Lucky Lucky side bet pay table change

[BlackHead]

Also I read your post over at GC some time ago it really helped out.


Sent from my iPhone using Tapatalk

[BJGenius007]

BlackHead,

I began by calculating the house edge and the Effect of Removal (EoR) for each of the various ranks by applying Combinatorial Analysis in Excel. Here are those results for a 6D game (they'd differ a bit for 8D):

Pay Table S21 10:1

Original Pay Table

Hand
Paytable

Hand
Paytable

Suited 777
200

Suited 777
200

Suited 678
100

Suited 678
100

Unsuited 777
50

Unsuited 777
50

Unsuited 678
30

Unsuited 678
30

Suited 21
10

Suited 21
15

Unsuited 21
3

Unsuited 21
3

Any 20
2

Any 20
2

Any 19
2

Any 19
2

Loser
-1

Loser
-1







Edge
-5.3120%

Edge
-2.6623%

A
-0.2670%

A
-0.2845%

2
0.2624%

2
0.2700%

3
0.2032%

3
0.2093%

4
0.1427%

4
0.1467%

5
0.0922%

5
0.0952%

6
-0.3330%

6
-0.3284%

7
-0.6417%

7
-0.6377%

8
-0.4012%

8
-0.4002%

9
0.0816%

9
0.0795%

10
0.2151%

10
0.2125%

J
0.2151%

J
0.2125%

Q
0.2151%

Q
0.2125%

K
0.2151%

K
0.2125%

As shown, the effect of lowering the payout on the Suited 21 to 10:1 from the original 15:1 produces an overall player's advantage of -5.3120% vs. the original -2.6623%. However, this change has very little effect on the EoR values, so the New Lucky Lucky Count (NLLC) tags (as I named them when I posted my original research on this topic back in 2009 on the Green Chip pages at http://www.bj21.com/greenchip/misc_b...lucky-results/) of -1, +1, +1, 0, 0, -1, -2, -2, 0, +1 will still function adequately.

I then ran two 1-billion-round CVData sims for a 6D, S17, DA2, DAS game with 4.75/6 pen (so 65 cards cut off) for a heads-up AP playing one spot and flat-betting $100. In the first sim, the AP bet $100 on the sidebet on every round. Those results showed that the new paytable requires a NLLC true count of +4 to be +EV (as you correctly guessed elsewhere in this thread). In the second sim, the AP still flat-bet $100/rd on BJ and, when the TC was +4 or higher, he also bet $100 on the sidebet. Here are the sidebet results from this second sim:

Units Bet
# Won
EV
WR

TOT
85,457,584
21,262,333
4.38%
$37.40

34
1

-100.00%
$0.00

33
1

-100.00%
$0.00

32
2
1
100.00%
$0.00

31
5
2
40.00%
$0.00

30
17
3
-47.06%
$0.00

29
45
12
371.11%
$0.00

28
63
9
33.33%
$0.00

27
202
70
173.27%
$0.00

26
301
85
75.42%
$0.00

25
551
153
23.78%
$0.00

24
1,391
390
48.81%
$0.01

23
2,205
579
25.13%
$0.01

22
4,168
1,111
29.97%
$0.01

21
7,968
2,197
37.16%
$0.03

20
15,839
4,170
35.87%
$0.06

19
26,873
7,123
31.61%
$0.08

18
47,606
12,673
28.41%
$0.14

17
80,712
21,315
25.06%
$0.20

16
140,084
36,707
21.63%
$0.30

15
223,668
58,403
20.56%
$0.46

14
370,912
96,558
19.22%
$0.71

13
617,299
159,422
17.50%
$1.08

12
977,002
251,641
14.77%
$1.44

11
1,467,176
375,548
13.11%
$1.92

10
2,386,035
609,058
11.75%
$2.80

9
3,635,245
922,700
9.69%
$3.52

8
5,586,487
1,409,321
7.94%
$4.44

7
8,220,134
2,060,888
6.22%
$5.11

6
12,891,410
3,209,843
4.39%
$5.66

5
19,148,704
4,741,195
2.94%
$5.62

4
29,605,478
7,281,156
1.28%
$3.78

As these results show, the AP earns $37.40/hr from the sidebet. When he places the sidebet, he has an average edge of 4.38%, and he wins just under one-quarter of his sidebet wagers. Unfortunately, he loses more $37.40/hr by flat-betting and playing B.S. for the BJ bet, so overall he's losing $5.67/hr. Now if you could find a game with better rules (say LS, for instance), better pen (recall I used 4.75/6 pen), or where you can bet MORE on the SB than on the BJ, you could get +EV even from the new, bad, paytable. Conversely, worse rules (H17, 8D) and worse pen (even additional players) will hurt.

Hope this helps!

Dog Hand

P.S. If you need help running the sidebet in CVData, just ask and I can talk you through it.

P.P.S. Norm, am I missing some better method of posting Excel spreadsheets?

The solution is the husband-wife team. Husband uses Hi-Lo and wife uses the Lucky-lucky count. Husband tells wife how to play her hand. Husband mimics his wife's side bet.