Betting according to EV.... not the true count

For a moment.....put card counting aside...

If you calculated EV on "deck composition", after every hand.
http://www.bjstrat.net/cgi-bin/cdca.cgi
(yes i know its impossible at the tables, just imagine on pc simulations)

And you ranged your bets accoring to the EV.
e.g. 4Deck game has approx -0.28 EV
The shoe played and 5x2's 5x3's 5x4's 5x5's were played...
The EV moves to +0.113

If i stepped up my bet +1 unit, for each +0.5EV, would that be the perfect betting pattern???

The perfect bet size is;

Bet = BR*EV*Kelly Fraction/Variance

Kelly Fraction is user defined but LTE to 1.0

0.25 < Kelly Fraction < 1.0 depends on your circumstances
1.0 < Kelly Fraction < 2.0 reduces your win rate
Kelly Fraction > 2 -----> Bankrupt

That is all there is to it.

Nicely written, but understand something: Kelly fraction < 1 reduces your win rate also. Kelly 0.5 is no different than Kelly 1.5, insofar as reduction of win rate is concerned.

That fact isn't apparent from your post.

Don

There is no reason to ever bet above Kelly. I thought that was obvious but you never know.

I decided to return to playing BJ after a 15 yr absence after going broke due to over betting. When I came back I studied RoR and the mathematics you put into BJA. You even helped me through some of the calculations. I came back with an anal focus on making sure my betting , BR, and RoR were appropriate since a lack of understanding in the nineties cost me fifteen years.

You were one of the first to focus on the critical aspect of RoR. Most of the authors before you made only passing and incomplete references to it.

Counting is easy. The hardest parts are getting the hours in and managing risk properly. Norm's software make the latter easy.

Thank you for your huge contributions!

I still struggle with the knowledge that based on getting in a maximum of one N0 I have a one in six chance of LOSING over an entire year. But I can't do much to change that. It is what it is. And I've had those stretches.

I was also the unfortunate member of a team for a relatively short time that had miserable results and lost the confidence of my team mates as a result. You covered this exact scenario in BJA as well under the team play section.

I have learned that over the long run the math works but the long run is much longer than most of us would expect it to be based on our intuition especially with the deterioration in game quality and increase in sweat.

My solution has been to stick to a RoR that allows me to play confidently even at the expense of EV as I got wiped out years ago which I didn't truly understand properly at the time.

I will probably never be a black chip player but I will always be in the game. And since my comeback I have been able to double my betting stakes as I've achieved success over time. I started at a 1% RoR and I've been gradually working towards quarter

As my BR has slowly grown I've been able to further reduce my risk and increase my win rate. The other benefit is more and more games become playable. When I started I was very limited in my game selection. Now I can play most games as long as I wong out appropriately.

Thank You!

I decided to return to playing BJ after a 15 yr absence after going broke due to over betting. When I came back I studied RoR and the mathematics you put into BJA. You even helped me through some of the calculations. I came back with an anal focus on making sure my betting , BR, and RoR were appropriate since a lack of understanding in the nineties cost me fifteen years.

Admirable that you learned from your mistakes and righted the ship. Good for you!

You were one of the first to focus on the critical aspect of RoR. Most of the authors before you made only passing and incomplete references to it.

One of the nice innovations was the "bumping into the barrier" concept that I got from the world of options and the formulas that pertain. No one before had ever considered the idea from anything other than the end-point viewpoint. And, as you know, that was the wrong way to do it.

Counting is easy. The hardest parts are getting the hours in and managing risk properly. Norm's software make the latter easy.

Thank you for your huge contributions!

You're welcome. And thanks for the acknowledgment.

I still struggle with the knowledge that based on getting in a maximum of one N0 I have a one in six chance of LOSING over an entire year. But I can't do much to change that. It is what it is. And I've had those stretches.

It's the unfortunate nature of the enterprise. Solo counting is so extremely hard for that very reason. You take all the swings by yourself, and the long run seems to take forever. Play too much and you risk both overexposure and burnout from the constant travel and trying to stay under the radar. Play too little and, well, you've covered that above. Finally, try the team approach, and, well, you've covered that below! :-)

I was also the unfortunate member of a team for a relatively short time that had miserable results and lost the confidence of my team mates as a result. You covered this exact scenario in BJA as well under the team play section.

Some personal experience went into that chapter! You are not alone, my friend.

I have learned that over the long run the math works but the long run is much longer than most of us would expect it to be based on our intuition especially with the deterioration in game quality and increase in sweat.

Yes, exactly right.

My solution has been to stick to a RoR that allows me to play confidently even at the expense of EV as I got wiped out years ago which I didn't truly understand properly at the time.

I will probably never be a black chip player but I will always be in the game. And since my comeback I have been able to double my betting stakes as I've achieved success over time. I started at a 1% RoR and I've been gradually working towards quarter [chips?].

Patience is a virtue. Discipline too.

As my BR has slowly grown I've been able to further reduce my risk and increase my win rate. The other benefit is more and more games become playable. When I started I was very limited in my game selection. Now I can play most games as long as I wong out appropriately.

It sounds like you've found a great approach that works for you. Stick with the game plan!

Thank You!

The pleasure is all mine.

Good luck to you, and Happy New Year.

Don

When you discussed "bumping into the barrier" in BJA, you calculated the SD for 100 coin tosses as SQRT 100 = 10. I also think I remember a math prof post making the same calculation. In math texts, I've seen the calculation SQRT (100*0.5*0.5)= 5

Why is there a difference?

which = 100 * 0.5 * 0.5 = 25

Stand Dev = Sqrt (25) = 5

The 10 = 2 Std Dev, which is often used in the Normal table. +/- 2 Std Dev constitute about 95% of the outcomes.

The sqrt(npq) formula applies in a different way to this discussion.

If I toss a coin 100 times, the mean is 50 and one s.d. in either direction is five. So, we can say (roughly: see below)) that a one-s.d. band around the mean is 45-55. But in my example, we're netting the results! If I get 45 heads (one s.d. according to the above method), I get 55 tails, and I am NET 10 paces to the left of where i started!

So the barrier example requires somewhat different thinking. Here, at the start, I'm standing ten paces to the right of the abyss. The calculation for a one-s.d. move in either direction, just as in blackjack, is simply sqrt(number of tosses/hands). At the END of 100 tosses, as explained on p. 123 (and above), I have a 16% (actually, more) chance of teetering on the edge. That's a one s.d. move to the left. And, the point is that, if I count the chances of falling at any time throughout the entire experiment, the probability is double, or 32%.

Now, actually, it's even worse than the above. The normal distribution works as a good estimate of binomial probabilities for large n. As it turns out, 100 isn't all that large. And, in the above example, the actual binomial probability turns out to be 18.4%; so my chances of becoming part of the landscape are a whopping 36.8%!

I knew that when I wrote the article, but I didn't want to confuse the issue and simply wanted to introduce the barrier concept.

Hope this helps.

Don

That passage in your book with the "mortal remains floating in the Colorado River" is quite a piece of literature and it explains the barrier syndrom with the resulting doubling of probabilities in the most poignant way.
As many players still underestimate the risk of ruin, I recommand everybody to (re)read that chapter on ROR. Even if the subject is difficult it's a fun read.

Francis Salmon

After already being way in the hole, I get wiped out in one shoe.

Dealer tells me "I've never seen anything like that before."

http://bj21.com/boards/free/free_board/index.cgi?noframes;read=136487

The question: If you were to assign an exact percentage of the chance of drawing a key card or from a grouping of key cards that is a deviation of basic strategy what would that percentage be? For example, you have a 13vs2. Basic strategy is of course to stand but HiLo I-18 calls for hitting at TC-1. Obviously if you have a large enough percentage of {6-8} available you would hit. With HiLo and other counting methods that assign a tag value to follow a single point on a number line you can have TC1 or 2 and still have a high enough ratio of cards available in this {6-8} grouping in the remainder of the deck to hit the 13 regardless of what the dealers upcard is! If you have a 52% chance of drawing a {6-9} on a 12 are you going to stay on the 12 because the dealer is showing a 6? What of this huge prevalence of this {6-9} grouping with regard to the dealer? There is also a 52% chance of his hole card being in this grouping, along with a similar chance of drawing another! These are the times when the HiLo player sees that had he hit he would have made a great hand and ultimately the dealer bangs out a 6-7-8, a 6-6-9, perhaps an 8-8-6, etc. By knowing the exact ratio of a potential key card or card grouping for a specific hand you can make more informed playing decisions.

Composition dependent play is to have more information and the ability to make instantaneous calculations with this added information means making better playing decisions. In this particular scenario, those neutral cards are not so neutral! Playing based on a single fixed point means the betting decision coincides with the playing decision. The playing decision for a given hand at a given count is always the same is always the same with these counts, for instance with a TC2 you always always always stand on 12vs3 and you always always always hit at less than TC2. With a system based on deck composition and ratios of card groupings, the playing decision can vary with the same TC because the TC is used for betting but the playing decision is based on deck composition rather than the TC. In other words, a 12vs3 at TC2 can have different playing decisions because the playing decision is based on availability of key cards and key card groups. You can have 12vs3 at TC2 but the ratio of {6-9} compared to other card groups gives a high probability of drawing what is a key card for this particular hand. All of this can be calculated to an exact percentage of drawing from a key card grouping for a specific hand in real time and without a computer, at least by me. I don't know of anyone else that does what I do is all. It's a very obscure and difficult counting method that requires memorizing a short stack of tables along with some serious speed math to carry out the calculations, so it's not for everyone. Composition dependent play can be done without a computer though... I am living proof of that.

Blackjack has been very good to me over the years and I never would have even gotten off the ground way back when without the information I derived from Peter Griffin, Don Schlesinger, Ken Uston, Stanford Wong, D.H.Mitchell and Arnold Snyder. These are the people who gave me the groundwork to evolve to what I am today. Never met any of them but I've had a lot of great years because of these great authors and their contributions to the blackjack community.

Here's one. Every casino I patronized yesterday saw every dealer consistently ending up with a total of 20 or 21. Lost 8% of the bank in one day of lousy flux.

Very nice description/report/analysis Ivan.
Do you or would you happen to know if any on this board or in general, are there any players who know about and use the key card concept?

For any that don't know about it there is a nice explanation by David Sklansky in "Getting The Best Of It".

Robbie

I agree that all the books you mention are superb and I found another book extremely informative, "The Worlds Greatest Blackjack Book" by Lance Humble and Cooper. That and one of Uston's books were the best for me when I knew nothing about blackjack. Those were good until I learned a complete system from the "Silver Fox", Ralph Stricker.

Robbie