Poll

No votes (0%)
No votes (0%)
2 votes (16.66%)
7 votes (58.33%)
No votes (0%)
1 vote (8.33%)
No votes (0%)
1 vote (8.33%)
1 vote (8.33%)
9 votes (75%)

12 members have voted

Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26432
Joined: Oct 14, 2009
August 24th, 2020 at 10:12:43 AM permalink
3 Giving 8 is a baccarat side bet found at the Hippodromo casino in London. It pays 200 to 1 if the Banker hand starts with a two-card total of three and the Player hand draws a third card of an 8.

For my analysis, please see my new page on 3 Giving 8. I welcome all comments.

The question for the poll is would you bet the 3 Giving 8?
Last edited by: Wizard on Aug 24, 2020
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
MattUK
MattUK
  • Threads: 13
  • Posts: 138
Joined: Jul 25, 2017
August 24th, 2020 at 11:21:25 AM permalink
Thank you Wiz! I saw it in Hippodrome and other London casinos as well, eg. Horizon, which is metres away on the same square.
https://horizonscasino.com/casino-games/baccarat/
I think it's available to whoever pays for the licence.
I could bet with a really decent bankroll if there's a chance to enter EV+. However, my guess is that could happen rarely and to sit by the table without betting at all for an hour or so is not suspicious at all. Otherwise, it's like Triple in Sic Bo paying 200 (which would be 6.94% edge, as it's 1/216 vs. 1/220.11, not to mention far easier).
Is the "Red 8 pays even" in the making? :-)
Last edited by: MattUK on Aug 24, 2020
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
August 24th, 2020 at 11:22:59 AM permalink
I wouldn't but I've seen people do it in the same way as they make the specific tie bets. I guess it's the thrill on a nice payout every now and then.

btw the Empire, according to their webpage, only give 180 to 1 and describe...
Quote: https://www.thecasinolsq.com/assets/caesars/images/files/10-CEUK-PuntoBanco-V1-Sept2019.pdf


THE HOUSE EDGE FOR PUNTO BANCO:
BANK IS 1.06%
PLAYER (PUNTO) IS 1.24%
ANY EGALITE AT 9-1 IS 4.84%
3G8 (THREE GIVING EIGHT) IS 17.77%
UR WAY EGALITES SIDE BET
TIE 0 13.40%
TIE 1 11.80%
TIE 2 10.10%
TIE 3 10.90%
TIE 4 12.90%
TIE 5 12.50%
TIE 6 13.30%
TIE 7 8.20%
TIE 8 12.20%
TIE 9 11.60%

MattUK
MattUK
  • Threads: 13
  • Posts: 138
Joined: Jul 25, 2017
August 24th, 2020 at 11:31:28 AM permalink
Quote: charliepatrick

btw the Empire, according to their webpage, only give 180 to 1


Yes, that's true. I saw the table with my eyes.
Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26432
Joined: Oct 14, 2009
August 24th, 2020 at 12:23:24 PM permalink
Thanks for the feedback. I see I forgot to properly leave a link. Here it is -- 3 Giving 8.

That omission may explain the results of the poll so far:

"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
unJon
unJon
  • Threads: 14
  • Posts: 4570
Joined: Jul 1, 2018
August 24th, 2020 at 12:40:02 PM permalink
Quote: Wizard

Thanks for the feedback. I see I forgot to properly leave a link. Here it is -- 3 Giving 8.

That omission may explain the results of the poll so far:


Wiz, how often would this be +EV if you counted 8s?
The race is not always to the swift, nor the battle to the strong; but that is the way to bet.
Ace2
Ace2
  • Threads: 32
  • Posts: 2672
Joined: Oct 2, 2017
August 24th, 2020 at 12:56:16 PM permalink
For infinite deck 200 to 1:

1 - (25+16*5)16 / 13^5 * 201 = 9.05% edge
It’s all about making that GTA
Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26432
Joined: Oct 14, 2009
August 24th, 2020 at 2:54:48 PM permalink
Quote: unJon

Wiz, how often would this be +EV if you counted 8s?



Good question. Thus, the "Is this countable?" choice in the poll. Perhaps I'll answer that, which would be a solid day of work.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
unJon
unJon
  • Threads: 14
  • Posts: 4570
Joined: Jul 1, 2018
August 24th, 2020 at 3:07:42 PM permalink
Quote: Wizard

Good question. Thus, the "Is this countable?" choice in the poll. Perhaps I'll answer that, which would be a solid day of work.

Thats what I voted for!
The race is not always to the swift, nor the battle to the strong; but that is the way to bet.
ThatDonGuy
ThatDonGuy
  • Threads: 117
  • Posts: 6218
Joined: Jun 22, 2011
Thanked by
MattUK
August 24th, 2020 at 4:06:29 PM permalink
For counting 8s, here's what I get after 50 million 8-deck shoes:

8s In The ShoeHouse Edge% of hands
325.5152.159
316.1962.937
306.4513.089
296.383.108
286.3423.11
276.4673.11
266.4063.11
256.3453.11
246.0563.11
236.3933.109
226.3553.11
216.1853.109
206.4643.109
196.3743.11
186.4283.11
175.823.11
166.0643.11
156.3583.11
146.273.11
136.2373.11
126.123.109
116.3213.11
106.2093.11
96.243.109
86.1583.109
76.4523.11
66.0653.109
56.4923.11
46.2563.11
36.2763.109
28.593.087
127.692.878
01001.891


Apparently, the best time to play is at the start of a new shoe, assuming none of the 8s were burned.
unJon
unJon
  • Threads: 14
  • Posts: 4570
Joined: Jul 1, 2018
August 24th, 2020 at 4:18:16 PM permalink
Quote: ThatDonGuy

For counting 8s, here's what I get after 50 million 8-deck shoes:


8s In The ShoeHouse Edge% of hands
325.5152.159
316.1962.937
306.4513.089
296.383.108
286.3423.11
276.4673.11
266.4063.11
256.3453.11
246.0563.11
236.3933.109
226.3553.11
216.1853.109
206.4643.109
196.3743.11
186.4283.11
175.823.11
166.0643.11
156.3583.11
146.273.11
136.2373.11
126.123.109
116.3213.11
106.2093.11
96.243.109
86.1583.109
76.4523.11
66.0653.109
56.4923.11
46.2563.11
36.2763.109
28.593.087
127.692.878
01001.891


Apparently, the best time to play is at the start of a new shoe, assuming none of the 8s were burned.



I don’t follow. If the shoe is half over and no 8s have come out, that would be better than at the start of a shoe.
The race is not always to the swift, nor the battle to the strong; but that is the way to bet.
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
August 24th, 2020 at 4:57:24 PM permalink
Quote: unJon

I don’t follow. If the shoe is half over and no 8s have come out, that would be better than at the start of a shoe.

I tend to agree.

For a decent simulation you need to develop a running count, for simplicity have -12 for any Eight and +1 for everything else. (If it was just Red 8's then -25 and +1.) Then divide the running count by the number of decks left in the shoe and remember the results for all the different counts. In theory if there are more 8's to come than usual it should drift towards the player's advantage. The trick is to find out where this happens and how often the player has an advantage.

In essence this is how you check for countability of a game or sidebet.

There's a measure which looks at the average if you sat and watched lots of hands, only betting $100 on all profitable hands, and seeing how much on average you won for every 100 hands watched. You can then compare this with other games (e.g. Blackjack) or just look at how much one might make per hour.

As an example in Blackjack, six decks, 66% penetration about 25% of hands would be played and the profit would be about $18 per 100 hands watched. If the penetration goes to 83% the profit is $27 per 100 hands. If there's a CSM it's under $2!

So I'd guess with Baccarat going down to the end (including the burn cards) some opportunities will happen occasionally.
Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26432
Joined: Oct 14, 2009
August 24th, 2020 at 5:37:02 PM permalink
I like to start any discussion about card counting with the Effect of Removal. To be specific, remove a single card from the shoe and calculate the house edge? What is the house edge according to all ten types of cards? It would be easy to say that removing an 8 would be bad, but let's see if the other cards are all equally good to be removed.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
gordonm888
Administrator
gordonm888
  • Threads: 60
  • Posts: 5001
Joined: Feb 18, 2015
August 24th, 2020 at 6:06:41 PM permalink
Quote: Wizard

I like to start any discussion about card counting with the Effect of Removal. To be specific, remove a single card from the shoe and calculate the house edge? What is the house edge according to all ten types of cards? It would be easy to say that removing an 8 would be bad, but let's see if the other cards are all equally good to be removed.



For this bet, I can do it in my head.

If there are 10% more 8's remaining in the deck then there should be, on average, then that will wipe out the house advantage. If the number of remaining 8's is elevated by >10%, then the player will have an advantage making this sidebet.

Its an easy calculation to do approximately, because of the nature of bacarrat: the probability of the player getting a two card total of 3 is not greatly affected by removal of any card except possibly a 3.

The probability of any dealer having a 5 or less, is not greatly affected by the removal of any card.

But the probability of the player drawing an 8 is directly and linearly affected by the number of 8's remaining in the deck, and thus the probability of winning this side bet scales approximately linearly with the number of 8's remaining.
So many better men, a few of them friends, are dead. And a thousand thousand slimy things live on, and so do I.
Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26432
Joined: Oct 14, 2009
August 24th, 2020 at 6:50:27 PM permalink
Quote: gordonm888

the probability of the player getting a two card total of 3 is not greatly affected by removal of any card except possibly a 3.



I would argue that the 3 exception is pretty obvious. The first hurdle to jump over is a Banker total of 3 and that is going to usually happen with a 0 and 3 as the first two cards.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
Marcusclark66
Marcusclark66
  • Threads: 33
  • Posts: 1140
Joined: Mar 26, 2020
August 24th, 2020 at 6:54:29 PM permalink
Quote: Wizard

I like to start any discussion about card counting with the Effect of Removal. To be specific, remove a single card from the shoe and calculate the house edge? What is the house edge according to all ten types of cards? It would be easy to say that removing an 8 would be bad, but let's see if the other cards are all equally good to be removed.



I might be wrong but I believe they're talking about a 38 special, the same as a Panda 8 in the United States. Where the bank wouldn't draw if the bank had three and the player drew up to an eight or anything that's comprised of an 8 on the third card.

But in this particular case they are saying the bank has to have only 3 points for the player to make a panda 8. It does not matter what card comes out or what combination of cards as long as it's a 3 card eight. I believe that's what they talking about?
Marcus Clark. Real Person; AKA MarcusClark66. *Professional Casino Security Expert. *Certified EMT *Certified Company Firearms Instructor. *Certified Gaming Regulations Interpreter for Corporate Applications. *Domestic UrbanTactical Combat Casualty Expert. *Tic-Tac-Toe Expert (Real Competitive Versions) & Mastering Chess. *Honorary & Official #1 Fan of the MDawg Adventures Club. *Mastering Cracking it. Bit-by-Bit, Piece-by-Piece Crediting Forum Members. *Certified Casino Property Entry & Exit Point Analyzer *Baccarat Winning Session Record: 12 out of 12 & 1 out of 1 Mini Session. Baccarat Losing Session Record: 2 Losing Sessions.
ssho88
ssho88
  • Threads: 53
  • Posts: 657
Joined: Oct 16, 2011
Thanked by
unJon
August 24th, 2020 at 6:58:16 PM permalink
Quote: unJon

Quote: Wizard

Thanks for the feedback. I see I forgot to properly leave a link. Here it is -- 3 Giving 8.

That omission may explain the results of the poll so far:


Wiz, how often would this be +EV if you counted 8s?



8 deck, play until last 14 cards(81.2 hands/shoe), 1 pay 200(HE = -8.6822658%), simulation results :-

a) Counting system : unbalanced simple running count system, tags from Ace to T(+1, +1, -3, +1, +1, +2, +2, -10, +2, +1).
b) ev/shoe = 3.85 unit(flat bet)
c) bet frequency/shoe= 21.4 rounds
d) average ev(%) = +18%
e) variance = 239

You will strike once for every 170 rounds of betting.

It is a high variance game and good luck !
Last edited by: ssho88 on Aug 25, 2020
ThatDonGuy
ThatDonGuy
  • Threads: 117
  • Posts: 6218
Joined: Jun 22, 2011
August 24th, 2020 at 9:38:11 PM permalink
Quote: charliepatrick

I tend to agree.

For a decent simulation you need to develop a running count, for simplicity have -12 for any Eight and +1 for everything else.


Using +12 for an eight and -1 for everything else, after 20 million shoes, I am getting some strange results:


CountHouse Edge
-25-9.899
-24-8.845
-23-8.772
-22-7.062
-21-5.891
-20-9.191
-19-9.532
-18-8.474
-17-7.839
-16-6.816
-15-5.179
-14-4.24
-13-3.083
-12-2.326
-11-1.243
-100.03
-90.792
-82.348
-7-3.344
-6-1.193
-50.366
-42.672
-34.182
-25.595
-16.619
07.56
18.801
210.522
310.718
411.957
512.938
622.297
722.322
822.907
922.86
1022.712
1123.107
1223.408
1323.923
1423.936
1524.251
1624.205
1724.894
1824.926
1925.053
2025.328
2125.67
2226.057
2325.671
2426.837
2527.125

unJon
unJon
  • Threads: 14
  • Posts: 4570
Joined: Jul 1, 2018
August 24th, 2020 at 9:39:51 PM permalink
Quote: ThatDonGuy

Using +12 for an eight and -1 for everything else, after 20 million shoes, I am getting some strange results:


CountHouse Edge
-25-9.899
-24-8.845
-23-8.772
-22-7.062
-21-5.891
-20-9.191
-19-9.532
-18-8.474
-17-7.839
-16-6.816
-15-5.179
-14-4.24
-13-3.083
-12-2.326
-11-1.243
-100.03
-90.792
-82.348
-7-3.344
-6-1.193
-50.366
-42.672
-34.182
-25.595
-16.619
07.56
18.801
210.522
310.718
411.957
512.938
622.297
722.322
822.907
922.86
1022.712
1123.107
1223.408
1323.923
1423.936
1524.251
1624.205
1724.894
1824.926
1925.053
2025.328
2125.67
2226.057
2325.671
2426.837
2527.125



Do you mean -12 for an 8 and +1 for everything else?
The race is not always to the swift, nor the battle to the strong; but that is the way to bet.
MattUK
MattUK
  • Threads: 13
  • Posts: 138
Joined: Jul 25, 2017
August 25th, 2020 at 2:20:00 AM permalink
Intuitively, the edge is out there, from time to time, BUT:
1) the shoe won't run until last 14 cards. As usually, the dealer can change it in the middle when things are only starting to be interestning.
2) the edge happens so rarely the player would have to sit in a witness mode for most of the time and that would become to much to bear for everyone after a few minutes. He could save the seat by betting 5 on Player and waiting for the edge to bet 100, but that makes it even more uphill.
3) realistically, it would take a few hours session to have any realistic chance to beat the house.
4) even then it's very volatile and there may be no hit the entire night which would be bitterly frustrating. Especially seeing someone winning big on roulette, eg. 3500 from the 100 you've lost again.
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
Thanked by
unJon
August 25th, 2020 at 4:00:34 AM permalink
I've done 10m shoes with 90% penetration (763m hands) and the bet is countable at 200/1 but not at 180/1. The figures are only slightly higher with 95%.

At 200/1 the bet goes +EV at about +5.2.

With 10m shoes there is noise around the cutover (4.7-5.6) due to the high variance of the bet itself and a range of .48%-.51% winners. With only 1m shoes it's even noiser (3.6-9.7) (i.e. 3.6 was positive and 9.7 was negative). I think this may have been what you were seeing in your results.

PayoutPenetrationProfit per 100 handsHands played
200/1
90%
$75.86
14.36%
180/1
90%
$1.22
0.99%
Note: I need to review these figures as I think they're based on hands played rather than hands watched. This would imply the profit made would be more.
Last edited by: charliepatrick on Aug 25, 2020
ThatDonGuy
ThatDonGuy
  • Threads: 117
  • Posts: 6218
Joined: Jun 22, 2011
August 25th, 2020 at 6:30:35 AM permalink
Quote: unJon

Quote: ThatDonGuy

Using +12 for an eight and -1 for everything else, after 20 million shoes, I am getting some strange results:



Do you mean -12 for an 8 and +1 for everything else?



No, it's +12 for an 8 and -1 for everything else. The higher the count, the fewer 8s are in the deck, so the less likely that the player's third card will be one.
Why, yes, this is counter to the usual "higher counts are better for the player," isn't it? I've never been much of a card counter...

Note that the values are the House Edge; the APs are the negative numbers.

Also note that this is not a True Count.
Last edited by: ThatDonGuy on Aug 25, 2020
gordonm888
Administrator
gordonm888
  • Threads: 60
  • Posts: 5001
Joined: Feb 18, 2015
August 25th, 2020 at 6:51:26 AM permalink
Quote: ThatDonGuy

No, it's +12 for an 8 and -1 for everything else. The higher the count, the fewer 8s are in the deck, so the less likely that the player's third card will be one.
Why, yes, this is counter to the usual "higher counts are better for the player," isn't it? I've never been much of a card counter...

Also note that this is not a True Count.



It absolutely needs to be analyzed with a True Count -i.e., adjusted for number of remaining cards. Sorry, ThatDonGuy, but your table is not particularly helpful

Look at Ssho88's post. Ssho88 is very knowledgable about counting at Bacarrat and is one of the top math guys in this forum, although he is "under the radar." I trust his numbers.
Last edited by: gordonm888 on Aug 25, 2020
So many better men, a few of them friends, are dead. And a thousand thousand slimy things live on, and so do I.
ThatDonGuy
ThatDonGuy
  • Threads: 117
  • Posts: 6218
Joined: Jun 22, 2011
Thanked by
gordonm888unJon
August 25th, 2020 at 7:01:27 AM permalink
Quote: gordonm888

It absolutely needs to be analyzed with a True Count -i.e., adjusted for number of remaining cards. Sorry, ThatDonGuy, but your table is not particularly helpful


Got it. I also changed the count to -12 for each 8 and +1 for each other card, to make the positive counts better plays for the player.
After 50 million shoes, here's what I get:


Keep in mind the numbers are House Edge - the negative numbers are the APs
TCHouse Edge
-1025.429
-9.525.203
-923.707
-8.523.731
-823.007
-7.521.951
-720.925
-6.520.494
-619.222
-5.518.679
-517.782
-4.517.218
-416.422
-3.515.569
-314.461
-2.513.704
-212.951
-1.512.125
-111.251
-0.510.667
09.407
0.57.746
17.246
1.56.474
25.593
2.54.887
34.084
3.53.201
42.591
4.50.852
50.478
5.5-0.211
6-1.083
6.5-1.821
7-2.82
7.5-3.365
8-4.226
8.5-4.809
9-6.06
9.5-7.77
10-6.905


Looks like it's an AP with a TC of +5 1/2
ssho88
ssho88
  • Threads: 53
  • Posts: 657
Joined: Oct 16, 2011
August 25th, 2020 at 9:03:11 AM permalink
8 deck, play until last 14 cards(81.8 hands/shoe), 1 pay 200(HE = -8.6822658%), I would like to propose these two counting systems.

a) Unbalanced running count system, tags from Ace to T(+1, +1, -3, +1, +1, +2, +2, -10, +2, +1), bet when RC >= 34. Please note that RC trigger point will change slightly if the penetration change.

b) True count system, tags from Ace to T(+1, +1, -4, +1, +1, +2, +2, -10, +2, +1), bet when TC >= 4.2.


Both counting systems give similar ev/shoe but RC system is much easier to apply in the casino.
Marcusclark66
Marcusclark66
  • Threads: 33
  • Posts: 1140
Joined: Mar 26, 2020
August 25th, 2020 at 10:05:36 AM permalink
Quote: Marcusclark66

I might be wrong but I believe they're talking about a 38 special, the same as a Panda 8 in the United States. Where the bank wouldn't draw if the bank had three and the player drew up to an eight or anything that's comprised of an 8 on the third card.

But in this particular case they are saying the bank has to have only 3 points for the player to make a panda 8. It does not matter what card comes out or what combination of cards as long as it's a 3 card eight. I believe that's what they talking about?



Yes it is with only an 8 card.

However I like the two , 200 to 1 Wagers better in the United States now where the first one is a 3 card 8-9 does not matter which side has a three-card eight and the other side has to have a three card nine, that hand wins 200 to 1. The second one is a three-card 7-7 tie both sides have to have a three-card 7 and then it's 200 to 1.
Marcus Clark. Real Person; AKA MarcusClark66. *Professional Casino Security Expert. *Certified EMT *Certified Company Firearms Instructor. *Certified Gaming Regulations Interpreter for Corporate Applications. *Domestic UrbanTactical Combat Casualty Expert. *Tic-Tac-Toe Expert (Real Competitive Versions) & Mastering Chess. *Honorary & Official #1 Fan of the MDawg Adventures Club. *Mastering Cracking it. Bit-by-Bit, Piece-by-Piece Crediting Forum Members. *Certified Casino Property Entry & Exit Point Analyzer *Baccarat Winning Session Record: 12 out of 12 & 1 out of 1 Mini Session. Baccarat Losing Session Record: 2 Losing Sessions.
gordonm888
Administrator
gordonm888
  • Threads: 60
  • Posts: 5001
Joined: Feb 18, 2015
Thanked by
unJoncharliepatrick
August 25th, 2020 at 10:16:02 AM permalink
Here are EORs for this sidebet assuming 8 decks and a payout of 200 for a win. This is the effect on player EV of removing one card of a given rank.

Ten: +0.002948
9: + 0.00472
8: - 0.02636
7: + 0.004731
6: + 0.004724
5: + 0.002934
4; + 0.002915
3: - 0.008746
2: + 0.0003557
So many better men, a few of them friends, are dead. And a thousand thousand slimy things live on, and so do I.
gordonm888
Administrator
gordonm888
  • Threads: 60
  • Posts: 5001
Joined: Feb 18, 2015
August 25th, 2020 at 10:49:39 AM permalink
Given a baccarat shoe that is 4 decks remaining except for 4 extra cards with a rank of 8:

A-7, 9: 16 cards each
8: 20 cards
Tens: 64 cards

I calculate that the player EV for the sidebet with this shoe composition is +0.10264553.

If you lower the number of 8s in the above shoe from 20 cards to 18 cards (only two excess 8's) the calculated player EV on the sidebet is 0.013387138.

That was what I meant in my earlier post where I mentioned "10% surplus 8's." We would expect roughly 16 8's to remain in this shoe (above) but if the actual number of 8's is approximate 10% higher than 16 than the house edge will be approximately 0.0.
So many better men, a few of them friends, are dead. And a thousand thousand slimy things live on, and so do I.
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
August 25th, 2020 at 11:01:30 AM permalink
Quote: ssho88

...counting systems.

a) Unbalanced running count system, tags from Ace to T(+1, +1, -3, +1, +1, +2, +2, -10, +2, +1), bet when RC >= 34. Please note that RC trigger point will change slightly if the penetration change.

I tried a modification of the first count as the program need a balanced count (-10.7/-3.3) and based on 4.8 or better it gave a better return than the simple count (based on 90% penetration). Obviously 3's help the dealer get a 3 so it is a factor, which your count includes.

My original figures incorrectly used profit per hands wagered rather than watched, so the profit should have been about $300. Using the count above it goes up to about $350.
Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26432
Joined: Oct 14, 2009
August 25th, 2020 at 12:21:44 PM permalink
Quote: gordonm888

Here are EORs for this sidebet assuming 8 decks and a payout of 200 for a win. This is the effect on player EV of removing one card of a given rank.

Ten: +0.002948
9: + 0.00472
8: - 0.02636
7: + 0.004731
6: + 0.004724
5: + 0.002934
4; + 0.002915
3: - 0.008746
2: + 0.0003557



Thank you!
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
MattUK
MattUK
  • Threads: 13
  • Posts: 138
Joined: Jul 25, 2017
August 25th, 2020 at 3:36:51 PM permalink
If you're going to break Hippodrome and Horizons, the maximum bet on 3G8 in both casinos is £50, winning 10k.
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
August 25th, 2020 at 4:26:41 PM permalink
Thanks for the EORs, I've based the third version of counting on these. The counts where bets are made are 4.0 and 8.9.
AKQJT54 : 1.0 976 : 1.6 8 : -8.9 3: -3 2: 0.1
OddsPenetrationProfit/100 hands ($100)Times wager made
200/1
90%
$356.27
23.01%
180/1
90%
$160.31
10.54%
gordonm888
Administrator
gordonm888
  • Threads: 60
  • Posts: 5001
Joined: Feb 18, 2015
August 25th, 2020 at 5:13:26 PM permalink
Quote: charliepatrick

Thanks for the EORs, I've based the third version of counting on these. The counts where bets are made are 4.0 and 8.9.
AKQJT54 : 1.0 976 : 1.6 8 : -8.9 3: -3 2: 0.1

OddsPenetrationProfit/100 hands ($100)Times wager made
200/1
90%
$356.27
23.01%
180/1
90%
$160.31
10.54%



FYI, the EORs for a payout of 180/1 are almost exactly 10% lower than for 200/1. Not very surprising.
So many better men, a few of them friends, are dead. And a thousand thousand slimy things live on, and so do I.
ssho88
ssho88
  • Threads: 53
  • Posts: 657
Joined: Oct 16, 2011
August 25th, 2020 at 5:23:19 PM permalink
Quote: charliepatrick

Thanks for the EORs, I've based the third version of counting on these. The counts where bets are made are 4.0 and 8.9.
AKQJT54 : 1.0 976 : 1.6 8 : -8.9 3: -3 2: 0.1

OddsPenetrationProfit/100 hands ($100)Times wager made
200/1
90%
$356.27
23.01%
180/1
90%
$160.31
10.54%



Here are the EORs(Ace to T) generated by combinations analysis, see link below.

https://ibb.co/4KFFjG5
gordonm888
Administrator
gordonm888
  • Threads: 60
  • Posts: 5001
Joined: Feb 18, 2015
August 25th, 2020 at 6:21:35 PM permalink
Well I am using a comb. math spreadsheet that was designed for a much more complicated bacarrat calculation, and I see I had a weird error in data entry that affected my calculations by a small amount. I am now reproducing Ssho88's numbers which are


Rank
EOR (pay 200)
EOR (pay 180)
Ten
0.002519
0.002268
9
0.004159
0.003745
8
-0.02451
-0.02207
7
0.004187
0.003771
6
0.004181
0.003747
5
0.002636
0.002374
4
0.002506
0.002257
3
-0.008293
-0.007468
2
0.002536
0.002284
A
0.002519
0.002268
So many better men, a few of them friends, are dead. And a thousand thousand slimy things live on, and so do I.
ssho88
ssho88
  • Threads: 53
  • Posts: 657
Joined: Oct 16, 2011
August 25th, 2020 at 8:53:59 PM permalink
Quote: gordonm888

Well I am using a comb. math spreadsheet that was designed for a much more complicated bacarrat calculation, and I see I had a weird error in data entry that affected my calculations by a small amount. I am now reproducing Ssho88's numbers which are


Rank
EOR (pay 200)
EOR (pay 180)
Ten
0.002519
0.002268
9
0.004159
0.003745
8
-0.02451
-0.02207
7
0.004187
0.003771
6
0.004181
0.003747
5
0.002636
0.002374
4
0.002506
0.002257
3
-0.008293
-0.007468
2
0.002536
0.002284
A
0.002519
0.002268



It’s amazing, you created this EOR calculator in such a short time !
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
August 26th, 2020 at 5:31:43 PM permalink
Thanks for the above. As a mathematical exercise (and also to have some coding for other things perhaps) I developed a program to go through all the possible combinations of cards (e.g. 0 0 0/0 0 0 to 9 9/9 9) counting the combinations for the number of decks (8 here) and, where appropriate, one card removed for a given value.
Parms: ndx:8 card removed:99 sidebetpays:200 Time:1:3:4:722
Overall Result:
TotHands: 4998398275503360
Banker: 2292252566437888
BEV: 0.9894209421575281
BHE: 1.0579057842471928
Player: 2230518282592256
PEV: 0.9876491867108339
PHE: 1.2350813289166096
Tie: 475627426473216
SideBetWon: 22708577366016
SEV: 0.9131773418174763
SHE: 8.682265818252366
Parms: ndx:8 card removed:99 sidebetpays:200 Time:1:3:4:802
xHE means percentage House Edge for easier reading.

I get very similar numbers so was wondering whether this is the correct way to work out EoRs.
I counted a picture as "0", so it's in a different order to yours!
Removed Card
House Edge
Difference
None
8.682 266%
0.000 000%
Ten
8.430 368%
0.251 898%
Ace
8.430 368%
0.251 898%
2
8.428 635%
0.253 631%
3
9.511 548%
-0.829 282%
4
8.431 670%
0.250 596%
5
8.418 674%
0.263 592%
6
8.264 200%
0.418 066%
7
8.263 542%
0.418 724%
8
11.133 015%
-2.450 749%
9
8.266 331%
0.415 935%

If you add a card instead.
Added Card
House Edge
Difference
None
8.682 266%
0.000 000%
Ten
8.933 155%
-0.250 889%
Ace
8.933 155%
-0.250 889%
2
8.934 861%
-0.252 595%
3
7.868 000%
0.814 266%
4
8.933 583%
-0.251 318%
5
8.944 703%
-0.262 437%
6
9.096 655%
-0.414 390%
7
9.097 298%
-0.415 032%
8
6.255 498%
2.426 768%
9
9.096 247%
-0.413 982%

After thought - the "difference" looks the wrong way round, it was because I was looking at the change as it affected the pllayer. So a House Edge dropping from 8.68% to 8.43% was good by 0.25% for the player.
Last edited by: charliepatrick on Aug 26, 2020
TinMan
TinMan
  • Threads: 20
  • Posts: 445
Joined: Nov 17, 2009
August 27th, 2020 at 8:05:47 PM permalink
Quote: Wizard

The question for the poll is would you bet the 3 Giving 8?



I would not. It also seems like a very complicated bet to explain succinctly. It's basically a 3-step parlay (banker 3, player must draw 3rd card and 3rd card 8).

And there's nothing elegant about it--e.g. there are BJ side bets that pay high return if player gets 2 As and dealer gets As (along those lines). That's visually obvious. If this were a BJ side bet it would be something like (slightly exaggerated) "Player get 4c4s, player splits, player gets exactly one 10 and dealer bust with a 25." It's too many different disparate parts.
If anyone gives you 10,000 to 1 on anything, you take it. If John Mellencamp ever wins an Oscar, I am going to be a very rich dude.
DogHand
DogHand
  • Threads: 2
  • Posts: 1485
Joined: Sep 24, 2011
Thanked by
gordonm888
August 31st, 2020 at 1:55:11 PM permalink
Quote: gordonm888

Well I am using a comb. math spreadsheet that was designed for a much more complicated bacarrat calculation, and I see I had a weird error in data entry that affected my calculations by a small amount. I am now reproducing Ssho88's numbers which are


Rank
EOR (pay 200)
EOR (pay 180)
Ten
0.002519
0.002268
9
0.004159
0.003745
8
-0.02451
-0.02207
7
0.004187
0.003771
6
0.004181
0.003747
5
0.002636
0.002374
4
0.002506
0.002257
3
-0.008293
-0.007468
2
0.002536
0.002284
A
0.002519
0.002268



gordonm888,

My 8-deck, 200:1 EoR's agree with these values.

Using these, I produced balanced tags of +3 for 0,1,2,4, and 5; +5 for 6, 7, and 9; -10 for 3; and -29 for 8; or 3,3,-10,3,3,5,5,-29,5,3 for A-X.

I then ran a 10-million-shoe sim with 15-24 cards cut off, but with one more round after the cut card appears (this is what Eliot used for his baccarat sims): this gave an average of 81.28 rounds per shoe.

For the 200:1 paytable, the SB is +EV for +13 and up, which was 24.93% of the rounds. When played, the sidebet hit rate was 0.59%, which gave an EV of +18.73%.

For the 180:1 paytable, the SB is +EV for +30 and up, which was 11.88% of the rounds. When played, the sidebet hit rate was 0.67%, which gave an EV of +20.58%.

Hope this helps!

Dog Hand
TinMan
TinMan
  • Threads: 20
  • Posts: 445
Joined: Nov 17, 2009
August 31st, 2020 at 2:07:25 PM permalink
Quote: DogHand

Quote: gordonm888

Using these, I produced balanced tags of +3 for 0,1,2,4, and 5; +5 for 6, 7, and 9; -10 for 3; and -29 for 8; or 3,3,-10,3,3,5,5,-29,5,3 for A-X.

Dog Hand



Impressive work!

Separately, FWIW, I’m sure some people could keep this count but to me this seems really challenging. More complicated than any BJ counting scheme ive heard of.
If anyone gives you 10,000 to 1 on anything, you take it. If John Mellencamp ever wins an Oscar, I am going to be a very rich dude.
TinMan
TinMan
  • Threads: 20
  • Posts: 445
Joined: Nov 17, 2009
August 31st, 2020 at 2:23:17 PM permalink
It just occurred to me that you may be able to use the scorecard to make keeping the count easier. Not ideal to have physical evidence of your counting but with a well thought out camouflage possibly indecipherable
If anyone gives you 10,000 to 1 on anything, you take it. If John Mellencamp ever wins an Oscar, I am going to be a very rich dude.
  • Jump to: