BJ SIMULATION WITH HI-OPT 2 INDICES

Hi,

I need someone to simulate below online video BJ game :-

1) Single deck
2) Peek rule
3) H17
4) No surrender but player allow to bet the insurance side bet( 2 : 1)
5) DAS, Double after split
6) Double on hard 10 or 11 only
7) Original hand split once to form 2 hands, NO resplit.
8) Split Ace only receive one card
9) Bet max 3 hands per round
10) Shuffle after each round.

The house edge = 0.37% per hand and 0.33% per wager. Player can bet up to 3 box per round and player can see at least 7 cards(including dealer's upcard) before HIT/STAND/DOUBLE/SPLIT.

It is online casino video BJ game, player will be given bonus(extra free credit) when make deposit, however player have to meet the wagering requirement before allow to make withdrawal. I manage to win consistently and hope can increase the profit with INDICES PLAY. With INDICES PLAY, player can increase the playing efficiency and betting 5 or 10 times higher in box 2 and box 3 can reduce house edge per wager and therefore can increase profit further.

Please help me to simulate this games under these scenarios :-


1) In order to verify my results, just simulate it AGAIN base on basic strategy(NO INDICES PLAY) and get the house edge, HE1

2) Simulate it with full HI-OPT 2 indices, update the true count(TC) after each card dealt and play according to the updated TC, and get the house edge, HE2. I can send you the HI-OPT 2 indices if you want it.


I suggest to bet 1 unit, 5 units and 10 units in BOX 1, BOX 2 and BOX 3 respectively. Perhaps you can suggest a better bet size.

I would like to know the difference between HE1 and HE2. I can share with you how I make money from online BJ games( or any other low house edge games).


Cheers

James

Hi James,

So you're looking to depth charge the SD game based on the fact that the site lets you vary your bets between the betting squares? What's the MIN and what's the MAX you can bet? After all, you can bet 1x1x10 but if the table max is $25, well that's not going to make things very interesting. Next, you didn't give any of the wagering requirements, so that you could figure out your return straight up on the game, and then your return if in fact the game with HI-OP II indices could get a positive expectation on box 3.

Lastly, for the people whom might simulate this for you (I don't have any software currently to do the simulation) you might want to be a bit more courteous? Posting a thread saying "hey someone simulate this for me" is a bit on the nose. Someone had to purchase the simulation software, then they have to set up the trial exactly as you're mentioning, run and examine the results. To me this is a bit more work than I'd be willing to do for someone randomly online, but some people do enjoy running them and reporting. I'd just highly recommend you purchase simulation software yourself. It's really not 'that' expensive and at the end of the day it'll pay for itself 100x over if you're any serious about blackjack games.

HI-OPT-II is a count that is designed to support decisions about when to increase bet size based on the cards that were played on previous dealt hands. But your online BJ game reshuffles every hand, so that is not a decision you can make. Instead, you want to make optimal decisions when faced with these situations:

- Insurance (1st 2 cards vs Ace)
-16vTen
-11vAce
-12v 4,3,5
-13v2
-15vTen
-10 v Ace

The Hi-OPT-II count/indices tables are just dreadful for what you are trying to do.

In your proposed situation, with one deal against three hands before a reshuffle, the rule on when to take the insurance bet (on all 3 hands) is: Take insurance vs a dealer ace when all 6 of your cards are A-9, or "not Tens."

6 "Non Ten" cards vs. Ace
Frequency of occurrence = 0.006933 (almost 0.7%)
EV of Insurance bet = +0.03333

Using the Hi-OPT-II criteria, which says you need a count of 4 to make the insurance bet, you will make the wrong decision on these kinds of hands

You have 6 cards that are NOT Tens, but the count is 3 or less (Vs Ace)
Example: Player: 92, 88, 63 versus Ace This hand has a HI-OPT-II count of 3 and your simulator will not take insurance, but the insurance bet with these cards is EV= +0.0333

Hands where you have 2 Tens and the count is 4 (Vs Ace)

Example: Player: T5, T4, 54 versus Dealer: Ace This hand has a HI-OPT-II count of 4 and your simulator will take insurance, but the insurance bet with these cards is a loser, with a negative EV: EV= - 0.03333
Making a 16 v T decision
This is the 2nd most important indice decision, you will have this situation over 2% of the time.

Hi-OPT-II is a terrible count for making this decision. It counts the 6 as if it is a low card, but in this case it acts like a high card and will make your 16 go bust. Since over half of your 16vT hands will be T-6 v T, and because the 6 is handled wrong, Hi-OPT-II will have you making some T6vT decisions that are wrong by as much as 3%.

HI-OPT-II assigns the 9s a count of 0, but for a 16vT decision it pretty much acts like a Ten.

Example: Here are two hands, each with a HI-OPT-II Count =+1, and the first decision is to act on a 16 v Ten:
1. Player: T6 99 62 vs Dealer: Ten
Stand on 16 ,EV= -0.5534 Hit on 16, EV= -0.4836 Optimum = HIT

2. Player: 97 T5 88 vs Dealer: Ten
Stand on 16, EV= -0.5008 Hit on 16, EV= -0.5109 Optimum = STAND

Same, count, but different optimal decisions. Hi-OPT-II is NOT the right tool, IMO.

Good work Gordon. As I tried to relate to the op by pm , rather subjectively, I admit: While counting is worthy at deciding when to Wong or ramp bets, it will be of negligible value after the wagers are placed. Those very occasional times where the micro count will dictate a variation from bs, will add very nearly nothing to the EV.

As mentionned above, if you are playing only one deal you don't want a counting system. Combinatorics are the optimal way to inform your decision.
Something like this: https://wizardofodds.com/games/blackjack/hand-calculator/
you might even want to use that if close calls don't happen very often. If they do it might be tedious to input cards on it every time and a chart might be better. Which you can write based on that.
The thing I am mostly worried about is being able to take your money out if you win. This method is known and online casinos can detect it. Also, as far as I know they have full discretion on bonuses and can freeze them (and even your deposits) if they don't like the way you are playing. That's a much bigger risk than bad variance imo. If you factor that in I doubt you have a profitable setting.

Quote: OnceDear

Good work Gordon. As I tried to relate to the op by pm , rather subjectively, I admit: While counting is worthy at deciding when to Wong or ramp bets, it will be of negligible value after the wagers are placed. Those very occasional times where the micro count will dictate a variation from bs, will add very nearly nothing to the EV.

Totally agree.

Taking insurance when none of the 6 cards is a Ten will reduce the house edge from 0.37% to 0.35%

From there, you can probably pick up another 0.05% (on the 3rd hand) by taking into account all the cards you see and playing computer perfectly, so you could probably get the House Edge down to 0.30% (on the 3rd hand.) The problem is that there is probably less than 2% of your hands for which the cards on the table will justify a deviation from Basic Strategy (and maybe less than 1%.)

Single-deck is an interesting topic though. There are lots of (infrequent) departures from Basic Strategy that are not in the indices, because counting systems can't catch them:
- split 6-6 v 8, if there are two 9s on the table (and the other cards you see are reasonably balanced in rank distribution.)
- stand on T-5 v T if there are two 6s on the table (and any other cards are reasonably balanced in rank distribution.)
-stand on 8-4 vs 3. And any 12 vs 3 with an 8 on the table is a good candidate for standing when the other cards are reasonably balanced in rank distribution. And if you can see 2 8's on the table, standing a 12 vs 3 is usually a great move.

Thanks Gordon.

If I can lower house edge by 0.07%, and base on my total wagering per year, then I would probably "save" about $28k per year !

LOL

Quote: ssho88

Thanks Gordon.

If I can lower house edge by 0.07%, and base on my total wagering per year, then I would probably "save" about $28k per year !

LOL

Hey buddy! LOL, great to see you back on the forums, ssho88.