Case study: Splitting 66v3 or not @ RC7 Double deck

[Secretariat]

Works the way it always works. You set where to start and stop in the shoe, you set the number of cards of each value, and it millions or billions of hands for each index.

Thanks!

[k_c]

Thank yous so much KC.
I will study that in the next couple of days.

How long does it take to program and come up with indices like that?
Not sure if CV Data can do that.

Level 1 counts are fairly efficient. It takes 5 seconds or so to compute each set of data for A-5 using HiLo on my older laptop. Other hands could require more or less.

k_c

[Secretariat]

Level 1 counts are fairly efficient. It takes 5 seconds or so to compute each set of data for A-5 using HiLo on my older laptop. Other hands could require more or less.
k_c

KC, I am not done yet with A-5 and various situations but while working on this with your software, something else crossed my mind and it's EV and bet variation

Attached is a document at the RC0 level DD, H17, DAS that shows how many units I would bet at neutral counts based on my understanding of the numbers generated by your software.

It's surprising to see how the EV goes up (except at line 7) on whether the number 789s goes up or goes down although it goes up much faster as more 789s are played.

Do I interpret your numbers in a proper way and does the betting pattern look legit?

Please note that the bets and EV are on the second page as the file has been broken in two parts

Thanks again for your tremendous help KC.

Secretariat
BJ Secretariat KC 789 B Chart.pdf

[k_c]

KC, I am not done yet with A-5 and various situations but while working on this with your software, something else crossed my mind and it's EV and bet variation

Attached is a document at the RC0 level DD, H17, DAS that shows how many units I would bet at neutral counts based on my understanding of the numbers generated by your software.

It's surprising to see how the EV goes up (except at line 7) on whether the number 789s goes up or goes down although it goes up much faster as more 789s are played.

Do I interpret your numbers in a proper way and does the betting pattern look legit?

Please note that the bets and EV are on the second page as the file has been broken in two parts

Thanks again for your tremendous help KC.

Secretariat
BJ Secretariat KC 789 B Chart.pdf

All of your EVs are correct. However, what you have to realize is that these EVs are for using the best composition dependent strategy. (because compute mode was set to Best strat)

Let's take a look at a specific example for illustration -

shoe composition (2,3,4,5,6,7,8,9,T,A): {4,4,4,4,4,0,0,0,16,4}
COMPUTE MODE: Best strat
overall EV using best CD strategy: +2.795%

Leave the shoe composition as is and set compute mode to Basic strat (rather than Best strat)
shoe composition (2,3,4,5,6,7,8,9,T,A): {4,4,4,4,4,0,0,0,16,4}
COMPUTE MODE: Basic strat
overall EV using basic CD strategy: -1.005%

In order to realize EV of +2.795% you need to deviate from basic strategy

You can analyze which hands should deviate from basic strategy. While leaving compute mode set to Basic strat using the same shoe composition you can input and compute possible player hands. This will show which hands are played sub-optimally by using basic strategy rather than best strategy. For example, input and compute a player hand of T-6. You will find that the strategy versus up cards of 3,4,5,6 is Stand*. The asterisk is a visual cue that although basic strategy is Stand, it is not the best strategy.

Although an EV of +2.795% is possible in the above example, it may not be realistically attainable because a person may not be able to learn and implement the necessary strategies. In order to play successfully a blackjack player needs to honestly assess his capabilities.

k_c

[Freightman]

Secretariat said

It's surprising to see how the EV goes up (except at line 7) on whether the number 789s goes up or goes down although it goes up much faster as more 789s are played.

Sure sounds similar to QTC (Quality of True Count)

[Secretariat]

All of your EVs are correct. However, what you have to realize is that these EVs are for using the best composition dependent strategy. (because compute mode was set to Best strat) In order to realize EV of +2.795% you need to deviate from basic strategy Although an EV of +2.795% is possible in the above example, it may not be realistically attainable because a person may not be able to learn and implement the necessary strategies. In order to play successfully a blackjack player needs to honestly assess his capabilities.
k_c

This is awesome KC. Thanks for taking me to DisneyBJland and I'll have lots of fun there. I now have a more precise perspective on perfect betting".

My overall quest on this forum has been geared toward perfect card play and this comes down to the question that I asked before in a couple of other threads, which is: 1) how close to perfect card play can a human being get? In this particular situation, how close to a Max EV @ 2.795% can a human being get?

On a more general scope, 2) can perfect play generate 0%EV or better while flat betting?
How close to that can a human being get? Any data you can share?

Secretariat

[Secretariat]

Secretariat said


Sure sounds similar to QTC (Quality of True Count)

Nice to see you Freightman. HAHA, I was pretty sure you would show up at some point. I have been thinking a lot about QTC but could not quite associate an EV number on it although I saw it could be beneficial. This confirms what you've been saying all along and I can see how powerful WH combined with QTC can be.

[Secretariat]

Secretariat said
Sure sounds similar to QTC (Quality of True Count)

After comparing Perfect Strategy EV with Basic Strategy EV on the deck compositions mentioned above
I must say that BS EV is way below perfect strategy in some situations.
KC's warning is important.

In fact, in all 9 deck compositions we should bet no more than one unit playing BS.

I know that you and I, Freightman, can do better dans BS, but how much better is the question?
To be precise how much better in the 9 deck compositions attached?

Of those 9 deck compositions, Freigthman, in which situation would you bet multiple units?

[Freightman]

Secretariat said

Of those 9 deck compositions, Freigthman, in which situation would you bet multiple units?

I’m not sure that I understand the chart. I want to. A good conversation would be required. In any event, my basic premise is
1. Using halves to calculate both RC and TC - the way it’s always done
2. Using proportions of 10’s v 6789 and 2345 to assign QTC
3. Using a 2 ramp betting system to determine if you are betting on the lower vs higher ramp.I kinda sorta limit bets below lower ramp level if composition is really really lousy

Regardless of TC, if 6789’s are > 10’s, you are never on the higher ramp.

Appears you’ve gone further to limit lower ramp to 1 unit under certain card compositions regardless of true count. I can see that. One thing to be cautious of is the higher frequency of 1 unit bets vs overall rounds which would tend to make you stand out more - especially in a 2 ramp system.

Im ahead of th3 game in any event, but I can’t compute the additional potential EV of your premise vs mine.

[Secretariat]

Secretariat said


I’m not sure that I understand the chart. I want to. A good conversation would be required. In any event, my basic premise is
1. Using halves to calculate both RC and TC - the way it’s always done
2. Using proportions of 10’s v 6789 and 2345 to assign QTC
3. Using a 2 ramp betting system to determine if you are betting on the lower vs higher ramp.I kinda sorta limit bets below lower ramp level if composition is really really lousy

Regardless of TC, if 6789’s are > 10’s, you are never on the higher ramp.

Appears you’ve gone further to limit lower ramp to 1 unit under certain card compositions regardless of true count. I can see that. One thing to be cautious of is the higher frequency of 1 unit bets vs overall rounds which would tend to make you stand out more - especially in a 2 ramp system.

Im ahead of th3 game in any event, but I can’t compute the additional potential EV of your premise vs mine.

That would be nice to discuss it but prior to that, I think we need KC's help.
OK, perfect card play, or humanly optimal card play is one thing. Betting is another one.

Regardless of card play, here's a way to look at the situation with a deck starting at 44444444-16-4 which always yields a RC or TC of 0

The first column is the sum of the odds of getting a blackjack, a two-card 20, plus the odds of dealer break Let's call it the Trifecta friendly condition. As we changed the number of 789s (second column) the combined odds of BJ, TT and DB (dealer bust) change.

Of course, there's no need to adjust our playing strategy when we receive BJs or TTs but can we derive a winning betting strategy at RC 0 based solely on the number of 789s. This seems to be what you are doing Freight, but the BS strategy doesn't yield a profit according to KC's tool. KC we need your input on this.

Trifecta condition Number of 789s (down first and up below)

0,4469	12
0,4706	9
0,4996	6
0,5347	3
0,5769	0
0,4469	12
0,4276	15
0,4120	18
0,3993	21
0,3893	24

[lij45o6]

@Secretariat: If you are interested in studying Blackjack with regards to perfect play, consider this book.


The appendices are a bit maths heavy, but the general English within the chapters can get you through. Try to read and re-read as much as possible.

[Secretariat]

@Secretariat: If you are interested in studying Blackjack with regards to perfect play, consider this book.

The appendices are a bit maths heavy, but the general English within the chapters can get you through. Try to read and re-read as much as possible.

Thanks Dogman. Awesome book indeed. Makes me regret my daydreaming in math classes. HAHA. I have it and I refer to it quite often. Mr. Griffin really had a solid grasp of the game.

[Secretariat]

The first reply was based on 2D, H17, DAS at 44444444-16-4 and various 789s ratios
Things change a bit at S17.

Actually, even when the BJstrat mode (versus perfect strategy mode) is set at BS, the EV is positive when things are perfectly balanced.
That puzzles me somewhat. KC???

EV goes up with only 9x 789s and even more with only 6x789s.
EV goes down with BS when the number of 789s goes up.

0,1318	12
0,2279	9
0,2186	6
-0,0023	3
-0,642	0
0,1318	12
-0,0166	15
-0,192	18
-0,3832	21
-0,5864	24