Baccarat Card Counting

Hi all!

I was hoping you guys could help me understand the Wizard's card counting advice for baccarat. Here is what confuses me: As he notes, high cards are good for the Player and low cards are good for the Banker. As such, why would it be better to bet on the Player when the count is -4 or lower, given that you minus 1 from the count when high cards (5-8) are dealt? Shouldn't the Player become a favourite when the count is positive instead of negative, since this would mean that more low cards than high cards have left the shoe?

I'm sure I'm just misunderstanding, but any help would be much appreciated.

It is not the same as Blackjack where the dealer can bust more easily with a stiff hand ( dealer is forced to draw to at least 17)and the player is more likely to get a bj with a 3/2 payout . In baccarat , if the player is more likely to draw a "small card" ,i am assumming 2 to 5 . it helps the player to get a better total due to the game's rule. the surplus of "small cards" doesn't hurt the banker . both the banker / player have the same chance of getting a natural , unlike bj a natural for player pays 3/2 while a natural in baccarat does not.

Thanks sg, that must be right, but I must say that I'm still confused. If you are right, then what does it mean to say that high cards favour the Player and low cards favour the Banker?

The card drawing rules in Baccarat are designed to give the Banker a little bit favor; they are non-symmetric. As a result, when the Banker wins, one pays a commission, balancing the odds. After the rules have set, it's hard to explain, intuitively, why high cards favour the Player and low cards favour the Banker (because of the rules). I imagine Wizard's statement is an empirical result based upon thousands of computer simulations, with effects of card removals.

Quote: J13

Hi all!

I was hoping you guys could help me understand the Wizard's card counting advice for baccarat. Here is what confuses me: As he notes, high cards are good for the Player and low cards are good for the Banker. As such, why would it be better to bet on the Player when the count is -4 or lower, given that you minus 1 from the count when high cards (5-8) are dealt? Shouldn't the Player become a favourite when the count is positive instead of negative, since this would mean that more low cards than high cards have left the shoe?

I'm sure I'm just misunderstanding, but any help would be much appreciated.

Here is my article on it:

https://www.888casino.com/blog/baccarat-tips/card-counting-in-baccarat

Here is my video on it:

https://youtu.be/tYQk7hZ_jAU

Thanks Eliot, that video is really enlightening! Unfortunately, I can't access the article -- 888 casino can't be accessed in Australia without a VPN. Is the material published anywhere else? Thanks again!

Quote: J13

Thanks Eliot, that video is really enlightening! Unfortunately, I can't access the article -- 888 casino can't be accessed in Australia without a VPN. Is the material published anywhere else? Thanks again!

Thanks, with the permission of 888, I re-posted the article on my website:

https://advancedadvantageplay.com/2021/05/29/why-baccarat-card-counting-sucks/

8 DECK, PLAY UNTIL LAST 14 CARDS, 100,000 shoe simulation.


A) BANKER BET

1)Strategy : Computer Perfect play
2) Bet size : Adjust according to current EV & Var, KELLY RATIO = 1, Fixed Bankroll = $1,000,000.
3) Average EV* = 1.35%
4) Rounds/shoe = 0.29
5) Total Winning/shoe = $24.14
6) The earliest round(with +ve EV) occurs in round no 63




B) PLAYER BET

1)Strategy : Computer Perfect play
2) Bet size : Adjust according to current EV & Var, KELLY RATIO = 1, Fixed Bankroll = $1,000,000.
3) Average EV* = 1.87%
4) Rounds/shoe = 0.24
5) Total Winning/shoe = $34.25
6) The earliest round(with +ve EV) occurs in round no 60


C) TIE BET(1 PAY 8)

1) Strategy : Computer Perfect play
2) Bet size : Adjust according to current EV & Var, KELLY RATIO = 1, Fixed Bankroll = $1,000,000.
3) Average EV* = 9.58%
4) Rounds/shoe = 0.26
5) Total Winning/shoe = $178.51
6) The earliest round(with +ve EV) occurs in round no 68




D) TIE BET(1 PAY 9)

1) Strategy : Computer Perfect play
2) Bet size : Adjust according to current EV & Var, KELLY RATIO = 1, Fixed Bankroll = $1,000,000.
3) Average EV* = 7.63%
4) Rounds/shoe = 2.58
5) Total Winning/shoe = $794.64
6) The earliest round(with +ve EV) occurs in round no 36

If you bet on 20 tables(online games) at the same time with the help of computer, your total winning/shoe = 20 * (24.14 + 34.25 + 178.51 ) = $4738, So worth to play ?

Pleaae note that Computer Perfect play only applicable for online casinos.
* Average EV = Total winning per shoe /Total Betting amount per shoe

I understand you are using perfect Kelly betting for this simulation. Even then, on a more typical $50k bankroll the win rate is about $1/shoe for the Banker and $1.50/shoe for the Player. Given that a shoe lasts about 90 minutes, you're talking under $2/hour with a $50k bankroll and computer-perfect Player/Banker betting.

Yep, it really is sad.

I have my own perfect play article that maximizes win-rate to get a limit on what's possible.

https://www.888casino.com/blog/baccarat-tips/computer-perfect-play-against-baccarat

Quote: teliot

I understand you are using perfect Kelly betting for this simulation. Even then, on a more typical $50k bankroll the win rate is about $1/shoe for the Banker and $1.50/shoe for the Player. Given that a shoe lasts about 90 minutes, you're talking under $2/hour with a $50k bankroll and computer-perfect Player/Banker betting.

Yep, it really is sad.

I have my own perfect play article that maximizes win-rate to get a limit on what's possible.

https://www.888casino.com/blog/baccarat-tips/computer-perfect-play-against-baccarat

I reposted my article on computer-perfect baccarat play for those who can't connect to 888.

https://advancedadvantageplay.com/2021/05/30/computer-perfect-baccarat-play/

Quote: teliot

I reposted my article on computer-perfect baccarat play for those who can't connect to 888.

https://advancedadvantageplay.com/2021/05/30/computer-perfect-baccarat-play/

For BANKER bet, when cut card place at 14 cards, my wager per shoe is 0.293 compare to your 0.1633, I think the reason behind is because I assumed no burn card before first round dealt from a new shoe.

Could you please change your program to no burn card and re-simulate it again ? Just want to verify my results.

In many, maybe most, casinos, dealer cut more than one deck of cards.

Quote: UCivan

In many, maybe most, casinos, dealer cut more than one deck of cards.

There are some online casinos only cut 7 cards !

Quote: ssho88

For BANKER bet, when cut card place at 14 cards, my wager per shoe is 0.293 compare to your 0.1633, I think the reason behind is because I assumed no burn card before first round dealt from a new shoe.

Could you please change your program to no burn card and re-simulate it again ? Just want to verify my results.

My cut card has one more round dealt after the cut card is dealt, so it goes much deeper than yours, possibly down to the last 2 cards with a 14 card cut. Also, you are betting Kelly, I am betting flat, so your max wager will be much higher in units on a $1M bankroll. Or am I misunderstanding?

Sorry, my coding days are over. Just re-broadcasting old news.

Quote: teliot

My cut card has one more round dealt after the cut card is dealt, so it goes much deeper than yours, possibly down to the last 2 cards with a 14 card cut. Also, you are betting Kelly, I am betting flat, so your max wager will be much higher in units on a $1M bankroll. Or am I misunderstanding?

Sorry, my coding days are over. Just re-broadcasting old news.

My cut card has one more round dealt after the cut card is dealt as well. kELLY BETTING will not change the wager per shoe.

I think the only different parameter is the INITIAL BURN CARD.

Quote: UCivan

In many, maybe most, casinos, dealer cut more than one deck of cards.

Most times I have seen a new baccarat deck, they shuffle, cut, count 7 or 8 cards from the back for the shuffle card, load the shoe.
Draw one, face up, burn that many cards face down, and open the table for play.

Procedures surely differ from house to house, but cutting a deck off the back wouldn't seem to offer an enhancement to security or revenue.

Deleted