does taint any risk-averse calculations that are based on the lower bankroll estimates. If you choose to use risk-averse numbers, then the appropriate index number literally depends on the size of your bet on that specific hand at the time, which seems ridiculous to have to consider. And it is not like risk-averse indexes can magically make the game far less risky. They have a very small effect, despite the considerable complexities they cause. I prefer (and use) straight index numbers, and comfortably know that I’m doing the right thing because I am not risking anywhere near a Kelly fraction of my bankroll anyway. (In that case, risk-averse and non-adjusted indexes are identical anyway!)
I mention the differences between his indexes and mine inside the lesson at https://www.blackjackinfo.com/blackjack-school/lesson-14-advanced-course-part-2/ .
But the discrepancies do make me uncomfortable, and they require a lot of explanation to readers. Witness this lengthy reply! Since I cannot get in touch with Maître du jeu to discuss a resolution, I have left things as they are so far. But, as time passes, I become more convinced that I should simply edit all of his index numbers so that they are consistent with mine. I spent hundreds of hours optimizing my numbers, and have a great deal of confidence in their utility. Generating index numbers is much more involved than most people realize. Each number is also dependent on the choice of index numbers further down the decision chain. That is, determining an accurate index for 14vT also depends on the values chosen for 15vT and 16vT, since those are among the possible outcomes for hitting 15vT. It is a lengthy recursive process. I don’t know what method GameMaster used to generate his numbers, but I doubt that his methodology was as careful as mine. For that reason, I completely recommend my values. If there is a difference, I say use my number.
The good news is that when the index differs by more than a point or two, that indicates a situation where the edge likely changes slowly as the count changes. That means the most contested numbers are likely those that matter the least, where a point or three won’t affect your results much at all. The truth is that if you were to use all of his numbers vs all of my numbers, the difference in expected results is small.
Thanks, Ken, for the very detailed response. Your last point is especially helpful in explaining the larger discrepancies which were particularly concerning/confusing to me.
So for 9 vs 2 >> – 1, 2 which has a large difference and crosses the +/- threshold, the edge changes slower and the basic strategy decision to ‘double’ (index -1) is close enough that after factoring in risk-aversion it would change the BS decision to ‘hit’ (as the index is >=2)?
On a side note, though I know the incremental expected value is small, are you able to provide or can I purchase your full index sets beyond -5 to 5 for the six sets of rules if you have them?
During the iterative process of generating the indexes, I truncated the list to only -5 to +5 at each step of the process, so I do not have the outlying indexes. Wong’s Blackjack professionnel has a wider range for shoe games in H17 and S17 if I recall correctly.
Hi, I’m somewhat confused about the „european“ strategy (6 decks, S17, D9, DAS, NoSurr, NoPeek(!), no refund(!!) of split or double on dealers BJ) calculated by your engine. Although I understand that 11 vs. ten might run into a dealers BlackJack, it is still 15 ten’s for me to get 21 and only 4 aces for the dealer to reach a BJ (counted per deck). So why not DD on 11 vs. ten?
Next point is not splitting 8s against ten or ace. I can easily follow not to split 8s against A due to „a lot of ten’s around there“. But again, why not splitting that ugly 16 points against a dealers ten, who needs a „seldom“ ace to knock me out? Even more confusing: if 8s should not be splitted against ten, why should aces be splitted against ten?
Thanks for your support and…great website!
The short answer is: That’s just the way the numbers work out. Just trust the strategy.
But I’m sure that is not a very satisfying answer for you. So, I’ll try to give you some more insight. What about doubling 11vT? You can’t just look at the likelihood of you drawing a ten versus the dealer drawing an Ace. Most of the time neither of those things will happen. The tricky part is comparing all the other possible outcomes. You’ll make money when you double 11vT, but it won’t be as much as you will make on average by just hitting. For one thing, when you double you give up the ability to draw more than one card. If you draw an Ace,2,3,4, or 5, you would like to be able to hit again, but you are not allowed to. All of those factors go into the calculations, and in a no hole card game you should not double 11vT. (In fact, in the S17 game you describe, you shouldn’t double 11vT even if the dealer did take a hole card!)
The other decisions have similarly complicated explanations. There are rarely easy answers for the “why” of basic strategy. The calculations take all the possible outcomes into account, and show you the most profitable (or more frequently, the least unprofitable!) way to play each decision.
Just as much odds of bad play ending in a positive outcome (Take the 5, leave a bust card)
Not true, Black Jack is a game of 2 no more no less. It’s you and the dealer. Whatever anyone else does, shouldn’t affect you. Unless you were counting.
Its simple…. you are prob playing a game with even money black jack