A small dilema.

Me BR1 (1402 chips) acting first on the last hand and my opponent BR2 (1400chips). I have made the minimum bet 100 and my first two cards was totaling 11. My opponent made a bet of 175 (I think he should bet at least 203 ?) and he got a soft 14. Dealer was showing a 10 up card.

I was unsure if to double my 11 or to surrender it and leave my opponent to work out his soft 14 to at least a push against the dealers 10. Ultimately I have decided to surrender but I still dont know if it was the correct play or not at all ?

 

Why Not Double For Less?

How about putting out $75 and doubling for less?

 

PlayHunter plays at blackjack21.com (or a skin thereof). Doubling for less is not allowed.

(As an aside, it occurs to me that a useful feature in the forum would be the ability to define the various rule sets somewhere, and then link to them from posts rather than keep repeating them.)

 

Well....... first of all, I'd have blasted away almost all-in - bet the 1400 and keep the 2 back. Since your lead is less than the min bet, by betting the min it allowed your opponent options that his 175 bet brought into play. He can double and take the high from you even if you get three bets on the table - and he can surrender to take the low if he sees you standing weak. That bet takes all those options away, all he could do is take the low with a bet structured so that he blows you away with a natural but allow him to surrender if you're weak. 1250 would do that.

As far as your eventual situation, I'd have doubled that nifty 11. Forces him to double for the high.

 

Yes, London Colin is right this situation happened at a skin of Blackjack21.

Minimum bet is 100, and maximum bet is 1000. - So no, I could not bet 1400 at once. (but anyway, I think a push/win+lose/push scenario it worth about 7% for my opponent and a minimum bet is still the better option, is not so ?) - Double for less is also not allowed unless the remaining un-bet chips are less than the original bet. (for example if I were max betting 1000, then I could do a double for the remaining 402 chips going all in, but not less than that)

And by doubling my 11, I have figured that I was going to force my opponent to double only when my double totals 20 and 21 which is about 39%, but for any other scenario, he could play out his hand til he will equal my total if i draw a total of 17-18-19, or surrender if my double was drawing stiff.

But the real questions on this issue are: 1. what success rate have a soft 14 versus a dealer 10 (where dealer already checked for blackjack) ? and 2. should my opponent have bet at least 203 or a maximum of 195 so he can take the surrender low if my ending hand result would be underdog to dealer?

 

perfect bets

PlayHunter,
You and your opponent made perfect bets.

With your lead of 2 points and betting 100 minimum his response by any number between 153 and 195 have the same chances. A bet that looks ” rounded” has additional value because it doesn’t make obvious that it overcomes opponent by “1” point for some specific outcome. 175 was the most “rounded” from the range.
His (bot?) bet have all of these advantages:
Surrender bets your losing
Single win beats your win
Blackjack beats your double
Double bet beats your triple

If you bet big he can take low with benefits of double.
Even if you could bet almost all-in your opponent almost certainly would take low but bet high enough so his bj would beat your win. He also could play one point higher than your hand to take away some of your pushes. Your chances would be somewhere around 47%.
By taking low you have to surrender most of the hands giving away some of the hands that using basic strategy would cause win/push-push but that still would get you around 50%.

Once you surrendered your hand your opponent has to win/push A4 vs. dealer’s Ten, which makes him winning his hand 36% and pushing 10%
If you double he should hit to at least 17 and one point higher for your totals lesser than 20, otherwise double himself for a total chance of beating you 55%.

S. Yama

 

Yeah, not knowing the 1,000 max bet limit certainly shot a hole in my theory, didn't it?
Betting the minimum would be the call with that limit in effect.
Forces opponent to win, or push to your loss.

Mea culpa, but didn't know one rather important fact at the time.

 

Sure it is my fault too for the confusion, I should made clear the betting limits and rules regarding double for less from the start in my first post. Thank you all !

 

According to Yama's post above, betting the min is still the way to go, even if you can bet 1400.

 

Can I ask how you obtained these numbers? As PlayHunter said, they are the crux of the matter when deciding whether surrender is better than doubling. (The figures for doubling being available in Ken's tables.)

One of the many tasks I have on a to-do list that will most likely never get done is to modify a CA program so that it outputs probabilities of winning, pushing, etc., rather than EVs. Does such a thing already exist?

 

I have also wondered what your win/push % is for starting hands vrs dealer up card.

I have never been able to find a table of them.

I would imagine that Gronbog has probably generated a table of them in some of his past sims.

Larry

 

Hi all,

Yes my sims can generate probabilities of success for many kinds of goals. If you have specific questions you want answered then please fire away. Answers will have to wait until next weekend because I'm on vacation (only online briefly because my wife wanted to check her email.

 

I don't really have a specific hand for you to look at. The situation in the thread where surrender is available make for some interesting decisions. Normally when you take the low playing first, your opponent has no choice but take the high. You could then play the hand per Wong's table 5. However, if both bet correctly, you play the hand, and surrender would allow the opponent to take away the low, then you have to think about the situation.

It is probably unlikely that you could afford to bust, since your opponent could then surrender for the win.

The whole situation should probably be looked at in three scenarios. Dealer has 2-6, 7-10 and Ace up. These scenarios have a significant break in dealer bust percentages. Depending on which group dealer is in makes the opponents win/push percentage on his hand the determining factor on whether or not you should surrender to keep the low. I wouldn't be surprised that if dealer has 7-A, then you would be better off surrendering almost all hands.

Larry

 

Sim Results

OK, I've had a chance to read this thread carefully and have some sim results for some of the questions asked.

Looks like you made the correct choice.

Percent Success Rate:
Surrender: 51.66%,
Hit: 43.95%,
Double: 43.58%

Note that doubling was your worst choice, but not by much behind hitting.

These results are for an optimal startegy player playing behind you, but they hold for both a basic strategy player or a skilled player following you if you surrender, since BR2's optimal strategy, once you surrender, is the same as basic strategy.

As noted above, basic strategy is optiimal for your opponent, once you have surrendered. Vs the dealer's 10, the results for soft 14, after the dealer has checked for blackjack are:

Wins 37.79% Losses 51.54% Pushes 10.67%

I am always impressed at the accuracy of S. Yama's calculations and this time is no different.

For win/push of soft 14 vs 10, his results are very close to my sim results (above).
His optimal strategy for BR2 should BR1 double also agrees with my sim result and his success rate is close to my result of 100 - 43.58% = 56.42% (see above)

 

With both players playing optimal strategy, probability of success for your opponent with a bet of 175, before any cards are dealt is 48.9% With a bet of 203, that improves to 49.08%.

The problem with this kind of computation/simulation is that no one knows the complete optimal strategy for these situations, nor would most tournament players stick to basic strategy on the last hand. Exact probabilities are likely impossible to obtain. When the results are this close, I am hesitant to declare one bet to be better than the other.

 

Yes, each time. This is amaizing indeed because since Mr. S. Yama does all these estimations instantly without any computer programs he must have a damn great native sense of math and probabilities. (wish I had this ability too)

 

I think it would be helpful if we can know the win and push percentages of the following soft hands: AA (when no split), A2, A3, A4, A5, A6 each against dealer 7, 8, 9, A and 10 (after dealer peeks) (Thanks to you, now we got A3 vs dealer 10 :cheers

 

Luckily, my sim can generate these kinds of tables. I generated three sets of tables which are relevant to this situation:

1. 6 deck basic strategy outcomes
   http://gronbog.org/results/blackjack/strategy/bjt.com/handResult.basic.6.noBJ.txt
2. 6 deck basic strategy outcomes with no splitting (to get the result for A,A played out)
   http://gronbog.org/results/blackjack/strategy/bjt.com/handResult.basic.6.noBJ.noSplit.txt
3. 6 deck win/push goal outcomes (since that is the goal of our opponent should we surrender)
   http://gronbog.org/results/blackjack/strategy/bjt.com/handResult.mustNotLose.6.noBJ.txt

Each set of tables contains outcomes (if appropriate) for:

• Initial Hard Hands
• Initial Soft Hands
• Hard Double Downs
• Soft Double Downs
• Individual Split Hands
• Overall Split Results

All results assume that the dealer has checked and does not have black jack. Each cell indicates (Win%/Lose%/Push%).

I hope these provide the information you need to crunch the numbers for yourself!

 

These results are more than helpful for us, and a brilliant work indeed.

A very BIG and warm Thank You ! from all of us Mr. Gronbog :cheers:

 

Updated Tables

Thanks for BlueLight for contacting me to point out an apparent problem with the

http://gronbog.org/results/blackjack/strategy/bjt.com/handResult.mustNotLose.6.noBJ.txt

table. The entries for hard double of 4 vs Ace showed a 100% success rate. In reviewing the table in more detail, doubling 4 vs 9 showed a similar result. At first I thought he had found a bug in my software, however I am happy to report that there is no bug.

It turns out that for the "must not lose" strategy, doubling of a hard 4 is only recommended in a couple of very specific multiple-split situations which occur very infrequently. Doubling hard 4 vs Ace, for example, is recommended for hands like

17
19 doubled
17
4
vs Ace

which, during the simulation, only occurred about once in every 182,000,000 hands. The simulation which generated the original table was 350,000,000 hands and it turns out that the double of hard 4 occurred twice and was successful in rescuing the entire 4-hand result both times.

I have regenerated the tables using simulations of 2 billion hands each in order to obtain more accurate results. Even so, the double of hard 4 vs Ace only occurred 11 times during that simulation.