I'm trying to track down betting EORs for three special cases:
1)Dealer has an ace
2)Player receives an ace
3)Neither dealer nor player get an ace
Can anyone point me in the right direction?
thanks yet again,
Myooligan
I'm trying to track down betting EORs for three special cases:
1)Dealer has an ace
2)Player receives an ace
3)Neither dealer nor player get an ace
Can anyone point me in the right direction?
thanks yet again,
Myooligan
> I'm trying to track down betting EORs for three
> special cases:
> 1)Dealer has an ace
> 2)Player receives an ace
> 3)Neither dealer nor player get an ace
> Can anyone point me in the right direction?
> thanks yet again,
> Myooligan
Betting EORs do not vary according to playing situation, they represent the absolute value of the cards for a specific game. It is just like another point count system, except it uses fractions, and the values (divided by remaining decks minus initial house edge) will give you your exact edge at any given time.
See page 522 of BJA3 paperback for the most (only) accurate sets of numbers on this issue.
Hope this helps,
TAO
> Betting EORs do not vary according to playing
> situation, they represent the absolute value of the
> cards for a specific game.
You are right that traditionally, they do not. However, it is easy to see that, given the special situation of knowing the value of one your cards (or the dealers) before the round is dealt, a new set of EORs can be constructed. For instance, if you know one of your cards is an ace and you also know the deck is ten heavy, you want to bet high. But if you know one of your cards is going to be a six in the same situation, extra tens in the deck would be less favorable. Clearly the EOR for a ten is different in these two cases.
I don't know that anyone has looked into this matter, but I imagine someone has a program that could be easily adapted to analyze this. That's what I'm really hoping for. I unfortunately don't have the programming skills to write it myself.
What you are talking about is playing EORs, which is a serie of custom made counts for every situation you might encounter, and this can be found in pp 506-521 of BJA3 pap. or chapter 6 of Griffins TOB. But you really have to see those charts for yourself as I cannot give you a useful answer to the specific questions you describe.
If you don't have the book, those pages alone are well worth the price of the book, and so is p.522.
I was told that my original post is difficult to understand, so I'm going to try again:
I'm trying to figure out the EORs - not the edge - under those special situations. For instance, given a dealer ace up(Case 1 below), what is the change in EV for taking a two out of the deck? a three? a four? etc. I'm not talking about the change in the difference between the hitting and standing EVs (that would be the playing EORs), but rather, the change in overall EV for the hand.
Hopefully this makes more sense, although it still sounds pretty convoluted when I reread what I've written.
> I'm trying to track down betting EORs for three
> special cases:
> 1)Dealer has an ace
> 2)Player receives an ace
> 3)Neither dealer nor player get an ace
> Can anyone point me in the right direction?
> thanks yet again,
> Myooligan
TAO,
Thanks for responding. I have TOB but the playing EORs can't provide the information I need(see my clarification post below). Again, those numbers tell you how much more favorable hitting is than standing (or doubling than hitting, etc.). So for instance if you have 16 v. a dealers 10, extra 8s in the deck will likely make both hitting and standing more unfavorable. The EORs in the TOB tables tells you how much more favorable the former change is over the latter. But I don't want to compare hitting versus standing; I want to know the know how much the value of the bet changes in absolute terms.
I haven't seen the BJA tables, but I suspect they won't have the data I'm after either. Simply put, the EORs I'm after are useless for typical card-counting, because they presume knowledge of receiving a precise card before the round is dealt. First off, the BJA numbers would have to go beyond Griffin's numbers by showing not just comparative EORs but also the absolute effect on the EV for the given hand. Second, Schlesinger's figures are presumably tailored towards card-counting, as Griffin's were, and therefore would omit EORs for decisions where never in a million years would a card counter deviate from basic strategy. But I would need to know those, too, to derive the EORs I'm after from those tables.
I think I'll need to get ahold of a combinatorial analysis program. . . still hoping someone with more experience can tell me where to find one. . . or join with me in the research.
> What you are talking about is playing EORs, which is a
> serie of custom made counts for every situation you
> might encounter, and this can be found in pp 506-521
> of BJA3 pap. or chapter 6 of Griffins TOB. But you
> really have to see those charts for yourself as I
> cannot give you a useful answer to the specific
> questions you describe.
> If you don't have the book, those pages alone are well
> worth the price of the book, and so is p.522.
> TAO,
> Thanks for responding. I have TOB but the playing EORs
> can't provide the information I need(see my
> clarification post below). Again, those numbers tell
> you how much more favorable hitting is than standing
> (or doubling than hitting, etc.). So for instance if
> you have 16 v. a dealers 10, extra 8s in the deck will
> likely make both hitting and standing more
> unfavorable. The EORs in the TOB tables tells you how
> much more favorable the former change is over the
> latter. But I don't want to compare hitting versus
> standing; I want to know the know how much the value
> of the bet changes in absolute terms.
> I haven't seen the BJA tables, but I suspect they
> won't have the data I'm after either. Simply put, the
> EORs I'm after are useless for typical card-counting,
> because they presume knowledge of receiving a precise
> card before the round is dealt . First off, the BJA
> numbers would have to go beyond Griffin's numbers by
> showing not just comparative EORs but also the
> absolute effect on the EV for the given hand. Second,
> Schlesinger's figures are presumably tailored towards
> card-counting, as Griffin's were, and therefore would
> omit EORs for decisions where never in a million years
> would a card counter deviate from basic strategy. But
> I would need to know those, too, to derive the EORs
> I'm after from those tables.
> I think I'll need to get ahold of a combinatorial
> analysis program. . . still hoping someone with more
> experience can tell me where to find one. . . or join
> with me in the research.
I had an idea for a new line of blackjack research which would be like "EORs of EORs", and the basic idea of this is that as the decks get depleted, the unbalances created in the card values would ultimately affect the playing EOR values of the remaining cards. I think this would be great to devellop a working understanding of the game. I also think that if there is such a thing as intuition in blackjack, that this is where it would be rooted. I speculate that Griffin had a similar idea roaming in his mind as was studying subsets.
It sounds to me like your idea is very similar and I would love to do or see research on this fascinating subject, but I am not sure I would have the ability to do very much about it.
Don, what do you think about this subject, and the feasability of such an endevor?
Thanks in advance,
TAO
> I had an idea for a new line of blackjack research
> which would be like "EORs of EORs", and the
> basic idea of this is that as the decks get depleted,
> the unbalances created in the card values would
> ultimately affect the playing EOR values of the
> remaining cards.
You're right that EORs are not accurate later on in the deck (technically, they're only valid for the removal of one single card from the top of the deck). But I don't know how much this line of pursuit could help a person's understanding of the game. . . Off the top of the deck, there are 10 possible EORs. But suppose you want to know your exact EOR for removing any two cards (two-card composition-dependant EORs?. Each of those 10 EORs will have 10 more "EORs of EORs," as you put it. So at the two card level, you have 100 EORs to get your mind around. At the three card level, you get 10^3 =1000, and 4 cards deep you have 10^4 = 10,000 different effects to think about. Then it gets more complicated, since you have to start considering the possibility that one denomination gets removed completely from the deck.
Patterns would surely emerge in the course of such a study, but you'd ultimately have to use a complex multi-parameter system to take advantage of them, so I suspect the gain from having a "feel" for how some of the subsets work would be offset by the lack of mathematical precision available for simple counting.
My math was wrong. . . if you combine matching combinations (2 and 4 is no different from 4 and 2) you get 10+9+8+ . . . +1 = 55 EORs at the two card level. I don't know the numbers after that, but they're gonna be high . . .
First of all you are right that you'd have to use a complex multi-parameter system to take advantage of them, and I don't beleive that this information could be used in the same way that index plays may be used to gain an immediate advantage. It would be more just a question of establishing certain guidelines for certain types of situations.
Example: I beleive that if the deck is very rich in aces, and the count is negative, and you have a stiff vs. a stiff, that the values of the EORs would be amplified in favour of hitting, since the combination of low cards and aces would synergestically work to prevent the dealer from busting.
But OTOH your concept of multiplying the amounts of EORs for every card seen is completely wrong. The EORs are quite simply a point count system that is generally valid for any remaing subset.
This line of pursuit could help one's underderstanding of the game in the sense that when you are being whipped by lady luck, you will have a better understanding of the forces at work, and ultimately, if this line is succesful you might learn to recognize some potentially desastrous situations and escape them before you are torn to shreds.
TAO
> Example: I beleive that if the deck is very rich in
> aces, and the count is negative, and you have a stiff
> vs. a stiff, that the values of the EORs would be
> amplified in favour of hitting, since the combination
> of low cards and aces would synergestically work to
> prevent the dealer from busting.
Sounds reasonable.
> But OTOH your concept of multiplying the amounts of
> EORs for every card seen is completely wrong.
Nope, didn't say that. Originally I thought squaring the number of single-card EORs would tell you how many possible two-card combination EORs are available; that formula is wrong but the general thinking is correct. Not talking about multiplying one EOR times another, which would give you a meaningless figure, as you say.
> The EORs
> are quite simply a point count system that is
> generally valid for any remaing subset.
They are generally reasonable estimates, I agree, although as I read it this totally contradicts your original idea:
> the basic idea of this is that as the decks get
> depleted, the unbalances created in the card values
> would ultimately affect the playing EOR values of the
> remaining cards.
In the latter case you're talking about shifting EOR values while the former presumes constant ones. Which ones do you mean?
> In the latter case you're talking about shifting EOR
> values while the former presumes constant ones. Which
> ones do you mean?
The EOR are constant values, what I am suggesting is a study that would discover exeptionnal situations in which the values would be shifting. And this does in fact exist, Griffin talks about certain weird subsets where either the player or the dealer have astronomical edges, there are reasons for this, it is just that we don't understand them.
TAO
Exceptional situations with an astrononimical edge use to exist in the old days of single deck dealt down to the last card. These conditions no longer exist. Maybe that's why there's not much interest or participation in the thread.
An interesting simple strategy of EOR that exist in todays world that you can use' is you're at 6-deck game and dealt a pair of aces. The dealer shows an ace and ask "Insurance? You notice the other three players have nothing but core cards(7,8,9's). Three decks remain and the true count is 2. The Hi-lo simulations says dont take insurance but they are wrong. The correct EV decision is take insurance. This is an example of a practical way to improve your game by EOR in today's world.
Computer simulations are only as smart as what they are programmed to do. Some will say, you MUST play according to simulations and any deviation is wrong. They are incorrect in what they believe to be true.
So one may ask the crazy question "Is it possible to play better than a computer?" The answer is yes,very possible.
Keep in mind if the above strategy is used there will be no HUGE advantage but a significant one.
Posting Permissions
-*-
Bookmarks