I am on holiday and I'm playing in a place where late surrender is allowed but since I've never faced this rule before, I don't know how to handle it. Does anybody know where can I get the charts for late surrender according to the true count (not only basic strategy, of course)?
If you are going to play a little Surrender, you might consider simply using the following Counters Basic Strategy. It is very simple:
Surrender 15, 16 vs. 9,10, Ace
This is Not Basic Strategy. You don�t surrender 15v9 in BS, and not 15vA in S17. But these plays are correct when you get a little count, and would be superior than using just BS.
If you are going to play lots of Surrender, then you should lean some index numbers. But the CBS will get you most of the gain.
You did not mention whether your games was H17 or S17. There are different index numbers for each. The CBS I gave you is a reasonable overall compromise.
Counter Basic Strategy (CBS) is affected by the bet spread and penetration, in addition to the rules. According to the Blackjack 6-7-8 software Version 2003 (available thru this site), which can develop CBS tables, for Late Surrender, 6 decks, hit 15 v. 9. This is based on a 16 to 1 bet spread and 83% penetration. CBS changes little with penetrations of 67 - 83%, but CBS tends to be quite sensitive to the bet spread. For a bet spread other than 16 to 1, CBS might indicate you should surrender 15 v. 9.
For MathProf's other surrender recommendations, I agree.
-Hal
If I'm sure that nobody is watching I'll sometimes hit 15 v 9 in a negative count. But given the fact that most casinos verbally call "surrender" it is best not to vary it that much. Almost the only surrender variation I use much is 14 v 10 @ +3.
With an exit strategy of -1 or -2 I'm sure we'd be much more likely to surrender as the negative counts would be a smaller segment of the distribution. It would also be interesting to go a step further and have RA/CBS for surrender. Without precise indeces being run, I use this as my rationale for surrendering 8,8 v 10 almost always.
I did 2 quick Sims with SBA. 5/6 of dealt. S17-LS-DAS. I used an optimal 16-1 play-all spread, as determined by BJRM: 3,7,10,13,16 at 1-5 true counts respectively.
One with Always surrender 15v0; the other with Never Surrender
Always Surrender:
Initial bet advantage: 1.31883%, St. Err. 0.018783%
Total bets advantage: 1.15050%
SCORE: 49.34
Estim. payoff per 100,000 rounds played is 3756.60, with estim. std. 1691.28
Average st. dev. per round: 5.34830
Av. std. per round per unit: 1.11601
Average bet per round: 2.84842
Initial bet advantage: 1.30067%, St. Err. 0.018812%
Total bets advantage: 1.13485%
SCORE: 47.84
Estim. payoff per 100,000 rounds played is 3698.75, with estim. std. 1691.06
Average st. dev. per round: 5.34759
Av. std. per round per unit: 1.11717
Average bet per round: 2.84373
I decided to repeat the Sims, with 1 Billion Rounds for each. Always Surrender is still marginally ahead of Never Surrender, but the differences have gotten much closer:
Never Surrender- 1 Billion
Initial bet advantage: 1.32579%, St. Err. 0.005949%
Total bets advantage: 1.15666%
SCORE: 49.71
Estim. payoff per 100,000 rounds played is 3772.13, with estim. std. 1691.94
Average st. dev. per round: 5.35039
Av. std. per round per unit: 1.11746
Always Surrender- 1 Billion
Initial bet advantage: 1.32519%, St. Err. 0.005940%
Total bets advantage: 1.15616%
SCORE: 49.81
Estim. payoff per 100,000 rounds played is 3771.34, with estim. std. 1689.75
Average st. dev. per round: 5.34346
Av. std. per round per unit: 1.11584
Average bet per round: 2.84588
It really looks like it doesn�t matter much here what you do.
However, a play-all spread for Shoe play is very weak, even at 16-1. A Wonging spread of 1-6 is stronger. In fact, a 1-6 spread, sitting out at -1 is stronger.
I am fairly confident that with more realistic spreads, always Surrender 15v9 would be clearly better than Never Surrender.
...are contained in my Flashcard spreadsheet, available at bjfonline.com (Free Software group on the message boards) or the CCCafe (Files section). You will need Microsoft Excel 97 or later.
The file doesn't work. I can download it but then I cannot open it with winzip. I forgot to say that the game is six decks, duble on any two cards, DAS and dealer stays soft 17. Penetration is about 70-85%.
The new file can be downloaded, but WINZIP will not open it.
Congrats on the work though. The v.1 file was really slick.
15 v. 9 has a surrender index of +2.2. So instead of worrying about spreads you just do some mental calculation and act accordingly.
Francis Salmon
With Wonging out at certain negative counts, it's certainly plausible that the CBS decision would be surrender 15 v. 9.
For 8,8 v. 10, even with play all, the CBS decision is surrender.
-Hal
how the true count is calculated? In the Blackjack 6-7-8 software Version 2003, which contains the facility for developing CBS, the true count can only be computed with rounding to the nearest whole number. For example, for a true count of 0 in the software, the range of the exact true count is in the range (-0.5, 0.5).
If using flooring to round, a true count of 0 would mean that the exact true count would be in the range [0,1).
-Hal
It is possible that this could be a factor. Obviously, the spread is a factor. How you compute the True Count obviously affects the betting.
Why don't you mention the details of the spread you used. At what TCs were the different bets placed? This could help us compare the results.
BTW, I think Play-All spreads should not be used for Shoe play. Even Wonging Out at -3 (with smaller spreads) would be better. You have to go to the bathroom sometime.
As I think I said earlier in the thread, I consider the 16-1 play all is weak. And so it may turn out that for your 16-1 spread, Surrender is Not CBS.
Let me conclude with one thought. If a play (such as Surrender 16v9) with an Index of 2, is NOT CBS for some spread, then I think the spread is too weak.
PS: I noticed that you mentioned surrender 88v10 was CBS. Did you compute this with DAS? It doesn't surprise me that it is CBS, but I was wondering if that was part of the parameter set.
It is quite common to go into a casino full of low stakes shoes and get the pit to set you up a higher limit game where you play almost entirely heads up. You get greatly increased speed but have to pay the price of playing through all the negative decks. If you move to another table during a negative count then they will likely change your special table back to $5.
I either move on or take a break and wait for the crowd to loosen up. Never force yourself to play under bad conditions.
Heads up play is more than twice as fast which effectively more than doubles the SCORE. The higher stakes player has to accept this. As the casino empties they shut down more tables. Often the near empty casino has the worst conditions as only one or two tables is open with all the players crowded in there.
Given the choice of playing heads up and having to play all vs playing with another player and getting out of half the really poor counts, I may still elect heads up. Sometimes it is a close call. With two other players at the table I'm sure the call is not close at all. I haven't simmed this, just my gut feeling.
Especially if you're dealing with mediocore conditions and using the same bet spread. The only time one should consider playing all heads up is if they have a huge spread(not to mention heat) because of decreasing their minimum bet. Sims will prove this. It's more efficient to consistently avoid negative shoes and simply leave the table when the count goes bad. You'll have a far better chance of winning more money while reducing risk of ruin. I've never seen a TC of more than -4 because I refuse to throw my money away. I believe one of the main reasons for being a long term succesful card counter is because I simply refuse to waste my time playing all.
