hit 12 against 4 in a (well-balanced) 1 deck game !!!

Hello,

is there any other Software out there, that advises to hit with hard 12 against the 4 in a 1 deck (still well-balanced) game, as my Software (Black Jack Advocate) does?

The reason for that it is better to hit is prosaically, that by hitting a small card you not only betters your hand but also increase the probability of the dealer to bust; if you hit a ten-counting card you would decrease the Probability of the dealer to bust, but as you already lost in that case, this has no influence to your expected value. Only the decreasing of the dealers probability to bust after you hit a small card makes a difference. And in a 1 deck game this difference is enough to make hitting to the better move. (You find the mathematically prove for that at http://www.bj-advocate.com/Calculation/Calc_Example.htm )

I would be interested if there is any other software with the same result. If not, I guess my software is the only one that calculates correct.

Ups...
Please change in my sentence:
>>Only the decreasing of the dealers probability to bust after you hit a small card makes a difference.<<
the word >>DEcreasing<< to INcreasing.
It was a finger error, and I unfortunately didn't use your Password-Function.
As I am german, its not so easy for me to write in english. I am better in math ;-)

Thanks.

I didn't see were you where removing the two cards from the players hand to make up the 12 total. The probabilities are also effected if the player's 12 is made up of T,2 (most likely) or 8,3 for example.

Interesting software however. How are you adjusting the deck if you would want to check a probability against an unbalanced deck?
Can you specify cards remaining and a count?

It seems to make no sense to calculate the expected value of any hand against any dealer's up card with the assumption that the shoe would be still complete - because if the shoe is complete, there is no hand and up card at all, and the calculated situation can never occur.

This is true, but try to imagine, that one deck more has been in the shoe, and that exactly this deck has been drawn out, and that the last 3 cards has been your hand and the dealers up card; this situation is possible and therefore the calculation make sense for a totally well-balanced shoe with the specified total of decks remaining.

In this case, it is better to hit, whatever your hard 12 is composed of (except of course 6 and 6 ;-)

With my software it is just a click on a button to adjust the remaining cards in the shoe to a specified situation, say 2 and Ten (and the dealers up card as well). The calculation for the advise and the expected value for this distribution of remaining cards is calculated instantly, off course using the same formula by considering the cards that hit.

For instance, your hand is 2 and Ten, dealers upcard is 4:
Expected Value (Hit)= -19.54
Expected Value (Stand)= -21.06
->you should hit.
(remaining cards before hitting: three of 2's, three of 4's, 15 of Ten-counting-cards, 4 of all the others),(Dealer hits soft 17)

In all the other cases (Players Hand 3 and 9, 4 and 8, 5 and 7), it you adjust the remaining cards in the same way (what is very easy with my software), you should stand (except of course 6 and 6, what you should split).

A count is unfortunately not yet implemented, but it will be soon in the next review.

l. Hit 10,2 v. 4 Hit in One and Two Decks if H17. Hit in seven or fewer decks if S17.

Also m. Hit 10,2 v. 6 in one deck if S17.

Basic strategy assumes removal of the cards in the player's hand and the dealer's hole card. This is why you are getting different answers. Using your method of calculation you might find differences in the following hands also:
13 vs 2
9 vs 3
10 vs 9
A,8 vs 6
A,7 vs 2
A,6 vs 2

Using HiLo count T,2 vs 4 top of the deck the count is +1. The index for T,2 vs 4 single deck is +1, so yes you should hit.

How do you plan on altering the deck composition to reflect certain counts? Peter Griffin in "The Theory of Blackjack" did some work on this. Basically, you need to add/remove an equal number of high and low cards and leave the density of neutral cards alone.

It is very interesting to calculate a totally well balanced deck. A totally well balanced is normally the implicitly assumption of a count of 0.

So, the question is, if your hand is hard 12 (what ever the cards are, it might be that you already have 3 cards), and dealers up card is 4, and your actual count is 0 (assumption: shoe is well-balanced), and there is around 1 deck left in the shoe, is it than better to hit or to stand. And I say, then it is better to hit! I know, that this might be not easy to accept, because BS seemed to be a holy dogma... (please dont misunderstand. I don't want to say BS is simply wrong, I say that in the case that I discribed here to hit is better that to stand).

I guess the common approach to calculate is only an estimation. A good one, but an estimation. My approach instead is another one and result in the correct value using theory of probabilities.

Please refer my site http://www.bj-advocate.com/calculation/calc_example.htm , where this is explained more detailed.

Ok, BS assumes removal of the cards in the player's hand and the dealer's hole card, but does it also take account of the cards that could be drawn if you hit and therefore recalculates the dealer's probability after hit, as my software does?

I talked about a situation, when the count is 0. BS says it is better to stand, and I say that the exact calculation results in the advise to hit (1 deck, well-balanced remaining cards in the shoe, H17 or S17, whatever your hand is composed of)

In other words: if you have hard 12 (composed of 3 cards for example), and the dealer has 4, and your actual count is 0 (then you assumes to have the well-balanced shoe), then, whatever you have in your hand, the correct advise is to hit.

Youre right, this does not only touch 12 against 4, but this is the one case for instance.

The things I plan on altering the deck composition to reflect certain counts I have many ideas, and I will integrete them one after another. I will explain that in more detail on my website.

You are trying to calculate something to the sixth decimal place that doesn't occur. If you have a multi deck game and you have a 12 vs 4 hand with 52 cards remaining (this is the situation you are talking about, right?) Even if the count is exactly zero at this point, What are the odds that these remaining 52 cards consists of exactly 4 twos, 4 threes, 4 fours, etc. This is what you are calculating. This is only one arangement that is possible with a 0 count deck and it is no more likely that a deck that has 8 twos and light 4 ten count cards, 2 extra sevens, and only 2 nines. (Which will probably give you quite a different answer.)

Your program could be very valuable if you adjust the remaining deck composition correctly by a user inputted count and a remaining card level. But, I'm sorry, you haven't discovered the meaning of life here.

Please post what modifications you are planning on making here, and let's talk about it. There are many knowledgeable posters here that would probably be glad to help you get a useful product.

you said above:

This is true, but try to imagine, that one deck more has been in the shoe, and that exactly this deck has been drawn out, and that the last 3 cards has been your hand and the dealers up card; this situation is possible and therefore the calculation make sense for a totally well-balanced shoe with the specified total of decks remaining.

This is N decks remaining. The rationale that you are trying use has no importance. We don't talk about a two deck game where exactly 52 cards remain with the exact composition of a fresh single deck. You are essentially talking about a basic strategy play if you are talking about full balanced decks. The cards in the player's hand and the dealer's upcard do make a difference. As the number of decks increase these few cards become less important. But they are quite important in single deck.

It is possible that you are saying something that makes sense but that I have totally missed your point. I apologize if that is the case.

You're right, thats exactly what I mean. Of course every different arrangement with count 0 will lead to different expected values. And this arrangement is exactly the same what BS is based and is calculated on. And BS says, that you should stand with hard 12 against a 4 (1 deck, well balanced), and I say you should hit.

The best would be of course to make a frequency weighted mean value of every possible arrangement, each value calculated by the approach I use. I think something like this will be integrated in a later state of my software.

Excuse me, I didn't understand what you mean with that I havn't discovered the meaning of life here. Is that an american proverb?

I will post here more, and my goal is to make a valuable software, and I am happy for any discussion and to learn people that probably be glad to help me to get a useful product.

:-)

Yes, I think you did not get the point. It is very possible that the reason for that is my bad english.

To explain why the cards in your hand don't make a diffence for the advise to hit or to stand in a special case, please imaging the following impossible but instructible situation:

You have a Ten-counting-card and a 2, and the dealer has a 4. In the shoe (just imagine) there remain only Ten-counting-cards. The expected value of hit or stand depends only on the Ten-counting-cards: you should stand, of course. There would be no difference if you would have a 9 and a 3 in your hand, or a 2, a 3 and a 7 !!!!

The calculation of the expected value for a given hand-total depends ONLY on the cards that you didn't see (e.g. two burned cards) and don't see (the remaining cards in the shoe).

The composition dependent exception to basic strategy you mean assumes that the shoe has been well-balanced or full before drawing the first three cards. I am talking instead of the situation, that the shoe is well balanced AFTER drawing the first three cards.

And of course I don't calculat N decks, because I change the probabilities after cards dealt out (but not for the cards that are already dealt out, as explained above).

I apologize for my bad english and the bad example...

Consult the back of Wong's Professional Blackjack for the correct two card hand strategy for various numbers of decks.

There is the generic number 12 (or any other number you might select) which gives the probabilities of drawing from a full deck. I think Griffin may have been the first to calculate that way although perhaps he was anticipated.

I think that is what you are describing.

From this generic probability, Griffin then calculated rates of removal. He calculated how things would change if a 4 is removed (as it is) from the deck to calculate the probabilities of hitting and standing. The 12,4 play is close so Griffin found a 10,2 12 should be hit and other combinations making 12 should not. This makes sense. 10 and 2 removed still leave cards like 5,6,7,8, and 9 which help the hand some of which would be removed if the 12 were different from a 10,2.

What I do not understand and what makes me suspicious your program is not working right is a generic 12 v. 4 should not be hit. That is, if you were playing double deck and at the start of a hand, the dealer had a four and the player had a 10,2; the dealer had a hidden card and 51 cards to deal and these 52 cards comprised a perfect deck, the play would be to stand. If your program says the correct play for this circumstance is hit, your program is wrong. If your program says the correct play is hit for one deck with a 4,2, and 10 removed, it is working right.

"What I do not understand and what makes me suspicious your program is not working right is a generic 12 v. 4 should not be hit."

So far, so good. Then, why do you change from "generic" to 10-2, below? Now, it's a different story!

"That is, if you were playing double deck and at the start of a hand, the dealer had a four and the player had a 10,2; the dealer had a hidden card and 51 cards to deal and these 52 cards comprised a
perfect deck, the play would be to stand."

But we are clear that, if this were still double-deck, but if the same scenario had a hidden card and 48 cards left to deal, the residue being a "perfect" deck, then, of course, we would hit, right?

The point is, by definition, BS tells us what to do only off the top of a freshly-shuffled pack and with the assumption that the three cards in question are removed. If this is not the case, we aren't talking about BS anymore!

Don

2 level count For S17 the index is +2 and H17 is +1

"That is, if you were playing double deck and at the start of a hand, the dealer had a four and the player had a 10,2; the dealer had a hidden card and 51 cards to deal and these 52 cards comprised a
perfect deck, the play would be to stand."

Thank you. This is exactly what I mean (except that the player could have any other combination of cards, e.g. a 2,3,7). I still stay with the opinion, that in this situation (player has hard 12, dealer 4, H17 or S17, one hidden card and 51 cards to deal remaining in the shoe, these 52 cards comprised a perfect deck) it is better to hit - not for pig-headed reasons, but for mathematically reasons. I will prepare a text, where I will explain that. I will prove it (or apologize). Just give me a little time.

ML wrote: "That is, if you were playing double deck and at the start of a hand, the dealer had a four and the player had a 10,2; the dealer had a hidden card and 51 cards to deal and these 52 cards comprised a perfect deck, the play would be to stand."

Don Schlesinger wrote: "The point is, by definition, BS tells us what to do only off the top of a freshly-shuffled pack and with the assumption that the three cards in question are removed. If this is not the case, we aren't talking about BS anymore!"

This is a very interesting point. I think you agree with me, that BS is commonly used if in balanced Card counting systems like Hi/Lo the count is 0. In the above mentioned situation, the count is 0 and I think people would use BS...

But in both cases you should hit in the situation I discribed.

Perhaps ML discribed it better:
"That is, if you were playing double deck and at the start of a hand, the dealer had a four and the player had a 10,2 (or any other hand - HG); the dealer had a hidden card and 51 cards to deal and these 52 cards comprised a perfect deck."

In this case, for both S17 and H17 my advise is to hit.