Bankroll memory

Everybody seems to agree that the cards have no memory. Previous losses or wins have no bearing on the future results. (Not talking about actual cards being played from a fixed deck as is the basis for counting)
Why is it then assumed that the cards have a memory of your bankroll? If you start playing with a $50,000 bankroll, how do the cards know what you have left of this bankroll?
Let's say you've hit that negative variance and you've lost $30,000 of your bankroll. Do the cards know you've lost this much? At that point, aren't you just as likely to lose another $30,000 as when you first started? And when you've lost $60,000 (I guess you had to borrow $10,000) you would still be at the same point of possibly losing another $30,000!

It is true that you are likely to lose another 30,000 of the 50,000 that is why there is an optimal betting proportion -Professional BlackJack- You bet a certain percentage of your bankroll based upon your current advantage, that is how you create your spread. For instance a $3000 bankroll = 3000 x .76 x .005 = 11.4 which means that at an advantage of .5% at +2 True count you would bet $11.4 roughly $10. at a true count of +6 you would bet $57 roughly $60. This is where you get the $5-$60 spread. If you had double this bankroll, the bets would be double. And if you had half this bankroll $1500 the you don't have enough to play =( because 5-30 yields virtually nothing in gains.

It doesn't matter what your bankroll is, as long as your betting strategy is the same, then your probability for losing X dollars in Y hands is always the same. What does change is your risk of ruin. If you start with $60,000 and play a strategy with a risk of ruin of say, 5%, and lose $30,000, your risk of ruin is now much higher than 5% because your bankroll is now $30,000. Thus you should change your strategy, as John Doe suggests, if you want to keep your risk of ruin small after sustaining large losses.

Everything else you wrote seems to be correct, but...

Why is it then assumed that the cards have a memory of your bankroll?

What makes you say that?

Why is it then assumed that the cards have a memory of your bankroll? If you start playing with a $50,000 bankroll, how do the cards know what you have left of this bankroll? Let's say you've hit that negative variance and you've lost $30,000 of your bankroll.

You no longer have a $50K bankroll, you have a $20K bankroll and you should adjust your bet sizes accordingly. Actually, you should probably have adjusted them long before you hit that point. Whatever your risk of ruin was at the $50K level is now much, much higher at the $20K level if you haven't adjusted your bet sizes.

That is essentially what Kelly betting is all about. You constantly adjust your bet sizes based on the amount in your bankroll and the advantage you have. It is impossible to do this on a hand-by-hand basis, but you can do it at certain strike points in your bankroll. This doesn't totally avoid risk of ruin, but it reduces it considerably. There are also table minimums and maximums that get in the way of pure Kelly betting.

I don't know of anyone who makes the assumption that the cards have a memory of your bankroll. They don't. You are the one in control of your bankroll and you have to make the necessary adjustments as you go along and as conditions permit.

So if you do run into a negative variance, you then have even less of a chance of ever getting ahead because you've been forced to reduce your wagers.
It sounds like you better hope for some positive variance to start with or you will be either going down the tubes or simply going nowhere. Forgive my scepticism but it looks like using a huge bet spread just increases your variance so much that it becomes a far more important factor than some small advantage. You either hit positive variance and espouse the wonders of card counting or you hit negative variance and swear off the hole thing... That is if you don't wind up laying in a Vegas gutter or scrounging for quarters to build up your bankroll again.

You'll get much more detailed information than you would in a couple of paragraphs answer (even from the esteemed Bootlegger).

You only employ the "huge bet spread" when the count is in your favor, i.e. the deck statistically favors the player.

If you bet larger when the deck statistically favors the casino, such as with progressions in a negative count, you'll lose all your bankroll eventually.

The proportion of bet size vs. total bankroll determines your Risk Of Ruin - the greater the ratio, the higher the ROR. It is possible to lose 60% of your bankroll, but with reduced probability if you resize your bets downward as the bankroll diminishes. When your bankroll increases, you resize your bets upward.

Wong's and Schlesinger's books are some good ones to start out with.

Oh yeah, what's the best way to prove that card counting works? Casino's throw out card counters

Wally

Let me clear up a couple of the points you made in:

So if you do run into a negative variance, you then have even less of a chance of ever getting ahead because you've been forced to reduce your wagers. It sounds like you better hope for some positive variance to start with or you will be either going down the tubes or simply going nowhere. Forgive my scepticism but it looks like using a huge bet spread just increases your variance so much that it becomes a far more important factor than some small advantage.

Suppose that you lost a large part of your bank. Now you are considering resizing your bets downward. If you downsizer bets, then you increase the probability that you will eventually get out of the whole, but it will take you a longer time to do it.

Second, on this issue of bets spread and variance. A bigger bet spread should have lower variance. Your big bets should be determined by your bank and utility (RoR), and do not depend on the spread. Your waiting best depend on the spread. A big spread means smaller waiting bets, and lower variance.

If the table minimum does not allow you to get such big spread, then you should increase the amount of Wonging you do.

Given a table minimum, you would get the lowest RoR is you Wonged in at high counts (like TC of 5) and flat bet the table min. Of course, it will take you a very long time to make any money with this approach, but your RoR will be very small.

Sure seem to win a lot of first hand deals and win alot with +3 negative count After the first hand. So do counters Ever follow a trend in winning as a percentage or not.If the Trend for the evening was more in favor on a particular night,do you heavy,stratigic hitters work it or advise against it in when asked, to sound correct.
I can hear dont gamble but invest,however do you notice this trend and work it.Not trying to be snide.I would say not every trend is tied with high counts.I have seen trends end quickly,however I also seen a entire decks dealt and win more often with no help of any kind to the dealer with a low count or flucuating count.Im not talking about max wager here, but how about mabey a few hands of flat bets.Be honest because I being very honest about noticing trends.

The problem with trends is that you don't know one existed until it is over. There is no rhyme or reason to trends. A perceived trend can end with the next hand, and the odds say that it will. To truly gain an advantage, one must bet when he knows he has the advantage. Counting is one way to make that determination. Simply following trends based on hands won or lost in the past tells you nothing about what will happen in the future.

You either hit positive variance and espouse the wonders of card counting or you hit negative variance and swear off the hole thing... That is if you don't wind up laying in a Vegas gutter or scrounging for quarters to build up your bankroll again.

This is a recurring them for new advantage players. They often underestimate the nature and intensity of the swings in blackjack. Many players (even those with good skills) get a horrible downswing right off the bat and give up. Others are very lucky right off the bat and never look back. Most players simply win more or less about what they are supposed to win.

Remember it takes a good number of hands (measured as the figure N0) to have an 83% chance of being exactly even or ahead (eV exceeds 1 standard deviation of variance). For a good game this figure can be as low as 7500 hands (about 75 hours of play at a good single deck game) and for an H17 DAS shoe game with 1.5 decks cutoff it can be as high as 50,000 or more hands (500 hours of play) Between the start of your carreer and reaching 1 cycle of play (N0) there can be some serious ups and downs.

"A bigger bet spread should have lower variance."

Looks to me that the bigger your spread (relationship between your small bet and big bet) the higher your variance in units. Of course, the relationship between your variance and ev will make it look a lot better but if you are looking at it in terms of units, the bigger the spread, the bigger the variance.

with a bigger spread so your variance in pure dollar terms is less.

Who said you have to keep the same RoR ? From the discussion, I just want to make it clear for the readers that if as an example, you are spreading 1-6 in a DD game ($25-150) and suddenly because of the game conditions (no heat, clueless pitboss,etc.) you decide to play a better game in term of EV by adjusting your spread to let say 1-10 ($25-250) you will encounter greater swings both in $ and in units (variance). Your example implies that you will lower your units but this will also lower your expectation in term of "pure dollar" as you say. This might not be optimum for some of us.

The word of caution here is: even if you are playing a better game in term of EV and DI by spreading more, beware of the implication it might have on your RoR. Make sure you are adequately bankrolled.

you said that a wider spread has higher variance...to check this you must use the same bet on the top end so if you are comparing a 1-4 spread ($50-$200) with a 1-8 spread you must look at $25-$200.

For a six deck shoe game S17 DAS rules a 1-4 spread in $50 units has a variance of $129 per hand, same game with 1-8 spread in $25 units has a variance of $94 per hand. The wider spread has a lower risk of ruin while maintaining the same top bet of $200.

Now if you are to keep risk of ruin constant, then a wider spread will also yeild a higher top bet so that a 1-4 spread with a $200 max might be equivelent in risk to a 1-8 spread with a $300 top bet. In this case the wider spread will have higher variance (but also a much higher DI and SCORE because eV will have increased at a much higher rate than variance.

I don`t know where you are going or where you want to take me. The thing is , I won`t discuss this till the next moon comes out. Clear and simple, a higher spread yelds a higher variance in terms of unit, period. And this is the only thing I was saying. We don`t need a sim to understand that reducing your small bets will be the way to avoid it in terms of dollar.

GMan wrote:

I don`t know where you are going or where you want to take me. The thing is , I won`t discuss this till the next moon comes out.

Fine. We don�t need to discuss it anymore either.

You started this sub-thread with a question to me �Are you sure?� The answer was YES. Bigplayer elaborated on it. I think everything that needs to be said has been said by BP and me.

I hope your serious that you want the thread to end, because I won�t post anything more. You may be one of this people that must absolutely have the last word, and if so, you will probably respond. However, I don�t intend to respond further.

Several can easily mistake this as putting more money on the table..as in a higher max bet. I had to read your post twice to understand what you were talking about. I think the simpliest way to say it is, "A higher spread will decrease variance if the max bet is the same "dollar" amount as the previous max bet of the smaller spread that was being used."

As long as one is not talking about "unit variance" which is as useless as the concept of "unit" itself. The bigger one's spread the smaller one's dollar variance will be as the bigger spread essentially means a smaller fraction of the max bet is put down in neutral and negative counts. We can see Griffin's discussion of the Opie and Greta games. Opie spreads with the advantage while Gretta makes a flat bet equal to the average advantage in plus ev situations.