tomchina123
tomchina123
Joined: Aug 26, 2015
  • Threads: 36
  • Posts: 257
Thanks for this post from:
ksdjdj
June 11th, 2019 at 3:40:40 AM permalink
Quote: ksdjdj

I haven't been able to understand the whole post, but you are on the right track when you said for an insurance side count you need to treat "A as 1".
If you have found a way to make it "... not hard burden for memory if only we know why it is like this", then good for you.
It seems like you are good enough to be able to use your strategy^*^, either with a team or by yourself.
^*^: Edit, this originally said "system", but then I remembered that on this site "system" is generally considered a "bad word".
But for someone like me, I would only be able to play at a live internet casino using a more complex count than "hi lo".
For keeping a side count, I would use:
1. A "true %" by using a spreadsheet to keep track of all the cards left etc, and when the "non-ten" (A to 9) cards are less than 2/3 of the shoe that is left, then it would be +EV to make the insurance bet.
2. Or a count like the one shown in the link below (for when your are in a "brick and mortar" casino):
https://www.888casino.com/blog/side-bets/card-counting-blackjack-insurance



That is also true, if the player's first card is worth X, the dealer shows the same card it is worth Y, and if the player looked at their cards after looking at the dealer's up-card, it is worth Z (see examples):

Example 1: if the player's first card is a 5 it is worth -19.65% (to the player on average) and if the dealer is showing a 5 in the same hand, it is worth +2.995% (on average).
Also, if the player hadn't looked at their cards until they saw the dealer's up-card, then the average EV of that hand would be +19.52%.

Example 2: if the player's first card is a 10 it is worth +14.35% (to the player on average) and if the dealer is showing a 10 in the same hand, it is worth -0.9487% (on average).
Also, if the player hadn't looked at their cards until they saw the dealer's up-card, then the average EV of that hand would be -17.36%.

Rules used for above examples: infinite deck game where BJ pays 3/2, and dealer stands on 17, double any 2 card total, split once and double after split.

Used MGP's BJ CA and http://www.bjstrat.net/cgi-bin/cdca.cgi^^^, for working out the above values.

^^^: used 41000000 decks for this site, because it doesn't have an "infinite deck option".

Hope this was helpful.

-------------------------

I agree with the above, and you (tomchina123) should ask if there are people on this site who are good at blackjack and can understand Chinese well, as I think a lot of people want to help you out, if they can (this seems unlikely, as I think they would have replied to you by now if they could).

Lastly, I congratulate you for using your non-native language to try to write a post that you think could benefit other readers and yourself.
English*** is one of the hardest languages to learn as a 2nd language.

***: English is my first language but I have a communication disorder, so I can find it hard to understand things or get my point across clearly.




i can only answer to some points:
i don't really need help. except i want to learn how to do ca of splits to 4 hands. otherwise, i have done what MJP and bjstrat did.
i just shared what i knew and found.
tomchina123
tomchina123
Joined: Aug 26, 2015
  • Threads: 36
  • Posts: 257
June 11th, 2019 at 3:49:40 AM permalink
Quote: DogHand

tomchina123,

Assuming the game is S17, according to the Wizard of Odds (see https://wizardofodds.com/games/blackjack/appendix/9/1ds17r4/ ), for 84 vs. 3 the Stand EV is -0.220209, while the Hit EV is -0.229755. Since the Stand EV is larger (in this case, less negative) than the Hit EV, BS is to Stand.

On the other hand, for 93 vs. 3 the Stand EV is -0.261815, while the Hit EV is -0.255712. Since the Hit EV is larger (in this case, less negative) than the Stand EV, BS is to Hit.

You are correct when you say "if there is a reason, then this reason works for the playing efficiency of all hands of blackjack": in every case, the optimal choice is the option with the highest EV.

Hope this helps!

Dog Hand



thanks for help. but i calculated this and knew this and shared the reason.
i don't know why all think little of me, even in China, i have a wechat group. there are some big money winners in my group. some wants me to be his student, some said: you finally understand sth of blackjack. in fact, this guy is a beginner.
why took me as a beginner because i don't win big money? but i won small moneys for many years. and Wizard, Eliot is not either big money winner.
i can do so many difficult calcuations. and write so many things(chinese website).
ssho88
ssho88
Joined: Oct 16, 2011
  • Threads: 29
  • Posts: 272
June 11th, 2019 at 7:55:57 AM permalink
Could you please show me your website ?
ssho88
ssho88
Joined: Oct 16, 2011
  • Threads: 29
  • Posts: 272
June 11th, 2019 at 9:02:24 AM permalink
Quote: tomchina123

that is language. i may feel ashamed to say i studied english as my major in college. but how about maths? i self learned to do a lot on blackjack, bac and poker(After read Great Eliot's articles)... is there no one who understands what i am trying to say?

maybe i just ask:
in single deck, 93V3, the right strategy is to hit, 84v3, it is to stand.
what is the reason behind it?
if there is a reason, then this reason works for the playing efficiency of all hands of blackjack.




Tom, I think what you said is only for SINGLE deck game . . .
tomchina123
tomchina123
Joined: Aug 26, 2015
  • Threads: 36
  • Posts: 257
June 11th, 2019 at 9:44:06 AM permalink
1st, website: i had made my own. but it was mainly the copy of wizard's site. daily visit: less than 3 persons. so i gave them up. this is a chinese blog website. or a high-end ask-and answer website, famous : https://www.zhihu.com/people/hu-xiao-jia-96/posts, chinese,no ads at all(i was wronged before) . baccarat info is not allowed. poker is ok, so i did a lot of maths on poker. and put them into articles. now blackjack. i still need to play to make money. i will fly to south korean tomorrow. i wrote for the reason only by this way,i can keep learning. now i may even cannot tell i learn to play or i play to learn. Life makes me a person like this, i cannot help. and i cannot help to say, i developed a world-class rental LED screen system,by many years, like this system, i know it is right by many ways. but it is not well accepted by the industry, if any would wonder what life i am living. i can tell: before this rental system, i made some good money by other LED screen designs with my Israel boss.

2nd. this hand is for single deck.
ssho88
ssho88
Joined: Oct 16, 2011
  • Threads: 29
  • Posts: 272
June 11th, 2019 at 2:28:28 PM permalink
Quote: tomchina123

1st, website: i had made my own. but it was mainly the copy of wizard's site. daily visit: less than 3 persons. so i gave them up. this is a chinese blog website. or a high-end ask-and answer website, famous : https://www.zhihu.com/people/hu-xiao-jia-96/posts, chinese,no ads at all(i was wronged before) . baccarat info is not allowed. poker is ok, so i did a lot of maths on poker. and put them into articles. now blackjack. i still need to play to make money. i will fly to south korean tomorrow. i wrote for the reason only by this way,i can keep learning. now i may even cannot tell i learn to play or i play to learn. Life makes me a person like this, i cannot help. and i cannot help to say, i developed a world-class rental LED screen system,by many years, like this system, i know it is right by many ways. but it is not well accepted by the industry, if any would wonder what life i am living. i can tell: before this rental system, i made some good money by other LED screen designs with my Israel boss.

2nd. this hand is for single deck.




Can't access to your website ?
gordonm888
gordonm888
Joined: Feb 18, 2015
  • Threads: 30
  • Posts: 1720
June 11th, 2019 at 5:30:57 PM permalink
tomchina, I have done calculations similar to what you have done, although I have done them for 1, 2,6 and 8 decks. The effects of EOR of various ranks on "close-call" decisions is far more important for 1 deck than for 6 or 8 deck.

The Hi-Lo count is pretty good for some decisions such as 15v10, 15v9 and Insurance, but is bad for 16v10, 10v10, 13v2, 12v3,4, etc.

Regarding your questions on doing combinational analysis on split pairs (with resplitting): LOL, good question, I don't know how to do it any better than the approximations used by Wizard and bjstrat in their CA hand evaluators; although the rigorous combinatorial analysis of split/resplit pairs is not a question with much practical significance. That's why we all use approximations.
Sometimes, people are just a bottomless mystery. And, after all, this is just a sh*tty little forum in the sun-less backwaters of the online world.
ksdjdj
ksdjdj
Joined: Oct 20, 2013
  • Threads: 65
  • Posts: 386
June 12th, 2019 at 3:40:25 AM permalink
Quote: tomchina123

i can only answer to some points:
i don't really need help. except i want to learn how to do ca of splits to 4 hands. otherwise, i have done what MJP and bjstrat did.
i just shared what i knew and found.


Here is an interesting link, but I don't know if the link below is helpful*** to you (as I haven't learnt those formulas, and I don't know if you can use those formulas for anything other than working out basic strategy,)

http://www.professorbray.net/Teaching/89s-MOU/2017-SummerTerm2/Papers/CWR_Paper1_Blackjack.pdf

***: page 21 is where it explains how to work out the EV of splitting.

There is a problem with it though because it says "To make things easier, we assume that player won’t split again after splitting.", and you wanted to learn how to do a "ca of splits to 4 hands".

Even so, someone better at this than me may tell you if this is a good starting point or not for what you want to do.

Update (4 am Pac time):

Also, with the bjstrat website, would changing the "shoe composition" from 4 to 2 for the cards you want to split help in getting a better estimate for what the EV is, when splitting to 4 hands in a 1-deck game ?
Last edited by: ksdjdj on Jun 12, 2019
gordonm888
gordonm888
Joined: Feb 18, 2015
  • Threads: 30
  • Posts: 1720
Thanks for this post from:
ksdjdj
June 12th, 2019 at 12:38:21 PM permalink
Quote: ksdjdj

Here is an interesting link, but I don't know if the link below is helpful*** to you (as I haven't learnt those formulas, and I don't know if you can use those formulas for anything other than working out basic strategy,)

http://www.professorbray.net/Teaching/89s-MOU/2017-SummerTerm2/Papers/CWR_Paper1_Blackjack.pdf

***: page 21 is where it explains how to work out the EV of splitting.

There is a problem with it though because it says "To make things easier, we assume that player won’t split again after splitting.", and you wanted to learn how to do a "ca of splits to 4 hands".

Even so, someone better at this than me may tell you if this is a good starting point or not for what you want to do.




The link you provided assumes no resplit and also assumes that both split pair cards have the same EV -which they do before you hit them, but it gives the 2nd hand no equity for having knowledge of what the cards were when the first hand was hit. So it is a below-average way of calculating the split pair EV.

Quote: ksdjdj

Also, with the bjstrat website, would changing the "shoe composition" from 4 to 2 for the cards you want to split help in getting a better estimate for what the EV is, when splitting to 4 hands in a 1-deck game ?



Confusing way to word your question, but my answer is NO. Bjstrat website calculator does indeed provide you with calculated values of the incrementally higher EV of resplit pairs - when you check "optimum strategy" rather than "basic strategy" in the inputs above. I don't think you can get a better analysis of 4 way split/resplit of pairs than the bjstrat calculator provides without doing a simulation.
Sometimes, people are just a bottomless mystery. And, after all, this is just a sh*tty little forum in the sun-less backwaters of the online world.
kuroshivo
kuroshivo
Joined: Jun 8, 2019
  • Threads: 1
  • Posts: 8
June 15th, 2019 at 6:53:36 AM permalink
Quote: gordonm888


Regarding your questions on doing combinational analysis on split pairs (with resplitting): LOL, good question, I don't know how to do it any better than the approximations used by Wizard and bjstrat in their CA hand evaluators; although the rigorous combinatorial analysis of split/resplit pairs is not a question with much practical significance. That's why we all use approximations.



would a free and open source blackjack engine help to do statistical (i.e. Monte Carlo) analysis?

I cannot post the link because I am still new tp the forum...

  • Jump to: