The Floating Advantage and Basic Strategy

I was reading up a little bit, mainly just out of curiosity, about the concept (The Floating Advantage) and started to think about it and it's implications. It's presented in "Blackjack Attack" by Don Schlesinger, and I'm pretty sure Arnold Snyder's website has a couple articles about it.

For those who are not familiar, (I'd imagine a lot are not) I'll summarize. A card counter keeps a running count of the proportion of high to low cards using tags. Then, they divide this count by the number of decks remaining to get a True Count, an estimate of the expected value of the next hand. It is assumed that a True Count has the same edge throughout the packet, however, it was discovered that this isn't true. In fact, given decent rules, a 6 deck game with a Running Count of 0 has a slight player advantage when only 1 deck is left. The actual effect is pretty much directly correlated to the same rules' edge with "n" decks.

This has no value to the non card counter as the effects of using non-optimal (basic) strategy against high and low counts will eat up this advantage. To the counter, this knowledge has some slight effect, but hardly enough to utilize heavily. (It's probably good to know that at a RC of 0 with 1 deck left has a slight edge, and that higher counts will have a higher edge later in the shoe than earlier if you use the proper index plays)

But here's what I was wondering. The explanation of why this effect occurs is essentially that the remaining deck essentially acts the same as a single deck over time, when the Running Count is 0. So it's like playing against 1 deck, but with the 6 deck rules. That being the case, would you use a single deck basic strategy at a RC of 0? And Double Deck strategy with 2 decks left? Or would you always use the multi deck strategy? Which is better at that point in the deck, and if you do use the "n" deck strategy, is it also best to use "n" deck indices?

Optimally, if I was a perfect blackjack playing machine, would I constantly adjust my indices according to the number of decks left? I realize that constantly adjusting not only the RC, TC, Floating Advantage effect and so called "Floating Basic Strategy and Indices" if they exist may just be overkill, but I'm mostly wondering if it would theoretically be applicable. (Even if not practical)

Has anybody read anything about this? I'm not sure if anybody has ever bothered to do any research on this, but I am looking for any literature I can find.

The Floating Advantage is both very technical and esoteric. Only the most mathematically curious blackjack students need to be aware of it. For the basics, please see my Ask the Wizard column #182.

What I'm trying to figure out is if anybody has ever researched whether the same effect that causes the Floating Advantage would cause you to change your Basic Strategy based on the number of decks remaining. I pretty much understand the concept as described in the Blackjack Attack, and understand that it's not that practically useful.

Mostly out of curiosity, and partly because if these changes exist, they would be easy to learn, I'm wondering whether the effect would cause you to play "n" deck strategy, where "n" is the number of decks left, at a RC of 0.

Say with 1 deck left in a liberal 6 deck game, the RC is 0. You bet, knowing that you're playing at about a .2% advantage if you use Basic Strategy. Would it, however, be more beneficial to you to use the one deck strategy? Say you get 3,5 v 6, the change in index is enough to justify doubling. However, say you get A,8 v 6. In 6 deck you stand. In single deck you double. The RC is still 0, so I think that the same effect that causes the Floating Edge may justify a doubling down in this scenario, but I may be wrong.

As far as I know, though, nobodies ever done any research on that. And it seems to me, since I have most of the differnces between 1, 2 and 4+ deck strategy memorized, it may be worth knowing if this effect is present.

BTW: I have read that before and it does a nice job of explaining and simplifying the known concept, but I'm trying to find if anyone has studied the effect of penetration and Strategy at a neutral count.

deleted

Quote: Boney526

What I'm trying to figure out is if anybody has ever researched whether the same effect that causes the Floating Advantage would cause you to change your Basic Strategy based on the number of decks remaining.

No, you wouldn't. The last deck in a six-deck shoe has the same odds as the first deck, for the non-counter.

deleted

Floating Advantage is an effect of card removal, so it is not governed by the true count (which is an estimate of card distribution).

I have no doubt that there is a better strategy for TC=0 at 1 deck left (i.e. better than basic strategy), but it won't be the single deck basic strategy.

The only question is: how much mathematics are you willing to "spend" ?

Quote: Wizard

No, you wouldn't. The last deck in a six-deck shoe has the same odds as the first deck, for the non-counter.

I realize that, I meant for the Counter. I guess my wording is just off - I just realized I forgot to mention that I meant for the counter, who knows the RC to be exactly 0.

Quote: MangoJ

Floating Advantage is an effect of card removal, so it is not governed by the true count (which is an estimate of card distribution).

I have no doubt that there is a better strategy for TC=0 at 1 deck left (i.e. better than basic strategy), but it won't be the single deck basic strategy.

The only question is: how much mathematics are you willing to "spend" ?

Haha, OK so I'm glad I'm not the only one who thinks that this would effect would probably exist. I'm not anywhere near good enough to do the math myself, I was wondering if anyone had ever seen any work done, or literature about, this topic.

I'm almost certain that there would have to be some effect, although my intuitive idea of it being correlated exactly to # of decks at RC 0 is probably wrong. I do, however, think that many single deck plays would apply if the RC is exactly 0. (Such as Double Soft 19 against 6).

Quote: Boney526

I just realized I forgot to mention that I meant for the counter, who knows the RC to be exactly 0.

"Seems that, at the one-deck level, extremely high counts produce less edge than expected for the basic strategist (many pushes) and the extreme negative counts were found to be even more unfavorable than previously thought (doubles, splits, and stands tend to be disastrous)." - Blackjack Attack, third edition, page 70.

Quote: Wizard

"Seems that, at the one-deck level, extremely high counts produce less edge than expected for the basic strategist (many pushes) and the extreme negative counts were found to be even more unfavorable than previously thought (doubles, splits, and stands tend to be disastrous)." - Blackjack Attack, third edition, page 70.

I guess I'm really, really unclear in my wording. (Haha, I think my wording may be causing you to misconstrue my question, so it's totally my fault if that's the case.)

I know the effect that the Floating Advantage has for the counter who uses Basic Strategy at a count of 0, and indices for other counts. In fact I've read most of the major work on that topic, just out of curiosity rather than practical usefulness.

I'm trying, and failing, to find out if anyone has done work on to find if the same effect that causes the Floating Advantage may cause strategy changes in neutral counts as the deck depletes. As far as I can tell nobody has ever done research on that topic. I'm trying to see if anybody knows if research HAS been done on the topic (neutral count strategy changes as the deck is depleted) and where I can find it. It may very well be that the effect I'm trying to find simply does not exist, but I'm interested because if it does exist, I think it would be easy to tack on to Basic Strategy.

So rather than looking for the effect on the edge for counters and/or non-counters, I'm looking for the effect on optimal strategy during neutral counts due to shoe depletion - if any changes exist at all. (So it would purely affect counters, if any effect even exists.) More specifically, I'm looking for any literature or research that already exists on this topic.

I do appreciate your trying to help me. This is a quircky, and not well known topic. So I didn't expect anyone to really come up with anything regarding my question. I was more or less hoping somebody had some clue where I can look for this (possible) effect on strategy changes.

" it would be easy to tack on to Basic Strategy." perhaps you might want to research Basic Strategy first. It is the best option under all conditions. Even Thorpe was aware that the count had a positive or negative effect. Basic is to limits your losses if you know ABSOLUTELY nothing else. That's all it is !

deleted

Quote: Boney526

So rather than looking for the effect on the edge for counters and/or non-counters, I'm looking for the effect on optimal strategy during neutral counts due to shoe depletion - if any changes exist at all. (So it would purely affect counters, if any effect even exists.) More specifically, I'm looking for any literature or research that already exists on this topic.

In other words: should you use single-deck strategy when there is 1 deck left in a shoe game at neutral counts?

I don't know...

Quote: dwheatley

Quote: Boney526

So rather than looking for the effect on the edge for counters and/or non-counters, I'm looking for the effect on optimal strategy during neutral counts due to shoe depletion - if any changes exist at all. (So it would purely affect counters, if any effect even exists.) More specifically, I'm looking for any literature or research that already exists on this topic.

In other words: should you use single-deck strategy when there is 1 deck left in a shoe game at neutral counts?

I don't know...

Doesn't matter, the casino is not dealing that last deck anyway.

Quote: Buzzard

" it would be easy to tack on to Basic Strategy." perhaps you might want to research Basic Strategy first. It is the best option under all conditions. Even Thorpe was aware that the count had a positive or negative effect. Basic is to limits your losses if you know ABSOLUTELY nothing else. That's all it is !

That's not true. If the count is high or low then the optimal strategy is different than basic strategy. I think that there would be some effect on optimal strategy based on the shoe's current penetration, although the effect, if it exists is likely small. Obviously, I know how to play Blackjack properly, I'm wondering if this obscure topic has an effect that I don't think has been studied, has been.

I'm still looking for any work that's been done on the subject (deck depletion and it's effect on optimal strategy in neutral counts.) I'm pretty sure that nobody has done work on that, though.

Trust me, I have basic strategy memorized. And most of the important indices, as well. But I figure that IF some obscure effect existed between deck depletion and optimal strategy - I'd like to know about it.

Quote: Buzzard

Quote: dwheatley

Quote: Boney526

So rather than looking for the effect on the edge for counters and/or non-counters, I'm looking for the effect on optimal strategy during neutral counts due to shoe depletion - if any changes exist at all. (So it would purely affect counters, if any effect even exists.) More specifically, I'm looking for any literature or research that already exists on this topic.

In other words: should you use single-deck strategy when there is 1 deck left in a shoe game at neutral counts?

I don't know...

Doesn't matter, the casino is not dealing that last deck anyway.

True enough. But it's a hypothetical question meant which, if true, would have other implications. And I don't necesarilly mean to say that it you would optimally use One Deck BS, only that there may be some changes to Optimal Strategy (similar to the existance and use of indices.) I'm more or less asking "does the effect that causes the floating advantage also cause floating indices?"

Huh - I guess that's the a much better way to phrase it. I'm looking for any research on that.

"does the effect that causes the floating advantage also cause floating indices?"

Yes if you are counting. No if you are using Basic Strategy. Maybe if ahigh is dealing the cards.

Quote: Ibeatyouraces

Only if you knew the exact distribution of each seperate rank. Card counting only gives an estimation. It is not exactly precise.

I would agree - but for some reason I keep thinking that there must be SOME changes to the strategy based on deck depletion, or perhaps an effect on the indices themselves. An intuitive example is A,8 v 6. I think, but do not have the ability to prove, that with 1 deck left and a RC of 0, the proper play would be to double regardless of the number of decks the game started with.

My ventures in card counting have, so far, been completely out of hobby rather than actual play. As in - I haven't yet put into practice counting in a live casino, so even IF I do, I likely wouldn't use any advanced theories until I was completely comfortable with basic counting, betting accurately and all the indices I need.

Because of that - I haven't bothered to buy any simulation software. If I ever do, I may try to see if I can find what I intuitively believe may give me a slightly larger edge. If I never do confirm my intuition, I obviously will just use the traditional playing strategies with indices that have been proven to be accurate most of the time, if I ever choose to count cards in live play.

Quote: Buzzard

"does the effect that causes the floating advantage also cause floating indices?"

Yes if you are counting. No if you are using Basic Strategy. Maybe if ahigh is dealing the cards.

It does? Can you reference me to any research on "floating indices?"

In case anyone was wondering - so far this post is all I found....


http://www.bj21.com/boards/free/free_board/index.cgi?noframes;read=109585

Quote: Boney526

but for some reason I keep thinking that there must be SOME changes to the strategy based on deck depletion, or perhaps an effect on the indices themselves.

A counting strategy is what it is. It is a designed strategy which performes better than basic strategy and is designed to be easy to use. It makes no sense saying "there must be SOME changes", because a counting strategy is a FIXED strategy by definition.

There is no question that other strategies are performing better. The (computer) perfect strategy for example, based on the exact statistics of all unseen cards. But the perfect strategy is not a counting strategy, and thus the perfect strategy does not have indices. The perfect strategy also does *not* change with deck depletion, because deck depletion is already been taken into account.

Quote: MangoJ

The perfect strategy also does *not* change with deck depletion, because deck depletion is already been taken into account.

I'm not convinced this is 100% accurate. I play & count against 8 deck shoes, and it hadn't really occurred to me until now that the OP might have a point.

Take a pair of 7s against a 10. In single-deck, you surrender, partly because of the known lack of 7s. In multi-deck, it's just a hit. When I'm down to 1.5-1 deck in my 8 deck shoe, at a neutral count, should I hit or surrender? I'm essentially playing against a 1 deck shoe that is randomly composed, but on average, it will look like a regular single deck.

There is no way that counting indices take this into account. At least, Hi-Lo does not have an adjustment table for how many decks you are currently playing against.

deleted

Quote: dwheatley

There is no way that counting indices take this into account.

In fact we perfectly agree here. Maybe I didn't make it clear what I would understand as "perfect strategy": the precise (i.e. computer) calculation of the EV for every possible playing option accoring to the given set of rules and observed cards, and selecting the option yielding the best EV to the player.


From my understanding, the original question was: "Does a counting system at TC=0 and 7of8 deck penetration perform worse than at TC=0 and zero penetration ?"
If this is answered positive, then there must exist a different counting system (i.e. with modified index tables) which is optimized to the state of 7of8 penetration, and performs better (but may fail on zero penetration). The improved performance at 7of8 penetration may then stated as "Floating Advantage".

My educated guess is: Yes this statement is true, simply for the fact that a counting system based on TC is a simplification with the assumption, that the removal of a single card does not alter the card distribution. This assumption is very good at low penetration, but very poor at deep penetration.
So yes, you can create a different counting system optimized to 7of8 penetration, and you could probably publish an academic paper on this topic.

" So yes, you can create a different counting system optimized to 7of8 penetration, and you could probably publish an academic paper on this topic. "

And academia would rather you not waste their time !

deleted

Quote: MangoJ

From my understanding, the original question was: "Does a counting system at TC=0 and 7of8 deck penetration perform worse than at TC=0 and zero penetration ?"

Yes. Blackjack Attack by Don Schlesinger has a whole chapter (six) devoted to the Floating Advantage. He has lots of tables showing the advantage by count and where in the shoe that count is. It does not indicate any strategy deviations account to Floating Count. At the end of the day it is an esoteric topic, and understanding it doesn't help your game much, if at all.

Quote: MangoJ

A counting strategy is what it is. It is a designed strategy which performes better than basic strategy and is designed to be easy to use. It makes no sense saying "there must be SOME changes", because a counting strategy is a FIXED strategy by definition.

There is no question that other strategies are performing better. The (computer) perfect strategy for example, based on the exact statistics of all unseen cards. But the perfect strategy is not a counting strategy, and thus the perfect strategy does not have indices. The perfect strategy also does *not* change with deck depletion, because deck depletion is already been taken into account.

Obviously I meant that there must be some possible changes to your playing strategy based off of the deck's depletion. \A computer could play perfectly, whereas a person would have to use some combination of most profitable and most probable to decide on strategy deviations. I wonder, though, if any indices change significantly enough at a the 2, 1.5 and 1 deck levels to warrant memorizing the "floating indices."

Oh well, it seems I'll never find them. Although I'm now very confident they do exist, as I've found other references to them online, but no specifics.

As I said, if I ever do buy simulation software I may run a simulation to find out how much indices float relative to shoe depletion.

Quote: Ibeatyouraces

You would still hit 7,7 vs 10 at this point.

I would, as well, only because I have no proof that standing is a better play. Intuitively, though, I'd think with 1 deck left and a RC of 0, it may be superior to stand. I really wish I could find some work on this, because if my idea is right, I really don't think adding a few indices that kick in "at the x deck level" would be hard. I also don't think changing a couple indices at certain deck levels wouldn't be too much effort.

deleted

I realize that you would only stand if you knew the amount of 6/7s left, according to most conventional counting strategies. The more I think about it, the more I think the correlation would be incredibly small. I do, however, think that it's possible a few index plays float according to a relatively strong correlation to decks remaining.

I'm not advocating anybody changed their play based off of my thought, obviously I have no proof of this concept, although I have found it referenced elsewhere online.