Soft doubles
- Thread starter newbctr
- Start date
Naturally worth less on a 6D (less TC flux) than a 4/2/1 deck.
However, I've never had to quantify their value:
Have you run a sim with your (or a typical...) betting ramp to check how much soft doubles are worth (%EV) ?
Of course, shrinks as your risk-aversion rises....
However, I've never had to quantify their value:
Have you run a sim with your (or a typical...) betting ramp to check how much soft doubles are worth (%EV) ?
Of course, shrinks as your risk-aversion rises....
kewljason said:
I dont use HILO but never double A2,A3,A4,A5 against a 2. Statistically it can be justified some of the time for your biggest bet in your ramp against a 3 but unless you have real deep pockets you are better off hitting. To much added variance for a very small gain in EV. A8 against a 5 or 6 is a good double for any bet larger than your minimum if you are ramping conservatively. If you use a risking ramp that ups your bet rather early your second raise in bet is when you would double A8 against a 5. Hard 8 requires a higher count against 5 or 6. Sorry I couldnt be more specific.
Gamblor
Well-Known Member
I play with the indices for:
A2 v 4-6
A3 v 4-6
A4 v 4,5
A5 v 4
A6 v 2,3
As kewl mentioned, purely EV wise its not that critical (no where near as important as illus. 18 and surrenders).
These are the indices that are close to zero, and in addition, I have a card eating fetish, so these are also the indices that will eat up extra cards at low counts and less cards at high counts (thats why I didn't mention A7 v 2 and A8 v 6, which is also close to 0 - these I use only on special occasions).
Also keep in mind this is just crazy Gamblor method, not really recommended in the literature.
A2 v 4-6
A3 v 4-6
A4 v 4,5
A5 v 4
A6 v 2,3
As kewl mentioned, purely EV wise its not that critical (no where near as important as illus. 18 and surrenders).
These are the indices that are close to zero, and in addition, I have a card eating fetish, so these are also the indices that will eat up extra cards at low counts and less cards at high counts (thats why I didn't mention A7 v 2 and A8 v 6, which is also close to 0 - these I use only on special occasions).
Also keep in mind this is just crazy Gamblor method, not really recommended in the literature.
Last edited:
If you highlighted the rest of my sentence you would see it is not statistically justified for low counts. It has a small increase in EV for very high counts but by taking these if you are optimally betting in the long run the resizing of your bets down due to the increase in variance actually costs you EV in the overall picture. Maximizing the hands EV doesnt always maximize total EV for your game (maximize long term profit IF YOU BET OPTIMALLY). Betting optimally is the highest EV you can have. This is called risk averse play. You are not giving up EV but gaining EV when you maximize your entire play not just each hand to hand match up.MangoJ said:
Okay, if I get you right you say that EV is positive on small counts, but the advantage is to small to justify an additional bet. That means that CEV (the log-utility equivalent certainty) is negative, and hence hitting is more healthy for bankroll. Then EV on larger counts is much larger, and CEV turns positive, thus doubling is the best choice.
It seems I just misunderstood you, when you said at "max betsize" it would be at higher counts.
It seems I just misunderstood you, when you said at "max betsize" it would be at higher counts.
Risk averse indices
Mangoj, I think I see the problem. My first sentence talks about dealer upcard of 2 and my second dealer upcard of 3. I think you are missing they are two different situations. I never said anything had a +EV at small counts. I am not sure were you got that idea from.
If you are unfamiliar with the risk averse strategy I think QFIT has a good explanation on his site. It is called "Risk-averse versus expectation maximizing indices". In his "How to create a blackjack strategy-card counting techniques" there is a subsection called "Index generation for card counting". In this subsection you find the "Risk-averse versus expectation maximizing indices" explanation. It is very brief but drives the point home. Maybe 10 or so indices need to have this adjustment to maximize long term profit as opposed to individual hand EV maximization.
QFIT, you can mail my commission to the usual address.:laugh:
Mangoj, I think I see the problem. My first sentence talks about dealer upcard of 2 and my second dealer upcard of 3. I think you are missing they are two different situations. I never said anything had a +EV at small counts. I am not sure were you got that idea from.
If you are unfamiliar with the risk averse strategy I think QFIT has a good explanation on his site. It is called "Risk-averse versus expectation maximizing indices". In his "How to create a blackjack strategy-card counting techniques" there is a subsection called "Index generation for card counting". In this subsection you find the "Risk-averse versus expectation maximizing indices" explanation. It is very brief but drives the point home. Maybe 10 or so indices need to have this adjustment to maximize long term profit as opposed to individual hand EV maximization.
QFIT, you can mail my commission to the usual address.:laugh:
Ferretnparrot
Well-Known Member
I use indexes for
A,9 vs 6,5,4
A,8 vs 6,5,4,3
A,7 vs 2
A,2 vs 4
A,3 vs 4
for Doubling
A,9 vs 6,5,4
A,8 vs 6,5,4,3
A,7 vs 2
A,2 vs 4
A,3 vs 4
for Doubling
Ferretnparrot,
Since you (& Jason) have two of the few posts in this thread that make much sense to me, can you expand and tell me your TC indices? I see jason's point... not much value is added. I know from experience that when I double a huge bet with A4, A5, A6 etc into a super high count, go figure, I get a 10 often; in this case, I am stuck hoping for a bust, although the chances are pretty good.
Since you (& Jason) have two of the few posts in this thread that make much sense to me, can you expand and tell me your TC indices? I see jason's point... not much value is added. I know from experience that when I double a huge bet with A4, A5, A6 etc into a super high count, go figure, I get a 10 often; in this case, I am stuck hoping for a bust, although the chances are pretty good.
I vehemently disagree with this statement. Here's just ONE example of a soft double - A6 vs. 5:kewljason said:
8 decks, S17, count is 0.
If you do not double your EV is 9.8%.
If you double, your EV is 9.6% of TWO bets. You gain 9.4% on the second bet. Why would someone
NOT want to put out another bet that earns 9.4%? MOST counters would - if they could - make their MAX bet for this kind of an advantage.
Obviously, as the count goes up, this figure goes even higher.
Methinks that TOO many people around here are TOO averse to risk. If you're THAT worried about risk, why not just stay home & never gamble at all?
Thats not a risk averse play. You double that on all counts.Sucker said:I vehemently disagree with this statement. Here's just ONE example of a soft double - A6 vs. 5:
8 decks, S17, count is 0.
If you do not double your EV is 9.8%.
If you double, your EV is 9.6% of TWO bets. You gain 9.4% on the second bet. Why would someone
NOT want to put out another bet that earns 9.4%? MOST counters would - if they could - make their MAX bet for this kind of an advantage.
Obviously, as the count goes up, this figure goes even higher.
Methinks that TOO many people around here are TOO averse to risk. If you're THAT worried about risk, why not just stay home & never gamble at all?
I was referring to the statement that soft doubles don't add much. This statement seems to imply that it's not really worth soft doubling at all, let alone learning the indices.
If you want to talk about the value of index plays; until the deck becomes VERY rich, index plays for ALL plays don't add much. Index plays for hitting & standing add very little. Index plays for splitting add little. As do index plays for HARD doubles. Even INSURANCE index plays don't add that much. When the count gets very high, index plays START to become much more significant, and this pretty much equally INCLUDES soft doubles.
It's when you put them all together that the cumulative effect becomes significant.
If you want to talk about the value of index plays; until the deck becomes VERY rich, index plays for ALL plays don't add much. Index plays for hitting & standing add very little. Index plays for splitting add little. As do index plays for HARD doubles. Even INSURANCE index plays don't add that much. When the count gets very high, index plays START to become much more significant, and this pretty much equally INCLUDES soft doubles.
It's when you put them all together that the cumulative effect becomes significant.
kewljason
Well-Known Member
:laugh:Sucker said:I vehemently disagree with this statement. Here's just ONE example of a soft double - A6 vs. 5:
8 decks, S17, count is 0.
If you do not double your EV is 9.8%.
If you double, your EV is 9.6% of TWO bets. You gain 9.4% on the second bet. Why would someone
NOT want to put out another bet that earns 9.4%? MOST counters would - if they could - make their MAX bet for this kind of an advantage.
Obviously, as the count goes up, this figure goes even higher.
Methinks that TOO many people around here are TOO averse to risk. If you're THAT worried about risk, why not just stay home & never gamble at all?
Are you seriously suggesting that I don't know basic strategy? Or that I am advising someone not make basic strategy plays?
Nobody is saying don't make basic strategy plays. The thread is about learning index plays for strategy variation, like A2 vs 4, A4 vs 3, A7 vs 2, A8 vs 5,6. And even then, I am not saying don't learn them. I am sure most of us have at least some soft double indices in our arsenal. But the truth is they don't add much value. If you learn just A8 vs 5,6 and maybe A7 vs 2, you would be capturing most of the small advantage of soft doubles index plays.
Of course not.kewljason said:
Yes. Maybe it was merely a poor choice of words on your part; but you did SAY "soft doubles don't add all that much value". This could easily be construed by a newbie as advising people to not make ANY soft doubles, or at LEAST telling them that it won't make much difference. There are already too many "variance chickens" around here; no sense giving them even MORE excuses to play scared.kewljason said:Or that I am advising someone not make basic strategy plays?
FLASH1296
Well-Known Member
MangoJ,
Kudos to tthree for his especially pragmatic feedback.
Here is a compendious hypothetical situation to ponder.
A Shakespearian choice — to Double or not to Double —
You know what Basic Strategy directs you to do —
"Nobility" has naught to do with the decision —
as your bankroll is deflating before your eyes.
Hitting, you know, has an expectation of X dollars on your initial wager.
Doubling, you also know, has an expectation of 1.05 X
The former decision "earns" you $1.00 on your $25 wager.
The latter decision "earns" you $1.05 on a doubled (now $50) wager.
With MANY 'units' to play, DOUBLING in order to MAXIMIZE your expectation,
even if the increment is minimal, is perfectly rational.
With FEW 'units' to play, HITTING in order to MINIMIZE your variance,
even if the sequelant reduction in risk is minimal, is preferred.
This is an analogue to "balancing" Offense with Defense.
Stretching your aggregate expectation is Offensive.
Reducing your variance is Defensive. N'est ce pas ?
Do yourself a boon and learn the approximate expectation on SOFT DOUBLES.
Those close decisions will be those that have indices of zero or thereabouts.
You need to learn that there are a few hands that are virtual "coin-flips",
e.g. Ace-Deuce vs. Five (A2 v 5) and Ace-Four vs. Four (A4 v 4).
Kudos to tthree for his especially pragmatic feedback.
Here is a compendious hypothetical situation to ponder.
A Shakespearian choice — to Double or not to Double —
You know what Basic Strategy directs you to do —
"Nobility" has naught to do with the decision —
as your bankroll is deflating before your eyes.
Hitting, you know, has an expectation of X dollars on your initial wager.
Doubling, you also know, has an expectation of 1.05 X
The former decision "earns" you $1.00 on your $25 wager.
The latter decision "earns" you $1.05 on a doubled (now $50) wager.
With MANY 'units' to play, DOUBLING in order to MAXIMIZE your expectation,
even if the increment is minimal, is perfectly rational.
With FEW 'units' to play, HITTING in order to MINIMIZE your variance,
even if the sequelant reduction in risk is minimal, is preferred.
This is an analogue to "balancing" Offense with Defense.
Stretching your aggregate expectation is Offensive.
Reducing your variance is Defensive. N'est ce pas ?
Do yourself a boon and learn the approximate expectation on SOFT DOUBLES.
Those close decisions will be those that have indices of zero or thereabouts.
You need to learn that there are a few hands that are virtual "coin-flips",
e.g. Ace-Deuce vs. Five (A2 v 5) and Ace-Four vs. Four (A4 v 4).
Last edited: