Simulations for playing multiple hands in negative counts

For those of you interested in this topic, here is a link to BJI. Dan Provost ran Hi-Lo simulations on playing multiple hands in negative counts vs the traditional counting method. In shoe games it can increase your hourly earnings by nearly double!

http://www.bjinsider.com/MemberLogin.php3?loginfile=newsletter_53_inv

This is a MUST READ ARTICLE!!!

-MJ

I've read two of Dan's BJI articles and both came to incorrect conclusions. I'd take the results with a grain of salt.

I agree in principle but the sims just don't tell the entire story. Generally, especially when heads up or maybe one other player, it is best to stick to two hands always, including negative counts. This is both for a better SCORE and for cover. But sims give values on a per round basis and don't account for more time needed to play two hands over one. How to quantify this is tricky. Suffice it to say that most serious counters have known for a good amount of time that multiple hands in negative counts is best.

At crowded tables when cover is not needed this is not the case. Also, the smaller your spread the less you gain. When your spread is very large, say one penny to $10,000, everyone can easily see, without sims, that playing six hands of one penny in negative counts would make you much more money than playing only one. The quicker you use up the low cards the faster you get to the rich deck where you put out big bets.

Of course if you can convince someone else that splitting to multiple hands with lousy counts is always a great idea - play at the same table:)

When you have to bet a certain minimum per hand which is usually the case, it can easily be seen that mathemathically the best method is to play just one hand in negative counts.Let's compare it with playing two hands: For the same money you can play two rounds thereby burning 33% more cards.
On the other hand, in a high count playing multiple hands allows you to reach a good spread without drawing to much attention and it reduces fluctuations.
In other words, I'm not convinced by the Grifter-method.

Francis Salmon

and can play anywhere from 1 to 6 spots. Let us also say that on all positive ev counts you will have 1 hand of 1 million dollars. I'm using an exaggerated example to make a point. Let's say that you can play as many hands as you wish during negative counts for as little as $10 each.

Do you not agree that in the case above that you will make more money, both in terms of win rate per round dealt and win rate per hour, if you played all available spots for $10 each rater than only one hand for $10?

Once you agree to the exaggerated example ( and I think you should agree that multiple hands in negative counts wins more) then it is simply a matter of narrowing it down to realistic spreads and scenarios to find just where this strategy will help you.

I agree in large part about your logical opinion. Previously I even went as far as to declare "card eating effect" as a misnomer. I still think this play should be avoided for most card counters. Let me add my point of view to support your argument. I would also like to point out one special situation in favor of card eating. :)

If the count is negative, the true count is more likely to remain the same (negative counts) for the rest of the shoe. You pay a price by playing multiple hands in a negative EV game hoping that the fluctuation will eventually give you a positive EV. But this happens less than 50% of the time as the true counts will more likely remain negative. That means you are paying a price for nothing more than half of the time. I would think it is cheaper to tip the dealer if she/he will agree to shuffle the shoe upon demand.

Since you pay a price for card eating, the price you pay should be minimum. That means you would bet the table minimum when playing multiple hands in negative EV. It makes sense only if your spread is huge. This would make your minimum bet negligible. That means the spread in a 6-deck game should not be 1:10 or 1:20 but 1:100 or higher. I remember BJTraveller did this in some foreign countries. With a wide spread you need a huge bankroll. In addition, this huge bet spread will draw too much heat in Las Vegas and is not advisable there. The other point is that in a single deck or DD game, the fluctuation in true counts is bigger than multi-deck game and you do hope a bigger wider fluctuation to turn the count to a positive territory.

I think card-eating should not be considered in our regular BJ plays. It is a disservice to most readers here to advocate it. Let�s leave it to high-profile professional BJ players. Oh, maybe you are one of them. :)

I can't see any reason why to play extra hands to eat cards in a 6D game. It's much beter to just wong-out and find another table or just sandbag a few hands until we get a new shuffle.

Even if "card eating" works, it should only be used if there is no other way to get to a reasonably timely new shuffle.

It matters not whether the count becomes positive in that shoe. Your total rounds dealt, in this shoe and all the rest you play, will have a greater frequency of positive ev rounds if you use up the cards in the negative counts more rapidly. This is a simple lever. Put more weight on one side and the other side goes up. The more weight you put on the negative ev true count frequencies, weight meaning more cards per round dealt, the more the frequency will go down and the more the positive ev frequencies will go up.

I understand you are describing a situation where the money lost on negative plays is negligible and the only thing that matters is the number of positive situations you get per hour.
But even in your extreme example I don't see the advantage of playing more hands.In order for your claim to be correct you and the dealer would have to be faster playing through 2 rounds of 6 hands than through 7 rounds at one hand which uses the same number of cards.I doubt that this is really the case because for the same number of cards used the dealer will have to make 12 settlement instead of 7 and you will also need more time to place and collect your bets and to play your hands.
An other point that is often forgotten is that by playing more hands you will actually miss some positive rounds because the count can become positive during the round and swing back before the next round starts.
Besides, you know very well that in real life the spread per spot is usually restricted to 1:50 or so, and even more important the player's bankroll is limited, so it does matter how much money you lose in negative plays.

Francis Salmon

What if you have enough cards that playing the same number of hands would give equal proportions of high and low counts. If you use more cards during poor counts then a greater proportion of the rounds dealt will be in high counts. Simple as that. Just heavily weight the trivial counts and use up the cards so that your big bets are in action more frequently.

And to elaborate on my response to Titan5 who attaches significance to the current shoe: if I am in a heads up play all situation, even if I knew with certainty that the current shoe would not have one more positive ev round dealt, I would still play multiple hands during the negative counts to depleat the shoe in as few rounds as possible in order to quickly get to the next potentially profitable shoe. This is both intuitive and easily proven mathmatically.

Run two sims and forget about the bet sizes. Keep the positive ev counts at the same number of hands always, either 1,2,or 3, doesn't matter. In the first sim play one hand only during negative ev counts. In the second sim play two hands during negative ev counts. Then compare the true count frequencies of the two sims. You will see that the sim where multiple hands were played in negative ev counts gives a much more favorable true count distribution. From there, just give proper weight to each in terms of money bet. You can determine just how much money you can afford to invest, based on your spread, to profit from creating the better true count distribution. In good single deck games it is possible to do this even with flat betting the total action as Grifter has shown, but it has been known about long before that.

Someone betting green or higher will often find a casino full of packed $5 and $10 tables. He may ask the pit to make a $25 table where he will play heads up and get many more rounds per hour. (happens all the time, not a rare thing) Now you will pretty much bet stuck in a play all situation. Multiple hands during negative counts will generally greatly enhance your SCORE in this situation. In fact, it is not really even a close call.

Would you play 7 hands of minimum bet at negative counts? Shark

It is generally not allowed. But on BJ machines where I could play several hands at once, dealing a 4 deck shoe, I used to play all available spots for min bets when ev was negative.

This is really not all that complicated for heads up situations. Just run some sims using various numbers of hands and paste the TC frequencies to a spreadsheet. I used to do this for various situations until I could intuitively tell which was best.

If you could play 20 hands at once you could really get a great win rate. Each negative ev round you would see many cards and see sizeable swing in the count. When positive you could go down to two hands and bet high into the rich count. When negative you could go back to 20 or 30 hands.

Look at it another way: imagine a backcounter who is just watching a game and putting nothing in action. Is there any value to the cards he is seeing? Of course there is. And if there is a value then it would be possible to calculate the value. In fact, you could come up with an exact amount that it was worth to him to see each card during the negative ev segments before he Wongs in. Now, just suppose that after you figure out how much it is worth to him to see each card you find that this amount is greater than the negative ev of playing a hand of blackjack with a minimum bet. That is, the value of each card * 2.7 more than negates the negative ev suffered by playing the hand. In that case there is a gain for playing.

To say that it is not possible to put more money in action and get a greater win rate during negative counts is to say that there is no monetary value to seeing cards. How much it is worth to see the cards is determined by the bets at the upper end. If the bets at the upper end have sufficient ev then it makes the bets at the lower end trivial such that seeing the extra cards is worth more than you could gain by sitting out.

DD', I picked that one from you several years ago. At the time, you were speaking vaguely about it on GC, but you encouraged people to sim the strategy. I ran the sims and incorporated it to my game.I used it for 250 hours alltogether. I felt that I was earning about the same as with playing head on 2 hands at all counts. However, this strategy provided cover and as a result, longevity.

I shared my SBA sim results with few AP's but most doubted them. Shark

Sure there is a value to seeing the cards and we will see them whatever the number of hands we play.I just don't understand why we would have to pay a higher price for them than necessary.

Francis Salmon

Take two kinds of players. One that is playing for small stakes, the kind that hunts a DD game $5-$40 spread with $5 min table limit, as opposed to a bigger player whose min bet is always well above the table min.

Does this change your answer at all?

For example, if I spread to 3 x $5, I could see that it would be faster to eat the negative counts away. But at most places where I have played, I could not do that. I have to bet 2x the table min for the second spot, and some places require 3x the min on all spots to play 3.

I have often used multiple hands, but for two reasons...

(1) increase pen, particularly in SD games dealt RO6 or RO7. With RO6, playing 2 spots (heads up of course) increases the total cards dealt which is a good thing.

(2) greed. With another player at the table, spreading to two or more hands in + counts gives me a greater probability of eating those extra snappers or good cards since I am playing more hands.

But if I find myself wanting to eat negative counts, when my min bet is already at the table min, what is the effect of playing two hands? Covariance usually says play two hands of .76 * bet where bet is the amount you want to bet on a single hand. But I can't play less than the table min. So it would seem that while I am eating the negative count cards, I am also bumping up my RoR as well since I am actually over-betting.

Am I thinking about these issues correctly? It seems clear, based on your comments, that should I play a good SD game, and assuming I would probably spread $10-$40 normally, playing multiple hands at negative counts could be profitable since I could play 2 hands at $5, which is lower than the covariance factor and would actually lower my RoR. But if I start at the table min... ????

Thanks...

and want to make more money. If a card is worth 2 cents and you can see it for 1 cent then I'll pay to see extra cards, even if I'm getting one free. Someone else may value their time less and be willing to sit around and see cards for free. He will make less money. Multiple hands in negatives creates a different game entirely, same type of effect as having better penetration. You create a more favorable true count distribution, get more high count rounds per hundred dealt. A more favorable true count distribution wins more money with a smaller spread.

Best is probably the spreadsheet procedure I suggested before, or perhaps it was on advantageplayer.com, where this same thread is simultaneously taking place. Put the TC frequencies for one, two, and three hands in negative counts each on a spreadsheet and multiply each by the $ in action and the ev of that count, then sum to the bottom. It will easily answer the question of how much you can invest in additional player hands during negative ev situations in order to profit from the strategy. 1:8 is a small spread and having to double the minimum for multiple hands often more than negates the card eating gain. Sometimes it is a close call.