Lucky Ladies Simulation Request

[Jabberwocky]

Jacobsen has calculated that a modified Ten Count is optimal for LL. I play the KB version in DD but with the 6 deck pay table which requires a higher trigger point before the bet becomes profitable. LL has low vulnerability in shoe games unless the pen is DEEP.

[Tarzan]

Jacobsen has calculated that a modified Ten Count is optimal for LL. I play the KB version in DD but with the 6 deck pay table which requires a higher trigger point before the bet becomes profitable. LL has low vulnerability in shoe games unless the pen is DEEP.

Yes, the pen makes a huge difference in the 6D, 8D game. For example if you follow the index exactly using perfect play and are making 45 cents on every $100 wagered with 5/6, under the exact same conditions with pen altered to 4/6 you are now making 11 cents on every $100 wagered. This difference in pen has a significant effect! It shocked me just how dramatically pen altered the theoretical result.

[Three]

It shocked me just how dramatically pen altered the theoretical result.

It is all related to the changes in TC frequency associated with deeper pen. The quick way to look at it is to look at how pen changes those extreme TC frequencies that you make your bets at. You make more bets and more bets at a higher advantage (higher TC).

[Three]

The reason why I choose to use KO to target this side bet rather than a different count is because I want to beat the game without the side bet.

I will just explain some things rather than bust your chops.

To get a high LL edge you need a count that pits T's as high cards against as many of the other cards as possible with all other cards weighted the same. To play BJ you want to count the aces and faces as high cards and well you know. It is easy to do both and this is an example of where getting past simple counting is of importance in BJ. If you want to target both BJ and the LL bet use an ace neutral count. Your LL count will be very strong. Much stronger than any ace reckoned count. With a side count of aces you can make LL even stronger and have a high PE and high BC count for BJ.

Now onto Tarzan's suggestions about the Q/Qh. The edge of LL depends on the frequency of hitting EACH payout. With counting it is hard to get a correlation for all the payouts and have a practical system. The gain from factoring in the highest payout probability change is huge. If there are lots of Qh left compared to what is expected the bet becomes profitable well below the Qh blind index because the top payouts happen more frequently than the Qh blind TC assumes. The other side of the coin is a deficit of Qh where you have fewer Qh left than are assumed by a Qh blind count (1 Qh left for each deck remaining to be played). If fewer than expected (a deficit of Qh) Qh remain the bet will not be profitable until the index is exceeded by a good bit due to hitting the top payouts less than the TC suggests would happen. By adjusting your betting for these changes in Qh density you tim off long run losing bets you would make otherwise and replace them with profitable long run bets that you would not have made. I assume you understand that the gain from side counting Q/Qh can make big differences in expectation and where that gain in EV is coming from.

You were choosing a bad count for attacking the LL bet (ace reckoned RC approach) in order to supposedly keep a strong BJ game when a strong count for the LL bet would have made your BJ game stronger by increasing PE and BC. Tarzan's 3 column with ace count is a perfect count for achieving both optimally. Of course few could or would go that far to play the game.

Before I get into this you can use IC as an approximate Q/Qh blind lucky ladies correlation. Adding an Q/Qh side count improves LL performance a lot for the reasons previously stated.

If you want to get the best of the BJ game and the LL side bet with something within normal human abilities try this count with a side count of aces:
A-T: 1, 1, 1, 1, 2, 1, 0, 0, -2 with an ace side count with the ace adjustment for betting as -3 per ace seen (not per ace deficit or surplus).
This main count has a PE of .68 and IC of .94 and would be very strong for the LL side bet.
The ace adjusted count for betting becomes an unbalanced count:
A-T: -2, 1, 1, 1, 2, 1, 1, 0, 0, -2
The BC would be .98 but with an imbalance of -12 per deck the imbalanced count would need to be true counted by adding 3 to the betting running count for each 1/4 deck seen and dividing that by the number of decks remaining. Add a side count of Qh and you would be crushing all aspects of the game. The Qh side count would be easy to keep because once enough Qh have been seen the LL bet can be ignored.

An easier version along the same line of thinking would more focus on LL and insurance with increased PE but hurting BC considerably (not a good move for shoe games):
A-T: 0, 1, 1, 1, 1, 1, 1, 1, 1, -2 with a side of aces and Qh.
Ace adjustment -2 per deficit ace for the betting count RC. These are all balanced counts. So the betting calculation is much simpler.
IC .98, PE .61 and BC .89 (ouch).

A good approach for an overall is Hiopt2/ace side count with a side of Qh.

A-T: 0, 1, 1, 2, 2, 1, 1, 0, 0, -2 aces counting as -2 per deficit ace and a side of Qh.
BC .98, PE .67, IC .91

As you can see the latter is not terribly complicated and has top BJ performance with good LL performance.

Now as for KO the stats look like this:
BC .98, PE .55, IC (approximate LL correlation) .78 which doesn't compare well to the other choices.

Basically a strong LL count is ace neutral which also cause strong IC and PE in BJ and must be ace side counted to get the best possible BC. You can make your own decisions from there for the best overall count for you. If you have never tried to develop any advanced counting skillz you may need to develop them or leave a lot on the table if you aren't able to or won't try. Remember all these advances will tighten the bell curves around your decisions which increases EV while lowering risk and increasing your optimal bets.

[seriousplayer]

I will just explain some things rather than bust your chops.

To get a high LL edge you need a count that pits T's as high cards against as many of the other cards as possible with all other cards weighted the same. To play BJ you want to count the aces and faces as high cards and well you know. It is easy to do both and this is an example of where getting past simple counting is of importance in BJ. If you want to target both BJ and the LL bet use an ace neutral count. Your LL count will be very strong. Much stronger than any ace reckoned count. With a side count of aces you can make LL even stronger and have a high PE and high BC count for BJ.

Now onto Tarzan's suggestions about the Q/Qh. The edge of LL depends on the frequency of hitting EACH payout. With counting it is hard to get a correlation for all the payouts and have a practical system. The gain from factoring in the highest payout probability change is huge. If there are lots of Qh left compared to what is expected the bet becomes profitable well below the Qh blind index because the top payouts happen more frequently than the Qh blind TC assumes. The other side of the coin is a deficit of Qh where you have fewer Qh left than are assumed by a Qh blind count (1 Qh left for each deck remaining to be played). If fewer than expected (a deficit of Qh) Qh remain the bet will not be profitable until the index is exceeded by a good bit due to hitting the top payouts less than the TC suggests would happen. By adjusting your betting for these changes in Qh density you tim off long run losing bets you would make otherwise and replace them with profitable long run bets that you would not have made. I assume you understand that the gain from side counting Q/Qh can make big differences in expectation and where that gain in EV is coming from.

You were choosing a bad count for attacking the LL bet (ace reckoned RC approach) in order to supposedly keep a strong BJ game when a strong count for the LL bet would have made your BJ game stronger by increasing PE and BC. Tarzan's 3 column with ace count is a perfect count for achieving both optimally. Of course few could or would go that far to play the game.

Before I get into this you can use IC as an approximate Q/Qh blind lucky ladies correlation. Adding an Q/Qh side count improves LL performance a lot for the reasons previously stated.

If you want to get the best of the BJ game and the LL side bet with something within normal human abilities try this count with a side count of aces:
A-T: 1, 1, 1, 1, 2, 1, 0, 0, -2 with an ace side count with the ace adjustment for betting as -3 per ace seen (not per ace deficit or surplus).
This main count has a PE of .68 and IC of .94 and would be very strong for the LL side bet.
The ace adjusted count for betting becomes an unbalanced count:
A-T: -2, 1, 1, 1, 2, 1, 1, 0, 0, -2
The BC would be .98 but with an imbalance of -12 per deck the imbalanced count would need to be true counted by adding 3 to the betting running count for each 1/4 deck seen and dividing that by the number of decks remaining. Add a side count of Qh and you would be crushing all aspects of the game. The Qh side count would be easy to keep because once enough Qh have been seen the LL bet can be ignored.

An easier version along the same line of thinking would more focus on LL and insurance with increased PE but hurting BC considerably (not a good move for shoe games):
A-T: 0, 1, 1, 1, 1, 1, 1, 1, 1, -2 with a side of aces and Qh.
Ace adjustment -2 per deficit ace for the betting count RC. These are all balanced counts. So the betting calculation is much simpler.
IC .98, PE .61 and BC .89 (ouch).

A good approach for an overall is Hiopt2/ace side count with a side of Qh.

A-T: 0, 1, 1, 2, 2, 1, 1, 0, 0, -2 aces counting as -2 per deficit ace and a side of Qh.
BC .98, PE .67, IC .91

As you can see the latter is not terribly complicated and has top BJ performance with good LL performance.

Now as for KO the stats look like this:
BC .98, PE .55, IC (approximate LL correlation) .78 which doesn't compare well to the other choices.

Basically a strong LL count is ace neutral which also cause strong IC and PE in BJ and must be ace side counted to get the best possible BC. You can make your own decisions from there for the best overall count for you. If you have never tried to develop any advanced counting skillz you may need to develop them or leave a lot on the table if you aren't able to or won't try. Remember all these advances will tighten the bell curves around your decisions which increases EV while lowering risk and increasing your optimal bets.

Tthree thanks for the good reply. I understand that a good LL count need to have high IC and PE. The counts you suggest requires keeping two side counts (ace side count and side count of Qh). I am only capable of keeping one side count in my head along with the running count. I like the first count that you posted:

A-T: 1, 1, 1, 1, 2, 1, 0, 0, -2

[Three]

The counts you suggest requires keeping two side counts (ace side count and side count of Qh). I am only capable of keeping one side count in my head along with the running count.

There are only 6 or 8 Qh in the shoe. It is easy enough to keep a count that stays that low another way than in your head.

[Cougfan]

There are only 6 or 8 Qh in the shoe. It is easy enough to keep a count that stays that low another way than in your head.

And if you happen to play DD with LL the side count it is super easy since it is just a yes/no question - have I seen a Qh since the last shuffle?

[Jabberwocky]

There are only 6 or 8 Qh in the shoe. It is easy enough to keep a count that stays that low another way than in your head.

http://apheat.net/2012/08/08/card-co...jack-side-bet/

[seriousplayer]

At TC+4 if you are far enough into the 8D and have enough surplus Q/Qd then you are in the positive expectation zone to make the LL bet (possibly not by much)... only one problem with all that, you are using KO so go by a RC, not the TC. I would have to convert the KO RC into what I'm used to, a deck composition and a TC to answer that. The person to answer that is someone who uses KO and has studied the LL sidebet. There was something else I saw pertaining to KO and "KO with side", which I assume is an (A) sidecount. KO "with side" fares better than without. You need to find the specifics for what you are doing, for KO and not what people using other counts are doing.

Personally I go by IC+7 using my insurance count rather than betting count. If there are surplus Q, Qh then I would bet it at less than IC+7. As I said, at IC+5, +6, I am two decks into the discard in a 6D game. I've only seen one Q come out of the deck so far and no Qh, meaning there are seven surplus Q's available in the remaining four decks, I'll place some sort of bet on the LL. If it was TC+8 or more and only one Q had come out with four decks remaining I'd bet more (which is usually only $25 max). I use the insurance count because the LL bet becomes linear much like the insurance bet does, so a specific density of {T} to all other cards along with a sidecount of Q is what you'd be going by. The idea of going by just a RC and not TC is difficult for me to fathom, so using KO your betting count is going to be sloppy. I'm guessing that the LL sidebet using KO would be the same, a little on the sloppy side. With that being the case, it wouldn't hurt to look at the exact point of LL being a worthwhile bet using KO and/or go with whatever the KO equivalent is of TC+7 or more (with Q in even distribution) to bet it. If you have the KO equivalent of TC+4 (my insurance count) on an 8D with 4 decks in the discard tray but a huge surplus of Q to include Qh in the remainder then by all means you should wager on the LL bet. If you have the KO equivalent of what would be TC+6 (my insurance count) at that same level of pen but a huge number of Q are depleted to include at least five or six Qh then you wouldn't and would go with a rock solid TC+7 or more to wager on it.

The edge increases with the number of decks due to the increased number of Q and in particular Qh so if it's golden in 6D, it's sure as heck going to be at least that good with 8D. Keep in mind the theoretical edge is based on the infrequent occurrence of some of those QQ and even QhQhvsBJ popping up but it goes up drastically with surplus Q, Qh in the remainder of the deck. If the density of {T} in the remainder hits about 35% or more the LL bet becomes worth wagering on. A sidecount of Q/Qh would be quite the fanciful thing to do if you really want to go hardcore, I suppose. I won't have an exact count of Q, only a ballpark within a card or two. In the heat of battle I already have enough to do. I will have an exact ratio of {T} to other cards using my insurance count. I would probably train to do an exact Q, Qh sidecount if I played LL more often than I do, it's relatively infrequent that I play LL or run into it.

I see people in the table just counting the queen of hearts for the lucky lady side bet. Without using a counting system count could you beat the lucky lady side bet just counting the queen of hearts alone by itself and play using prefect basic strategy?