blackjack switch & card counting?

#1
I was at the casino in the falls canada last weekend, and i noticed a game called blackjack switch. I didnt play it but you get two , two card hands with equal bets, and you can switch the cards after that, then i think you play it like a normal two hands but the dealed will push at 22. I noticed it wasnt a countinuos shuffle and the cards were dealt from a shoe. Is it possible to count cards at this game? If anybody knows. Im new at this idea of card counting. I noticed all other tables had CSM on the tables even 25 per hand games.:cool2:
 

Grisly Dreams

Well-Known Member
#2
Arnold Snyder's Big Book of Blackjack has a section on strategies for playing oddball games, and that section includes blackjack switch. I don't remember how much depth he goes into, though. Some of the games have actual counts, with accompanying indices, and I can't remember whether blackjack switch was one of those or not.
 

SystemsTrader

Well-Known Member
#3
It's best to avoid counting this game especially if you are new to blackjack. Don't use Snyder's Big Book of Blackjack he doesn't give you enough info to beat the game and his method doesn't cover how to play double downs or splits.
 

callipygian

Well-Known Member
#5
(1) Post this in the Variations forum, there are some threads in there already about BJ Switch.

(2) The problem with BJ Switch is that basic strategy is non-obvious. The no-brainer decisions account for maybe 1/3, the decisions that Snyder recommends covers about 1/3, but then the last 1/3 are not obvious at all. There are very few rules which can be used as guidelines and as a result I don't think people can bring down the house edge anywhere close to the theoretical 0.4%. More likely than not, it's 1% or higher.

(3) Counting works, and Hi-Lo is as good as any.
 

KenSmith

Administrator
Staff member
#6
callipygian said:
(3) Counting works, and Hi-Lo is as good as any.
Actually, ExhibitCAA had mentioned the possibility of him posting the effect of removal for this game. I suspect we may see a substantially different EOR than regular blackjack. I would be interested to see them. Hi-Lo may not compare favorably to other alternatives depending on the answer.
 

callipygian

Well-Known Member
#7
KenSmith said:
Actually, ExhibitCAA had mentioned the possibility of him posting the effect of removal for this game. I suspect we may see a substantially different EOR than regular blackjack. I would be interested to see them. Hi-Lo may not compare favorably to other alternatives depending on the answer.
I calculated it once and I don't remember it being significantly different from regular blackjack - with, again, the caveat being that people have to switch perfectly, which is non-obvious. The EOR's may be different if people are playing with sub-optimal switching strategy.

I'll post what I have if people think that the value of posting it outweighs the cost of disclosure. I'll give a few days for people to comment.
 

ExhibitCAA

Well-Known Member
#8
I gave a talk on this very topic at the gaming conference. I have more loose ends to finish off on the topic before I'd like to say anything, but I will say this: HiLo is fine. Traditionally, HiLo overvalues the 2, but in BJSwitch, the EOR of removing a 2 is higher due to the push on 22 rule, so HiLo comes out well. I created a variation called the "Switch Halves" count (still mulling over different names), with HiLo tags except the Ace is downgraded to -0.5 and the 9 is -0.5. [Incidentally, only two people--both longtime pros--at the conference asked for a copy of the paper.]
 

KenSmith

Administrator
Staff member
#9
Cool. Gotta attend one of those conferences sometime. A few years back I spent some time in the library at Reno looking through papers from previous conferences.

I'll be interested to see your results, especially given the comment about underweighting the Ace.
 
#10
ExhibitCAA said:
I gave a talk on this very topic at the gaming conference. I have more loose ends to finish off on the topic before I'd like to say anything, but I will say this: HiLo is fine. Traditionally, HiLo overvalues the 2, but in BJSwitch, the EOR of removing a 2 is higher due to the push on 22 rule, so HiLo comes out well. I created a variation called the "Switch Halves" count (still mulling over different names), with HiLo tags except the Ace is downgraded to -0.5 and the 9 is -0.5. [Incidentally, only two people--both longtime pros--at the conference asked for a copy of the paper.]
Nice. I use "Unbalanced Zen Halves" on 6:5 games- {-1,1,2,2,3,2,1,0,-1,-2}

Maybe RPM would be a slight improvement over High-Low for people who don't want to use a level 2 system.

Did you find a significantly different EOR for the ace between the original version of the game and the downgrade where a switched natural is treated like a drawn 21?
 

SystemsTrader

Well-Known Member
#11
callipygian said:
I calculated it once and I don't remember it being significantly different from regular blackjack - with, again, the caveat being that people have to switch perfectly, which is non-obvious. The EOR's may be different if people are playing with sub-optimal switching strategy.

I'll post what I have if people think that the value of posting it outweighs the cost of disclosure. I'll give a few days for people to comment.
Callipygian or Exhibit are you guys ready or willing to post the EOR for Switch? I would love to see them, I am curious to see how they compare to a regular blackjack EOR.
 
#14
Any pointers for EORs and atleast an approximate system?

Any pointers for the EORs and a good or atleast an approximate system to beat this?



Regards
Waugh
 

assume_R

Well-Known Member
#15
steve waugh said:
Any pointers for the EORs and a good or atleast an approximate system to beat this?

Regards
Waugh
If you look at http://wizardofodds.com/blackjackswitch you'll notice the complicated chart (in brown, with all the EV's) which is how you determine whether or not to switch your cards. I feel like until this can be simplified into a condense basic strategy chart, it will be impossible to play correctly. There's no way without the help of a computer that a person can memorize that entire chart, and be able to add 2 random numbers in that chart for EVERY hand. If that chart were somehow simplified into a basic-strategy-type chart, then the EOR's could be estimated presumably by sims. I'm curious to hear how exhibitcaa or callipygian went about their analysis of this game, and if they advocate using the hard-to-memorize chart on wizardofodds. Even on wizardofodds he links to a computer based calculator telling one whether or not to switch http://wizardofodds.com/blackjack/bj_switch_calc.html
 

assume_R

Well-Known Member
#16
On further thought, the basic strategy chart would be very very complicated.

Example: Let's you have a 10,6 on one hand and an 8 on your other hand. Do you switch? Well you have to consider exactly what composition is for that 8, as well as the dealer's up card.

This would be the basic strategy chart just for that specific situation (10,6 and 8) below. There would be a different chart for each composition dependent hand 1 situation and each total in hand 2.

Now that is simply the "brute force" b.s. chart, but there may be patters, such as "if you end up with a 20 on one of the hands, it is always better to choose that hand, unless you can possibly get a 21" which would make the generation of even simpler b.s. charts. Or perhaps the basic strategy chart for X,6 and 8 is the same as X,5 and 8. But coming up with those patterns would require a lot of analysis.

 

moo321

Well-Known Member
#17
Can't you just memorize a full EV chart? I mean it would be a massive bitch, but if the EOR is good, it could be worth it to play heat free with good EV. I mean, if guys used to memorize full index charts...
 

assume_R

Well-Known Member
#18
moo321 said:
Can't you just memorize a full EV chart? I mean it would be a massive bitch, but if the EOR is good, it could be worth it to play heat free with good EV. I mean, if guys used to memorize full index charts...
Okay, then you have to memorize the full EV chart, and then you have to calculate EV1 + EV2 ?>? EV3 + EV4 where EV1 and EV2 are the 2 ev's if you keep your 2 hands, and ev3 and ev4 are if you switch the 2 cards. So memorizing the exact decimal numbers to a few places for the entire would be only part of the problem. The other part is thinking about what your switched hands would be, and adding those high-precision numbers on the fly. A few errors per hour would definitely occur as well as a low # of hands per hour, at least for a few months. You could definitely round all the numbers, which would help. Up to you if you're up for the challenge and dedication it would require, which would be more complicated than a full index chart. And that's just for the B.S.. Image also having to count and memorizing additional index numbers, which would be a huge problem to calculate.
 
#19
Snyder's rules for switching and any decent count

I am wondering if we can use Snyder's strategy for switching and use any decent counting for betting an playing decisions.
Auto Monkey-would you be able to post or mail
me the tags and strategy charts?


Callipygian-Please post yours if you are willing to.

Regards
Waugh
 

assume_R

Well-Known Member
#20
steve waugh said:
I am wondering if we can use Snyder's strategy for switching and use any decent counting for betting an playing decisions.
Auto Monkey-would you be able to post or mail
me the tags and strategy charts?


Callipygian-Please post yours if you are willing to.

Regards
Waugh

What's "Snyder's strategy"?
 
Top