Long term profit is seeing the big picture

[tthree]

That's not so. I can write a program to do a full combinatorial analysis of a shoe before every bet and before every play, and it will still have nightmare shoes. Almost as many as you or I do. Sorry to tell you but extra work is usually a waste of time, especially if you expect it to turn bad shoes into good ones. It is a dismal thought but it is also reality.

I was not talking about turning bad shoes into good. I was talking about damage control. Turning a max bet loser or 2 into winners has a huge affect on that shoes bottom line. Those are your big downswings being made into smaller downswings. You might still have the nightmare but it is easier to go back to sleep.

 

[FrankieT]

Multiple hands reduce variance.
Always insure blackjack at all but minimum bet.
Realize insuring good hands at small insurance disadvantage reduces variance .

I never thought about insuring good hands and blackjacks at a small disadvantage to reduce variance. I thought of only doing it as camoflauge so when I do it on big bets it looks less conspicuous.

Which hands do you feel are worth insuring at a disadvantage:
20, 19, 11, and/or 10????

And at how much of a disadvantage???

What other top plays that you know of that reduce a great amount of variance with little EV loss?

EDIT: I just found this useful thread - http://www.blackjackinfo.com/bb/showthread.php?t=11175&highlight=variance+reducing+plays

 

[tthree]

I never thought about insuring good hands and blackjacks at a small disadvantage to reduce variance. I thought of only doing it as camoflauge so when I do it on big bets it looks less conspicuous, but obviously the fact that it greatly reduces variance for little loss to EV is also a big bonus. Of course I never put a ton of thought into variance reducing plays period.

Which hands do you feel are worth insuring at a disadvantage:
20, 19, 11, and/or 10????

And at how much of a disadvantage???

What other top plays that you know of that reduce a great amount of variance with little EV loss?

I usually stick to blackjack and 20. The others are good starts but lose to often. Variance is reduced because you usually win one bet or the other. When the chance of losing both gets to high the variance reducing affect quickly disappears.

Blackjacks I insure all but my base bets. I go to 1 TC below the index on 20. If your goal is cover you probably want to be more generous.

Look at the zero point cards for your count. Hand matchups where they are the key card(s) will have a small increase in EV as the index is exceeded. This is because your count is a poor indicator of the deck compositions affect on the hand matchup. By giving up the index play until an advantage threshold is reached it doesn't cost much and decreases variance a lot. This is especially true if it is a doubling or splitting index as the money you put at risk for a small gain in EV is doubled.

 

[psyduck]

That's not so. I can write a program to do a full combinatorial analysis of a shoe before every bet and before every play, and it will still have nightmare shoes.

I know combinatorial chemistry in terms of building a library, but what is exactly a combinatorial analysis on BJ? You mean all possible combinations?

 

[MangoJ]

I know combinatorial chemistry in terms of building a library, but what is exactly a combinatorial analysis on BJ? You mean all possible combinations?

Combinatorial analysis is a simple algorithm that is able to calculate exact expectation values, essentially by calculating all possible outcomes.

Much like you would calculate the probability of throwing a sum of 12 with 3 dices, you would write down all combinations, and weight them with their corresponding probability.

 

[psyduck]

Combinatorial analysis is a simple algorithm that is able to calculate exact expectation values, essentially by calculating all possible outcomes.

Much like you would calculate the probability of throwing a sum of 12 with 3 dices, you would write down all combinations, and weight them with their corresponding probability.

Thanks Mango.

 

[Automatic Monkey]

I was not talking about turning bad shoes into good. I was talking about damage control. Turning a max bet loser or 2 into winners has a huge affect on that shoes bottom line. Those are your big downswings being made into smaller downswings. You might still have the nightmare but it is easier to go back to sleep.

What makes you think a stronger count will allow you to do that once or twice in a shoe? The "right" (made with more information) play will sometimes result in a loss while the "wrong" one will result in a win. Look at all the times we see a ploppy clean out a rack while we are getting killed. The superiority of the decisions we make with exceptionally strong counts versus ordinary counts like High-Low is so small we can expect to turn a winner into a loser maybe once a week, not once a shoe. The difference is so subtle we can't perceive it, even though we swear we can.

 

[blackjack avenger]

SCORE and NO

Hm, well right now I play at $15 tables and at this point in my career my max bet is limited by my bankroll more than anything. So given I hold my min bet constant, my RoR is dependent on my spread. This is only really true for my personal situation because I can't play at lower minimums. But if I were able to, then yes you guys would be correct. Agreed?

high SCORE good
low N0 good
These you want to raise
Lowering them does not serve you.

Ex. In a play all 1 to 4 spread better then 1 to 3.

 

[blackjack avenger]

risk averse insurance

Or variance reducing means to take insurance less frequently, not more frequently.

 

[assume_R]

high SCORE good
low N0 good
These you want to raise
Lowering them does not serve you.

Ex. In a play all 1 to 4 spread better then 1 to 3.

Okay, here is some actual output from a simulation:

Spread 1
Spread = 1:5
Win Rate = $30 / hour
Std = $490 / hour
SCORE = 54
N0 = 18,000
RoR = 17%

Spread 2
Spread = 1:10
Win Rate = $49 / hour
Std = $710 / hour
SCORE = 71
N0 = 14,000
RoR = 25%

Now, normally I would go with Spread 1 because of the significantly lower RoR. You would suggest I would go with Spread 2? Wouldn't you be worried that there's a higher chance of going bust?

 

[blackjack avenger]

not quite

Okay, here is some actual output from a simulation:

Spread 1
Spread = 1:5
Win Rate = $30 / hour
Std = $490 / hour
SCORE = 54
N0 = 18,000
RoR = 17%

Spread 2
Spread = 1:10
Win Rate = $49 / hour
Std = $710 / hour
SCORE = 71
N0 = 14,000
RoR = 25%

Now, normally I would go with Spread 1 because of the significantly lower RoR. You would suggest I would go with Spread 2? Wouldn't you be worried that there's a higher chance of going bust?

If u use the same ror the higher spread will outperform. These ror's are both way to high whether fixed or resizing.

 

[tthree]

What makes you think a stronger count will allow you to do that once or twice in a shoe? The "right" (made with more information) play will sometimes result in a loss while the "wrong" one will result in a win. Look at all the times we see a ploppy clean out a rack while we are getting killed. The superiority of the decisions we make with exceptionally strong counts versus ordinary counts like High-Low is so small we can expect to turn a winner into a loser maybe once a week, not once a shoe. The difference is so subtle we can't perceive it, even though we swear we can.

You are forgetting adding the block side count and using the fact that you are using it to bet more when it will allow a better decision on your stiffs. Like I said I am not simply calling a stronger count as a higher level versus a lower level but your side counts and playing adjustments as well. While weak info can often result in a less accurate decision resulting in a win, I am not going to start playing basic strategy rather than counting. We are talking about long term profits and the big picture. I just know how often I see it make a difference. Usually when things are going really bad at a high count the dealer is getting seven thru ace up cards and we are playing stiffs. Choosing a game with late surrender and the information on your main counts zero point cards vastly improves your decisions in many of these situations. Not to mention that the surplus or deficit of these cards tell you the exact deck composition not just the ratio of high to low cards.

Since you mention HILO many hand match ups the ace acts as a low card but is reflected in the count as a high card. This results in bad decisions that lowers your win percentage for the hand match up especially if the high count is due to a lot of surplus aces rather than surplus tens. You want to not only get the big bets out at the right time but win the highest percentage of big bets you can. An ace side count would flag the strength or weakness of your indices in those hand match ups and adjust your TC accordingly for the decision.

I don't know why anyone would be satisfied with simply getting the big bets out at the right time. You should demand of yourself that you win the highest percentage of the big bets that you can. This is simple common sense. The people I hear arguing otherwise just don't want to put in the effort and give back twice their max bet or more if the decision is doubling or splitting for each win subtracted to the ratio of wins to total bets made for each situation. Wouldn't you want to add those wins. Some people might not be skilled enough to do this but practice makes you develop skills. You can burn the best games by trying to increase your bets/spread beyond casino tolerance or you can bring your strongest complete counting (not just your basic count) game to the casino and play within their tolerance. Then you won't be one of those whining about not being able to play any great games. You will be playing them.

 

[assume_R]

If u use the same ror the higher spread will outperform.

Okay, agreed, but I was pointing out that to keep my RoR to a reasonable level I have to lower my spread.

These ror's are both way to high whether fixed or resizing.

So here's another example of 2 choices:

Spread 1:
1x1 to 2x3 units
RoR = 9%
SCORE = 27
N0 = 37,000

Spread 2:
1x1 to 2x7 units
RoR = 25%
SCORE = 35
N0 = 28,000

These are pretty crappy conditions, but my point is that I want to keep my RoR under 10%. But to do that, given the table minimums, I can't spread as high as I'd like. Given the spread I'd want at this game, it would bring my RoR way too high.

 

[tthree]

high SCORE good
low N0 good
These you want to raise
Lowering them does not serve you.

Ex. In a play all 1 to 4 spread better then 1 to 3.

Do you really think Assume_R, math genius (respect not sarcasm), is not going to know the big picture of RoR, win rate, SCORE, N0 etc and try to balance them for his best game. He is trying to find the optimal spread/top bet etc for his big picture, not be confused by focusing to much on any one part of the big picture. The big picture is more than just the math. It must include the constraints on the game imposed by the casino's tolerance levels and your bankroll and how good a game you find yourself having at your disposal. He, as well as most if not all here, has room to learn and improve but real world improvement goes beyond the straight math of the game. What good is playing with a slightly bigger edge on paper when the only games it will allow you to have at your disposal take more than that back or have undesirable variance levels (RoR) when you combine the larger spread/top bet with the crappier rules.

 

[blackjack avenger]

rethink everything

Okay, agreed, but I was pointing out that to keep my RoR to a reasonable level I have to lower my spread.



So here's another example of 2 choices:

Spread 1:
1x1 to 2x3 units
RoR = 9%
SCORE = 27
N0 = 37,000

Spread 2:
1x1 to 2x7 units
RoR = 25%
SCORE = 35
N0 = 28,000

These are pretty crappy conditions, but my point is that I want to keep my RoR under 10%. But to do that, given the table minimums, I can't spread as high as I'd like. Given the spread I'd want at this game, it would bring my RoR way too high.

U don't spread as high as u like, u spread optimally or less due to heat.

I think in ur examples u set the first, then just raise the spread? This is wrong. The ror needs to be the same.

 

[tthree]

U don't spread as high as u like, u spread optimally or less due to heat.

I think in ur examples u set the first, then just raise the spread? This is wrong. The ror needs to be the same.

Maybe you could show him an example using his first spread as a basis and including TC/advantage criterion for each bet. I assume that is your basis for keeping the RoR the same.

 

[assume_R]

I think in ur examples u set the first, then just raise the spread? This is wrong. The ror needs to be the same.

I'm not sure I'm following, avenger. With the spread and N0 I'd like, how can I make the RoR the same? The table minimums don't go much lower. And this is with wonging out. My bankroll is way too limited at this point in my life, I do realize that.

Edit: Just so you know, I'm not disagreeing with you or trying to minimize your knowledge. I honestly do want the <10% RoR with the low N0. I'm seriously asking you how do I achieve it.

 

[tthree]

Hopefully Avenger will chime in but I think he is expanding the range of TCs that the ramp covers with a higher spread and the same RoR. By making the higher bets at a bigger advantage threshold they carry less risk.

 

[assume_R]

Hopefully Avenger will chime in but I think he is expanding the range of TCs that the ramp covers with a higher spread and the same RoR. By making the higher bets at a bigger advantage threshold they carry less risk.

So you're saying instead of betting optimally (which may be to put a max bet out at +4), perhaps wait until +6 to put that max bet out if it keeps the RoR the same?

So for example a given RoR of 10% might have:
1x1, 2x1, 2x3, 2x5, 2x5, ...

as the optimal betting, and to keep that 10% RoR, it might be possible to do:
1x1, 2x1, 2x3, 2x5, 2x5, 2x5, 2x7 or something?

whereas the "optimal" spread (minimal N0) might have a RoR of 20% with:
1x1, 2x1, 2x3, 2x5, 2x7

 

[blackjack avenger]

how do i get into these situations

:laugh:

Assume R

Sim these
Your bank your game, one hand only
13.53% ror Wong positive hands
13.53% ror play all
13.53% your Wong out point

I think you need to see what optimal spreads look like with no bet constraints.