Regarding counting and progression
  • I have been reading a lot of good posts lately regarding counting and progression here. I am impressed by counter's logical thinking, by progressor's empirical evidence, and also by the idea of those who try to combine counting and progression. Here is what I attempt to share:

    I have used Hi-Lo Count for betting after I had been convinced by its mathematical logic, but I didn't win money with it and I began to question its logic with the following: :?:

    a. back counting wasted a lot time;
    b. the count won't go high at all during a lot of games;
    c. uncertain total winning hands when the count is high and bet is big.
    d. betting before seeing the dealer's up-card (relative to the logical beauty of knowing the dealer's up-card when playing Basic Strategy, relative to also the logical beauty of modifying Basic strategy when counting)

    I then began to use Walter’s Progression with Quit Points that recommends a stop after a 4 streak of losing hands. I do play by the true count deviation of I18. The result was and still is a big success for many years since I put my hands on Walter’s book. I never lose with this approach. Only once in the beginning I did lose when I got carried away and deviated from the progression pattern and didn’t quit after 4 losers in a row as per Walter’s rule.

    As an PhD engineer by education, I know the limitation of sims, to me, blackjack is not only a mathematical game, but also a situational game. Linear thinking helps, but the reality is more non-linear and non-random, and "long run" is only a "theoretical construct" which should be assessed by process. When I see debates between counters and progressors, I always have a feeling that they are cross talking. Blackjack is a game of arts and science, not a game of science only.

    As for the long run, the human brain is notoriously bad at statistical comparison, and in fact, you may not have played enough hands to "prove" card counting statistically in real world. If you sat there 8 hours a day, 40 hours a week, for a year, flat betting with basic strategy, getting 50 hands per hour (full table), and you took detailed records so you know you lost 1% of your action, then you still would not be able to say with 95% confidence that your expected edge was worse than -.5%.

    (1.97 * 1.1 / sqrt(200000) = .69% as the +- uncertainty in the confidence interval.) Two years would be barely enough. If you're counting and spreading big based on the true count, then boost that to something like a decade.

    Any success in Blackjack depends not only on how to win, but also on how to protect from loss. You can't win very consistently if you let yourself involved in a long losing streak.

    Every system requires self control.
  • I once heard "Only in academia does a purist succeed." The same might also be true in blackjack; it's the eclectic that has the best results. ....or should I say the eclectic that has self control. Good comments Alex.
  • AlexD30 said:


    As an PhD engineer by education, I know the limitation of sims, to me, blackjack is not only a mathematical game, but also a situational game.


    I wondered if you could expand on these limitations a little. Do you mean situational, as in "real life", that other factors than those in a sim come into play? Interruptions, conversation, "real time anxiety", etc.? I've always questioned how accurate my results from simulations are. I find I'm very confident (and successful) with certain progs on a sim, only to find myself in a different world when I sit at the table. I think it's a fascinating, and overlooked subject. And you seem to have the knowledge to shed some light on it.

    Great post.

    John
  • I’m not capable to have a scientific explanation why it works. Because of my technical education I still rationalize using science/math. I don’t know how to pinpoint, but it does work. The best advice I can give is play your hands using true count strategy like religion while waiting for that hot streak where you can do no wrong and the dealer continues to pay. It is possible, in theory, to bet $$ by the count alone and play 21 at a financial advantage. It is extremely difficult to do in practice because of the horrendous fluctuations due to standard deviation.

    I think most would agree that counting will provide a marginal advantage if done well and if the counter is reasonably lucky. Unfortunately games are getting less favorable, and even in its heyday counting was a borderline proposition.

    I strongly recommend both progression systems: Walter’s and Dahl’s combined with “quit points” while using card counting for Mag7 or I18. Both systems gives the average player the best way to capitalize on streaks while at the same time reducing his risk of going broke. The methods reminds me of Kelly criterion applied to a theoretically negative expectancy game. You're always betting the minimum while losing. You are never risking more than your minimum bet while on a losing streak. When you hit a win streak with a $5 progression, and win $500, you can step up to the $10 progression. When you make another $750 profit (a total bankroll of $1250) you can step up to the $25 progression. You incrementally keep stepping it up when winning and down when losing. If you go below a total bankroll of $1250, you step back down to the $10 progression, and if your bankroll goes below $500, you're back at the $5 progression. So using this method, your getting the absolute most for your gaming dollar without risking big bets during losing streaks. It extends your playing time, and gives you lots of opportunities to hit hot streaks. Because you're always maximizing your wins, there are times when you can go from a total bankroll of $1000 all the way to $10,000 in one day of playing. It has happened to me many times to end up my day with over 10K profit while starting from a buy-in of $1,250 for green play. I hit some really good hot streaks, and I incrementally increased my stakes proportionally ending up betting with $500 chips before hitting the 10K mark.

    Some say that when you play a progression system you are increasing the amount of $$ at risk against the house edge. Therefore, your total losses will be higher than by flat betting. Well, in theory it is true, but lets assume that a flat bet player theoretically expects to lose $15/100, and a progression player should expect to lose $30/100. Isn't it worth risking that extra $15/hour knowing that you have the potential to really hit a big score? I think it is! - To be successful, you must decide exactly what you want to accomplish. There are two ways of meeting difficulties: you alter the difficulties, or you alter yourself to meet them.


    What would you attempt to accomplish if you knew you could not fail?
  • What would you attempt to accomplish if you knew you could not fail?

    I can answer that in two words:
    Gwen Stephani :lol:
  • Excellant Post Alex! :D a lot of food for thought :wink:
  • Alex: Sounds like a good plan to me! Speaking from personal experience, the toughest part of this plan is dropping back to a lower level after a losing streak starts to eat away at profits. Betting black chips sort of grows on you....
  • Yes, I agree... such wonderful posts by you, Alex!!

    Speaking of his progression, yes, it is indeed a wonderful progression and I'm so glad & thankful I found this website and read his book that thoroughly explains everything about it. However, I had a question about the BEST time to raise it incrementally, as in your case, from 5 to 10 to 25 etc.

    I am a green chip player myself. 2 days ago, I played his progression starting with a base bet of $50 & 2K bankroll. 50-75-100-125 & I was up as much as $2800 but ended up leaving the casino very content with DOUBLE my bankroll, $2000. This being said, at what time would have been the best time for me to increase my base bet to $100, and therefore play 100-150-200-250 progression? I had one REALLY good shoe where I probably profited $1500, should I have increased it right when I begun my 5th potentially winning hand? Or was I supposed to raise it at the start of the next shoe instead???
  • BuGhOu§eMASTER said:
    Yes, I agree... such wonderful posts by you, Alex!!

    Speaking of his progression, yes, it is indeed a wonderful progression and I'm so glad & thankful I found this website and read his book that thoroughly explains everything about it. However, I had a question about the BEST time to raise it incrementally, as in your case, from 5 to 10 to 25 etc.

    I am a green chip player myself. 2 days ago, I played his progression starting with a base bet of $50 & 2K bankroll. 50-75-100-125 & I was up as much as $2800 but ended up leaving the casino very content with DOUBLE my bankroll, $2000. This being said, at what time would have been the best time for me to increase my base bet to $100, and therefore play 100-150-200-250 progression? I had one REALLY good shoe where I probably profited $1500, should I have increased it right when I begun my 5th potentially winning hand? Or was I supposed to raise it at the start of the next shoe instead???

    This will a mixed message but I’m sure you will get the most of it.

    First, You have to reduce the house edge as close to ZERO as possible with flat bet only. This is accomplished by playing your hands by the true count. Using the I18 with INSURANCE in good games with great rules. Like for example: 2D, S17, DAS and late surrender. If you play this game with a base flat bet of $50 you can expect to lose $10 per 100 hands. This game is on Las Vegas Strip in some clubs and is the best around. There are some other games too.

    If you want to hit big scores then you should think differently then card counters. The thing I have learned during my years as a card counter is that the statistics are only statistics and real play is very different. For those players that claim to be BJ pros the statistics make a much greater impact because the amount of play gets them closer to a statistically perfect game. I my case, am looking to offset financial swings in my favor. First, It seems to me that your session bankroll of $2000 was way to small for $50 based bet. Anyway, I would have played this way: If I would have reached $1,000 profit I would have my base bet at $75. Keep betting at that level as long I’m still within $1,000 or more profit. When reach $2,000 profit I would have switch to $100 base bet .., and so on.

    To be successful you must decide exactly what you try to accomplish. I, personally am trying to get as much money out from the game in the shortest period of time. There is only one way to do this and that is: Do not get emotionally attached to small winnings. A small winning is something between $500 and $1000 if you play with $50 base unit. There are two ways of meeting difficulties: you alter the difficulties, or you alter yourself to meet them. You have to alter your way of thinking. I like to treat my small wining as money that I don’t have to supply while betting big. I play my base bet capped with chips from the winning pile and don’t get sentimentally attached until the winning pile become substantial. I have found that learning to take a profit out is the single most important key to the game.

    However, greed and fear cause most players to get burned and I venture to guess that it causes good pro players to get burned as well. Many times I have watched a nice profit turn into a big loss while betting big on positive counts while losing hand after hand again and again. Don’t bet big unless you won some money. Don’t be afraid to bet big if you are within profit. Learn to always take some cash out of the game no matter how small that amount is.

    You will only have a negative expectancy if you flat bet the game. Else, if the average value of winning bets (AWB) exceeds the average value of losing bets (ALB) by a percentage greater than the house advantage over a flat bettor, then you will have a positive expectancy in that game. How do you know for sure when you're going to win and when you're going to lose? You don't, obviously, not even card counters can teach you that trick! But by reducing your next bet in response to any loss and pumping it right after any win you will ensure that overall your AWB will greatly exceed your ALB. Result: consistent profits from sequences of bets in which the house won more hands than you did.
  • Alex: I've played a style similar to yours for the last 20 years, but never
    with a progression. My concept of winning is very simple and it does not
    matter what level of game you play, because your chance to win is like
    you say, "situational" & your ability to take advantage of those situations
    is the difference in how you do in this game.

    When we have a good session, the chance to improve to a better session
    is always there and in some cases a better session can be converted to
    a great session. Take your example of a green player moving from a BR
    of 1200 t0 10,000 in one day. What does that mean: Well, it improves our
    long term chances because as a green player I can go to the casino and
    drop my total session stake 8 streight times,which is not likely. All players
    have different skill levels, but regardless of what that may be, your chances will be improved by extending a good situation.


    Arts and Science: I think you mentioned this earlier and I'd like to
    express my view on the matter as I see it: B/S, B/S variations and
    card counting is most of the science part of the game. The "Arts" part
    is a matter that most blackjack players have a lot of trouble with because
    it involves money and tends to get personal. Can you view a chip; red,
    green,black or purple as just another chip in your quest for a better win.
    Can you see your stack of green chips as an opportunity to advance to
    another level or are you just happy to go home with what you view as a
    good win. What about the limits of win/loss goals that more or less assure
    that you will always be a loser. There are other aspects of "Art", but the
    important one is $$$ and I don't think that most can ever overcome the
    inherent pressures and indecision that comes with it.

    Ray
  • One thing to remember is the money management related to the number of hands won when the time comes to increase your bet. With a positive count, you don't win more hands but actually lose more hands because busts are more likely to occur. You get instead more blackjacks and double downs that raise your EV a little. If you don’t get your fair share of BJs and DDs you will lose big time if you bet big because the count says so. The entire card counting concept is build on the principle that the cards are random and that the player will get his fair share of BJs and DDs during high counts which is not always the case in real world.

    If you don’t get the cards you cannot possibly win, regardless how good you might be on card counting. The only chance to make money using card counting is if the player gets a normal distribution of BJs and DDs opportunities during high counts. In the long run on computer you may get what you suppose to get in regard to BJs and DDs but if you lose a portion of your bank and not make it to the long run that’s a done deal.

    The one thing that is undisputed in this game is the fact that “streaks” occurs during both positive or negative counts. Very interesting will be if you guys will run a simulation and record the number of hands won or lost during positive and negative. You will be surprised to find out that the player actually lose more hands during positive counts vs. the number of winning hands during the same period. Also the player lose more hands during negative counts then he wins. Now, if you bet more during positive you may possibly make a little more money only if you get a fair share of BJs and DDs else you will lose BIG.

    Here are some results of ten million rounds on a 2D, S17, DAS, playing perfect basic strategy with flat bet. Keeping the Hi-Lo count to register the count value related with winning or losing the hand.

    Round 10000000 was completed at :04-12-2004 21:58:11
    The seed was 78995.88
    Dealer has garnered 37491
    Player 1 won -37491 on a handle of 10000000 or -0.37491%
    Players results have a Standard Deviation of +/-3549.648 or 0.03550%
    A maximum of 86 cards were dealt from the shoe.
    Win on Positive = 1870244
    Win on Negative = 2075057
    Lose on Positive = 2192082
    Lose on Negative = 2164066
  • Alex: I have to assume that you use H/L for the purpose of playing
    your hands and not for betting. Is that correct? If this is not the case,
    I can't imagine how you determine when or how to introduce the
    progression into the game because there is no predictive tool that
    I know of that forcast the probability of streaks. My experience in
    winning streaks does not support the idea that the wins are one after
    the other and are more like 3 out of 5 or 6 out of 10 with no specific
    order. Progressions need order and it's not all that hard to show why a
    player could very well miss out in a winning situation because of the
    predetermined structure. That's not to say that progressions don't work.
    It's just that I can't understand how they could work and as a result, I
    just don't play them. Blind faith is not something that I can live with..

    Winning with the high counts is a very marginal situation to be sure
    and if one sits down at a 6 deck shoe game and plays shoe after shoe,
    his chances to win is even more marginal. It takes a combination of things
    and some luck to have a chance. Because the cards are as kind to the
    dealer as they are to the player, what then is the major difference? The
    big difference comes after both dealer and player receive their cards. In
    this situation the player flexibility for doubles, split or not, hit or not is the
    difference. Winning one or two extra doubles with a big bet is all that is
    needed. Not a sure thing to say the least!!

    One other Question: How do you resolve the situation when the count
    indicates a big bet, but the progression requires you to revert to a single
    unit. Or, you lose 4 in a row, but at the sametime the count improves
    to a +2, do you still walk?


    Ray
  • Ray said:
    Alex: I have to assume that you use H/L for the purpose of playing
    your hands and not for betting. Is that correct? If this is not the case,
    I can't imagine how you determine when or how to introduce the
    progression into the game because there is no predictive tool that
    I know of that forcast the probability of streaks. My experience in
    winning streaks does not support the idea that the wins are one after
    the other and are more like 3 out of 5 or 6 out of 10 with no specific
    order. Progressions need order and it's not all that hard to show why a
    player could very well miss out in a winning situation because of the
    predetermined structure. That's not to say that progressions don't work.
    It's just that I can't understand how they could work and as a result, I
    just don't play them. Blind faith is not something that I can live with..

    Winning with the high counts is a very marginal situation to be sure
    and if one sits down at a 6 deck shoe game and plays shoe after shoe,
    his chances to win is even more marginal. It takes a combination of things
    and some luck to have a chance. Because the cards are as kind to the
    dealer as they are to the player, what then is the major difference? The
    big difference comes after both dealer and player receive their cards. In
    this situation the player flexibility for doubles, split or not, hit or not is the
    difference. Winning one or two extra doubles with a big bet is all that is
    needed. Not a sure thing to say the least!!

    One other Question: How do you resolve the situation when the count
    indicates a big bet, but the progression requires you to revert to a single
    unit. Or, you lose 4 in a row, but at the sametime the count improves
    to a +2, do you still walk?


    Ray


    From beginning I want to make it clear that I do not discount the card counting theory. I did for many years use card counting but I didn’t make serious money with. Of course I did have big wins when cards come my way but I also did lose big times when they didn’t and I was betting Max bets on positive counts while riding a losing streak. I come to the conclusion based on my observations that I can make much more money taking advantage of winning streaks regardless of the count. I did some serious simulations over the years and find out that raising the bet only after a winning hand helps reducing the fluctuations during losing streaks plus it compounds the money won while riding the winning streaks
    My method of betting in BJ is a money management that is a "streak system" and relies on the simple truth that while the majority of wins and losses are isolated incidents, many of them come in bunches. Anyone who disputes that streaks are commonplace and can always be relied upon to occur will obviously not believe that skillful betting can produce positive EV.
    The key is to correlate the value of your next bet to the depth of your winning goal that you are trying to accomplish, providing you with a constantly updated win target. At any point, a modest streak will take you within your target. You practically don’t have to know when a winning streak starts or for that matter will end. The only time when you have out there a bet larger then one unit is if you won the previous hand. You single bet unit is dictated by your bankroll while your Max bet is dictated by your previous winnings.

    Now, card counting theory says that your max bet is to be correlated to your bankroll. Well, according to the math, I agree, but the problem with that is that if you ride a losing streak while the count is positive you can lose a lot of money in the process. To make up for this you will have to survive a very long run and get your fair share of BJs and DDs. People have got to understand that the money in card counting comes from BJs and DDs with extra money on the table during positive counts. The number of winning hands is decreased during positive counts and the only thing that makes money for the card counter is that extra bonus on BJs if you get the cards. One extra BJ per hour for you and for the dealer will give you a half unit and another unit probably on doubling will put you at 1.5 units per hour. You have to understand that you are NOT any making money during Hit/Stand or Splits decisions no matter what the count is. Matter of fact your win-rate drops for Hit/Stand decisions during positive counts.


    :?: “One other Question: How do you resolve the situation when the count indicates a big bet, but the progression requires you to revert to a single unit. Or, you lose 4 in a row, but at the same time the count improves to a +2, do you still walk?” :?:

    Well, If the count indicates a big bet but my progression require a single unit, that means that I have been losing the previous hand(s). If I continue to bet big in this situation I may be riding a positive count losing streak that has the potential to wipe me out. If I lose 4 in a row but the count is at +2, I still walk regardless of the count. The only mathematical advantage the count gives is for deviating from Basic Strategy and INSURANCE. That has a solid math advantage. Even if the count is sky high but I lost the previous hand(s) I still bet one unit. The only reasonable hope during positive counts is if I get a BJ or a double down situation which are rare during positive counts anyway. Less 3,4,5 and 6 value cards to combine to a total for doubling down while the count is positive. During positive counts you get less opportunities of doubling on 9,10,11.
    Just run a simulation and tabulate the number of times you have DD during positive vs. negative and you will be amazed.

    The bottom line is the actually winning hands during different counts:
    Winning on Positive = 22.17%
    Winning on Negative = 25.26%
    Losing on Positive = 25.86%
    Losing on Negative = 26.70%

    As you can see, wining on negative, losing on positive and losing on negative are about the same percentage ( they are over 25% or more losing hands). Winning on positive has the smallest winning hands percentage and there is the exact moment when a card counter is trying to offset the difference based on BJs and DDs. Not a good choice by my book!
    Financial risk vs. reward doses not justify betting by the count. The margin of profit is so small compared with the risk exposure that it becomes ridiculous to me. I never won $10,000 betting by the count in my entire BJ career over 20 years. I’m not promoting progression to anybody but I can say that many times over and over I won way much more then $10,000 betting by some progression. The biggest money fluctuations that I had have been when betting by the count and the smallest ones when bet by prog.

    Plug that in your computer. :idea:
  • Alex: I'll never argue with success, but questions are unavoidable for the
    inquisitive mind. Thanks for your response...
  • Alex. I've got to agree with your progressive betting system.

    The last two times I have gained 65% and 190% of my bankroll.

    I use a version of your strategy with a couple of variations:

    1. I can only go to my local Casino lately and they only have so many tables to play at with my level of play. So if I do lose more than 4 hands in a row I have to keep plugging away until hopefully the tide turns.

    2. If my bankroll is getting low I will stop after three wins and go back to my starting bet.

    A couple of questions for you Alex.

    Do you go back to your original bet after a shuffle?
    What's the most hands you've won in a row while using your system? While I've used this system the most I've ever won is 8 consecutive.

    Spitoon/Andy

    PS. I am Andy Garcia in the top score listings.
  • spitoon said:
    Alex. I've got to agree with your progressive betting system.

    The last two times I have gained 65% and 190% of my bankroll.

    I use a version of your strategy with a couple of variations:

    1. I can only go to my local Casino lately and they only have so many tables to play at with my level of play. So if I do lose more than 4 hands in a row I have to keep plugging away until hopefully the tide turns.

    2. If my bankroll is getting low I will stop after three wins and go back to my starting bet.

    A couple of questions for you Alex.

    Do you go back to your original bet after a shuffle?
    What's the most hands you've won in a row while using your system? While I've used this system the most I've ever won is 8 consecutive.

    Spitoon/Andy

    PS. I am Andy Garcia in the top score listings.

    Well, practically makes no difference if you drop your bet to one unit at shuffle. According to Walter’s system he’s starting over at shuffle. I only drop my bet down to one unit after a losing hand. If the count is negative but I’m on a winning streak I keep betting into the progression. There is no point to drop at shuffle when the count is zero if during negative counts on winning streaks I keep on betting.

    The most I ever won in a row as far as I remember was about 12 hands last year. I remember I was in the beginning of my session betting with in green. I didn’t stop/drop to minimum until have lost the last hand. The run was good for $2,400 profit after losing that big bet at the end.
  • AlexD30 said:


    (1.97 * 1.1 / sqrt(200000) = .69% as the +- uncertainty in the confidence interval.)


    Not according to my calculator. Square root of 200000 is 447.21, while 1.97*1.1 is 2.167. The ratio of the two is 0.0048, which I beleive is lower than 0.005

    Leave alone the fact that the z-score for 95% CI is 1.96, not 1.97. At least in the statistics graduate program they told us so. If they think different in the Engeneering Dept., I would not argue.

    Yeah, and why on earth you would need a two-sided C.I. for a hypothesis like this? For one-sided the z-score would be 1.645, the final answer would be 0.040

    Regards

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!