Canyonero
• Posts: 509
Joined: Nov 19, 2012
July 11th, 2016 at 3:11:23 PM permalink
On my upcoming Vegas trip, I would like to give baccarat a shot. While everyone is into trends and stuff there, I would like to make an informed decision when to bet player or banker. However the system must be simple and easy enough to be fun. Here is what I came up with:

If any of you computer and baccarat nerds could run a simulation on the combined house edge, I would really appreciate that!

Card values from 1 (ace) to 4 are worth their value as a negative number (easy enough):
A = -1
2 = -2
3 = -3
4 = -4

5 and 7 equal 3 points each
6 equals 4 points (this part needs to be memorized):
5 = +3
6 = +4
7 = +3

All other values are ignored.

Keep calculating a running count. Divide by the number of decks remaining in the shoe.

By default, bet banker. While your result above ("true count") is 6 or higher, bet player instead.

[If you do run a simulation, is switching to player at TC of 5 the better variant? Thanks!]
CyrusV
• Posts: 103
Joined: Aug 8, 2015
July 11th, 2016 at 3:29:17 PM permalink
There is an app you can install for your smart phone, can't post the link, just google "baccarat card counting app" by pegi 12

You could clandestinely use it in live play, or play around with it, say remove all the small or high value cards and you'll notice the expectation hardly changes. There has been enough chatter - articles over the decades showing counting doesn't work with Baccarat.
Canyonero
• Posts: 509
Joined: Nov 19, 2012
July 12th, 2016 at 12:42:59 AM permalink
Quote: CyrusV

There has been enough chatter - articles over the decades showing counting doesn't work with Baccarat.

Read my post again plz. We are not talking about AP or a positive expectation here.

The point is to switch to the player bet, when its house edge becomes lower than the banker bet. The switch happens around 1.14 %. The idea is to develop a simple way to find that point.
CyrusV
• Posts: 103
Joined: Aug 8, 2015
July 13th, 2016 at 2:21:47 AM permalink
I understand, apply a count and when certain conditions are met bet the Player, no positive expectation expected. In reality betting the Player based on high card depletion never happens, as the app I mentioned above proves. Install the app remove all the high cards, press calculate, betting Bank is always recommended. Even it you did deplete the high cards to such a degree (in your case, apply values and a count, don't see the difference), which you can do with the app, you wouldn't see such a condition at the tables.
Canyonero
• Posts: 509
Joined: Nov 19, 2012
July 13th, 2016 at 3:59:53 AM permalink
I believe you are incorrect. I played through 2 shoes and both times the player bet HE went below 1% at some point (while banker went up accordingly).

Also, if you look at my count system, the TC of 6 isn't that hard to reach:

Just assume 4 decks left, with 2 extra 5,6,7 each removed and 2 more A, 2, 3 still in the deck

Running count: 32 - True count: 8

That example doesn't seem that much out of the ordinary to me...

Edit: In the above case: Player HE: 1.06%; Banker HE: 1.23% - bet the player!
Last edited by: Canyonero on Jul 13, 2016
charlesp
• Posts: 5
Joined: Jun 25, 2018
June 25th, 2018 at 9:43:41 AM permalink
I tested out your counting strategy this weekend in a live casino. Most of the shoes I played it was in reversed. High count was favoring the Banker instead and negative counts were favoring the player. Any suggestions?
DogHand
• Posts: 1661
Joined: Sep 24, 2011
Thanked by
June 26th, 2018 at 11:54:10 AM permalink
Quote: Canyonero

On my upcoming Vegas trip, I would like to give baccarat a shot. While everyone is into trends and stuff there, I would like to make an informed decision when to bet player or banker. However the system must be simple and easy enough to be fun. Here is what I came up with:

If any of you computer and baccarat nerds could run a simulation on the combined house edge, I would really appreciate that!

Card values from 1 (ace) to 4 are worth their value as a negative number (easy enough):
A = -1
2 = -2
3 = -3
4 = -4

5 and 7 equal 3 points each
6 equals 4 points (this part needs to be memorized):
5 = +3
6 = +4
7 = +3

All other values are ignored.

Keep calculating a running count. Divide by the number of decks remaining in the shoe.

By default, bet banker. While your result above ("true count") is 6 or higher, bet player instead.

[If you do run a simulation, is switching to player at TC of 5 the better variant? Thanks!]

Canyonero,

I just recently saw this post and decided to use my Baccarat simulator to answer your query.

I simmed 10-million shoes of 8D Baccarat with 13 cards cut off and one more round after the cut card. Using your tags, I calculated the TC distribution and the number of Banker, Player, and Tie wins for each TC from -10 to +10. I then calculated the EV for each of the three bets, assuming the usual values of Banker pays 0.95:1, Player pays 1:1, and Tie pays 8:1. Here are the results:

TC TCDist Banker Player Tie TC% EV(B) EV(P) EV(T)
10
65,887,581
30,126,823
29,575,801
6,184,957
8.086%
-1.450%
-0.836%
-15.516%
9
10,552,654
4,832,895
4,727,016
992,743
1.295%
-1.287%
-1.003%
-15.332%
8
13,227,922
6,058,109
5,922,600
1,247,213
1.623%
-1.265%
-1.024%
-15.142%
7
15,363,271
7,038,060
6,875,675
1,449,536
1.885%
-1.234%
-1.057%
-15.084%
6
19,481,657
8,927,560
8,713,768
1,840,329
2.391%
-1.194%
-1.097%
-14.982%
5
23,540,430
10,787,928
10,525,104
2,227,398
2.889%
-1.175%
-1.116%
-14.842%
4
30,305,664
13,888,887
13,547,616
2,869,161
3.719%
-1.165%
-1.126%
-14.793%
3
37,000,744
16,956,225
16,536,360
3,508,159
4.541%
-1.157%
-1.135%
-14.668%
2
48,453,029
22,218,114
21,641,351
4,593,564
5.946%
-1.102%
-1.190%
-14.676%
1
61,570,805
28,228,884
27,491,011
5,850,910
7.556%
-1.094%
-1.198%
-14.475%
0
94,112,811
43,151,735
42,011,185
8,949,891
11.550%
-1.081%
-1.212%
-14.412%
-1
70,824,112
32,485,643
31,600,205
6,738,264
8.692%
-1.043%
-1.250%
-14.373%
-2
62,506,191
28,674,509
27,874,635
5,957,047
7.671%
-1.014%
-1.280%
-14.227%
-3
47,553,774
21,815,591
21,205,758
4,532,425
5.836%
-1.011%
-1.282%
-14.220%
-4
38,005,404
17,439,925
16,940,949
3,624,530
4.664%
-0.981%
-1.313%
-14.168%
-5
29,122,511
13,367,371
12,976,356
2,778,784
3.574%
-0.952%
-1.343%
-14.125%
-6
23,958,152
10,995,874
10,673,008
2,289,270
2.940%
-0.947%
-1.348%
-14.002%
-7
18,873,407
8,659,348
8,406,417
1,807,642
2.316%
-0.954%
-1.340%
-13.801%
-8
15,941,583
7,317,991
7,097,284
1,526,308
1.956%
-0.911%
-1.384%
-13.831%
-9
12,481,126
5,726,688
5,557,864
1,196,574
1.532%
-0.942%
-1.353%
-13.716%
-10
76,071,402
34,947,788
33,737,940
7,385,674
9.336%
-0.707%
-1.590%
-12.620%
Total
814,834,230
373,645,948
363,637,903
77,550,379
100.000%
-1.065%
-1.228%
-14.344%

You wanted to know the TC at which the Player bet is better than the Banker bet. The table shows that at TC's of +2 and below, Banker is the better bettor's bet; at TC's of +3 and above, Player is the better bettor's bet.

However, as you can plainly see, the Banker and Player EV's are never positive in this range, so card counting the Banker and Player bets in Baccarat will not provide a +EV play for you. Thus, the best bettor's bet in this case is not to play ;-)

Hope this helps!

Dog Hand
Kielbasi
• Posts: 2
Joined: Jan 6, 2014
Thanked by
July 24th, 2018 at 8:00:21 PM permalink
i don't think I could ever understand the charting and other data in the game of baccarat. Just looks too complex.

What I do is observe the other players, and just see what the smart looking Asian women at the table are playing and just follow their lead. I've won using the system here in Pittsburgh