Math guys....

1357111317

Well-Known Member
#1
I have a question that hopefully the math guys will be able to help me with.

Lets say you are backbetting on someone in a positive count. A negative EV split comes up 88 v 8 9 X, 22 33 77 v 7, 22 33 66 77 v 2 and 22 33 66 77 v 3. If you dont split the backbet and the front person does you just get to put your bet on one of the hands.

First question. If you always played behind this person and this person played perfect basic stratagy what kind of EV would you save on 6D DAS H17.

Second question. At what kinds of counts would these negative EV splits become positive EV?

Thanks.
 
#3
Sonny said:
Check out the Backline Strategy sources below:

http://www.blackjackinfo.com/bb/showthread.php?p=21994

You can get an additional 0.2% advantage by using Wong's "sacrifice split" strategy. Grosjean's book will give you the exact critical indexes for making each play so you can potentially do better than that if you customize your strategy.

-Sonny-
It's very rules-dependent of course. I've never done it for H17 because I don't know a H17 store that allows it.

Most of the gain appears to come from 88 vs. 9-A. I suppose the higher the count, the more powerful the option of not splitting a back bet would be on those hands.

Here are some charts I derived- do they correlate to what has already been published?
 

Attachments

gibsonlp33stl

Well-Known Member
#4
Or as this article suggests:

http://www.blackjackforumonline.com/content/scavengerbjfo.pdf


There can be a large number of variations to splitting if it's not assumed that you have to split for the same amount. Everywhere I play allows splits/doubes for less. Interestingly - if you have say a $50 bet out for a 33 v 9, say you're playing at a $10 table. If you could split for less with a $10 bet...you gain value. I hadn't really thought of that before. Same scenario with an 88 v 8/9/T/A. Seems like if you play smart there's a lot of positives there to make up some EV. But then again, none of this takes into effect the count at the time...so these values are for a count of zero, i'm sure if the count is 7 TC and you have some max bets out, the indices would change.

Does anyone employ these techniques successfully?
 

callipygian

Well-Known Member
#6
1357111317 said:
88 v 8 9 X, 22 33 77 v 7, 22 33 66 77 v 2 and 22 33 66 77 v 3 ... At what kinds of counts would these negative EV splits become positive EV?
With the caveat that this is on an infinite deck (H17 DAS DA2 LS) ...

22 vs. 2: >+5
22 vs. 3: >+1
22 vs. 7: >+1
33 vs. 2: >+6
33 vs. 3: >+2
33 vs. 7: NEVER
66 vs. 2: >+11
66 vs. 3: >+5
77 vs. 2: >+6
77 vs. 3: >+2
77 vs. 7: >+8
88 vs. 9: NEVER
88 vs. T: NEVER
88 vs. A: NEVER
 

Sonny

Well-Known Member
#7
sabre said:
I have never seen a single casino that allowed splitting for less
In a backline betting situation where only one player decides to split it will be a split for less. It can also happen in situations involving coupons or scavenger plays. There are also similar opportunities to split for more.

-Sonny-
 

1357111317

Well-Known Member
#9
callipygian said:
With the caveat that this is on an infinite deck (H17 DAS DA2 LS) ...

22 vs. 2: >+5
22 vs. 3: >+1
22 vs. 7: >+1
33 vs. 2: >+6
33 vs. 3: >+2
33 vs. 7: NEVER
66 vs. 2: >+11
66 vs. 3: >+5
77 vs. 2: >+6
77 vs. 3: >+2
77 vs. 7: >+8
88 vs. 9: NEVER
88 vs. T: NEVER
88 vs. A: NEVER

Are those Hi Lo?
 

1357111317

Well-Known Member
#10
Automatic Monkey said:
It's very rules-dependent of course. I've never done it for H17 because I don't know a H17 store that allows it.

Most of the gain appears to come from 88 vs. 9-A. I suppose the higher the count, the more powerful the option of not splitting a back bet would be on those hands.

Here are some charts I derived- do they correlate to what has already been published?
I read on one of the links that if the players backbet is 7 times more than the front bet. You can start doing splits that benefit the backlining player. That shee that you provided AM says that there are only 3 situations that the backlining player can gain an advantage if his bet is 10X greater than the front player, those are 77 v 8,9,X
 
#13
1357111317 said:
I read on one of the links that if the players backbet is 7 times more than the front bet. You can start doing splits that benefit the backlining player. That shee that you provided AM says that there are only 3 situations that the backlining player can gain an advantage if his bet is 10X greater than the front player, those are 77 v 8,9,X
Right. DAS games involve a lot of splitting, and most of the power from this technique comes from the times the backliner never splits. It both increases advantage and decreases standard deviation.

The 77 plays I have are right around a factor of 7 so I guess it agrees with the other link. The ratio has to be pretty darned high before you start splitting 2's against an ace or 10. A lot of the value of doing it comes from the fact that you are going to be able to double if you catch a 9. If you're in a non-DAS game the chart is going to look completely different.

Now keep in mind my chart is not adjusted for count, so it's not going to be quite accurate in a real-world situation. In a neutral count doubling against high cards are marginal plays so you don't get as much benefit from splitting low cards against a 10 or ace. But in a real game if you have a partner backlining you with a big bet the count is going to be high and doubling 11 vs 10 or A or doubling 10 vs. 9 become more desirable plays, thus the ratio of front to back bet to make the defensive split profitable is going to go down significantly.
 

1357111317

Well-Known Member
#14
Alright thanks for the info everyone.

So basically my situation is a teammate will occaisionally place a backbet on me up to 10X my bet in good counts. Should I just use the indices Calligypian gave for when to not split the backbet as well?
 

callipygian

Well-Known Member
#15
1357111317 said:
Alright thanks for the info everyone.

So basically my situation is a teammate will occaisionally place a backbet on me up to 10X my bet in good counts. Should I just use the indices Calligypian gave for when to not split the backbet as well?
WHOA WHOA WHOA!

I answered a different question than the one you apparently meant.

Define the following:
EV(hit) = EV of hitting a splittable hand
EV(stand) = EV of standing on a splittable hand
EV(split) = EV of splitting a hand = 2*EV(onecard)
EV(onecard) = EV of just having one card (one to be dealt)

I answered the case where EV(split)>0. This is related to, but not necessarily, the only case where you want to split.

Let's say the backbet is n times the small bet, such that you're betting (n+1) units. You can't compare apples to apples any more:

(n+1)*EV(hit) - original hand unsplit, hits
(n+1)*EV(stand) - original hand unsplit, stands
(n+2)*EV(onecard) - low bet splits, big bet doesn't
(2n+2)*EV(onecard) - low bet and big bets both split

Now, if EV(split)>0, then n*EV(onecard)>0, which means that using the indices I suggested is a good rule. However, you'll miss the cases in (n+1)*EV(hit) or (n+1)*EV(stand) is greater than (n+2)*EV(onecard) but less than (2n+2)*EV(onecard). In these cases the low bet shouldn't split either to save the big bet. As n grows large, this matters less.

I don't have the setup to analyze this exactly - for all I know, this odd scenario will never occur. But it does underscore the importance of asking the right question! :)
 
#16
1357111317 said:
Alright thanks for the info everyone.

So basically my situation is a teammate will occaisionally place a backbet on me up to 10X my bet in good counts. Should I just use the indices Calligypian gave for when to not split the backbet as well?
No, wait. Use the .PDF chart in one of my posts in this thread as a first approximation. For all of the plays in orange, the back bettor never splits. You will make all the BS splits.

Being the ratio is only 10:1, the only non-BS splits the front player is going to make are 77 vs 8,9,10. And of course the back bettor will not split.

This will give you approximately 0.18% greater advantage as a partnership, with lower variance due to less money being put on the table.

Second approximation: some of the defensive splits, in the orange cells, become offensive splits at high enough counts and the back bettor can split along with you. Callipygian's indices tell you where that point is. For your first time out, I'd recommend just using the plain chart and add the indices when you are comfortable with that, just like when you went from using Basic Strategy to regular index plays.
 
Top