scobee 1: Floating Advantage and playing SAFE

[scobee 1]

I have found a casino that recently has begun placing the cutcard at the 5.25-5.5 level in their six-deck shoe game consistently. Because of this, the principles of floating advantage have suddenly become more than theoretical to me. The rules are S17, DA, DS, resplit to 4 spots, no RSA.

If I read Chapter 6 correctly, (tables 6.21a through 6.24a...page 98) your suggestions concerning bet sizing seem ambiguous. You have left the 'units bet' column open on many of the Chapter 6 tables apparently to stimulate the reader to re-inforce his comprehension of the floating advantage as it applies the more deeply one plays into the remaining cards in the shoe. I understand (I think) that a +2 TC during the first two decks dealt, is worth approximately 2/3 as much as a +2 would be worth at the 4.5 level. In this case, the SAFE play would be increase your units bet from 2 to 3 units. But if one has the additional advantage of a very deeply dealt shoe, and the increasing effectiveness of Basic Strategy, why wouldn't one's per hand expectation call for a doubling of the 2 unit bet to 4 units?

It appears that the advantage has doubled instead of increasing by half. If one should be so fortunate as to find a TC of >+5 at the end of this 5.25 or 5.5 game, why not increase the 12 unit bet called for to a maximum bet of 24 units? Why do I feel like I am missing something basic in evaluating 'floating advantage?

I am not a probability theorist. Having limited analytic tools (and temperament) I am confused as to why your recommended maximum bet remains the same. While the SAFE would suggest that in this rich environment, the usual ruinous risk associated with doubling your bets beyond the TC to bet ratio is nullified, why remain in the same ramp and maximum bet when the per hand expectation has improved from 3.02 to 6.0?

To summarize, I guess I am asking this:

1. Does the standard deviation remain the same as one gets deeper into the shoe that has this excellent penetration?

2. Is it the variance that makes for a negative expectation when you make a conditional play to double your usual 'units bet to TC ratio' ?

3. If one were lucky enough to find such a good shoe game, would it be feasible to just play with a 1-24 unit spread? Of course one would have to suppose that cover would become ever more a priority with this plan.... if it were indeed supported by computational analysis, which it may not be. Perhaps the smarter play would be to just double the unit size and remain with a 1-12 betting ramp from -1 to +6 TC. This game seems especially worthy of some extended analysis.

Sorry if the proposal is not clear... statistics are definitely a second language for me. What I need to know is would the SAFE in this game really give me the edge to go hog wild at the end of the shoe, if I have an extemely favorable TC.

Thanks for your patience with my pidgen english theorem. I would think that with these specific conditions, the additional risk would be worthwhile considering the probable payoff over time, but I am not sure.

scobee

[ET Fan]

[Don Schlesinger]

You can't imagine how much pleasure it gives me to actually have someone resurrect "SAFE" (an all but moribund concept!) and be talking about the floating advantage as a practical, rather than theoretical, concept!

When we deal more deeply into the shoe, we indeed do get greater edge towards the end, and that increase can be substantial. No simulators that existed back then permitted us to actually explore the correct bet variations, per shoe segment, the way we can now, with genial software such as Norm's CVCX.

But, the idea is not to simply double your spread, more deeply into the shoe, which would be dangerous, from a cover standpoint, but rather to understand that, because the edge that permits a 12-unit bet now comes earlier in the true count, we may make that max bet sooner, and, therefore, more frequently, towards the end of the shoe.

This, in turn, leads to substantially greater profits, without having to increase spread.

If Norm is reading this (and he will, because I'll send it to him!), perhaps he can tell us if, in its current form, CVData or CVCX can actually handle a change in the bet schedule depending on the floating advantage.

Don

[Norm Wattenberger]

Yes, CVData reports on floating advantage. You can divide the deck into eight sections by setting seven card numbers as borders. You can also define betting ramps by depth. You define up to six deck sections and attach a different betting strategy to each section. I just ran a one billion round sim with the specified rules using Hi-Lo Illustrious 18. The links below provide some results. The tables are a bit difficult to read on the web and the chart loses some color.

Floating Advantage Tables
Floating Advantage Chart

I agree that you would alter the ramp by deck depth, not the spread. For now you would have to determine the optimal ramp by trial and error. Determining the optimal ramp algorithmically is an interesting problem. I'll try to look at it when I have time.

regards,
norm

[scobee 1]

You're making me laugh again, old pro.

Actually, I am thinking of taking up the JaKal Method....it has all the research data of Target and the authentic observation of more than a hundred hours of success in real casino conditions.... although, unfortunately, no witnesses. :-(

If you must know, and I know you must, this was something that I first noticed as an anomaly while re-reading the 2nd edition. So it was actually the research that led to the proposal, not the other way around. While I was trying to soak up some wisdom during my hiatus, I found that the tables didn't quite make sense to me.

My imperfect understanding of The Book of Schlesinger, as well you know, is not just limited to Chapter six.

Since you have weighed in on the subject, so to speak, what would your optimal betting strategy be for this game?

[scobee 1]

I agree that you would
alter the ramp by deck depth, not the spread.

I am actually glad to hear that optimal spread for this shoe game remains the same. I was thinking I might have to dance at the table and guzzle Mike's Hard Lemonade to create some cover for such a radical spread. I am presently experimenting with your fine demo software on my emulated Windows '98 OS and am favorably impressed.. enough to send in an order soon. It is amazing to find such penetration around here, I was shocked to see it continue unchecked on my last visit.

The chart looks a little like a rainbow iceberg...if I can just put it in the path of this game we may have a Titanic event ahead of us.

Thanks for your time and attention..

scobee

[ET Fan]

I took one of your comments to mean you wanted to double your ramp, all through the shoe from beginning to end. It's been a while since I studied The Floating Advantage, but I'd be amazed if it supports doubling even at the 5.5/6 level.

Be aware that if you just keep your ramp as it is, you still profit from the FA. If your bet is a little low, risk goes down, and your score (lower case) goes up -- just not quite as much as it would if you ramped perfectly.

Also be aware that if you scale your bets according to a tool such as BJRM, the FA is already built in. If you increase your bets a little toward end of shoe, then you should really DEcrease your bets a little toward the start. Otherwise you are undermining the ROR calculation. I actually do this. I hold back a nickel 'till about 2.5 decks remaining, then at the 2d level I start adding a nickel. NOT REQUIRED and not even very precise. Just too lazy to map out the exact numbers.

Only a small part of the value of deep pen comes from FA. The main benefit is much greater volatility in the count, late in the shoe, leading to many more large and max bet opportunities. This should not increase your long run ROR, but will increase your session ROR or "trip ruin." Ie. Take more money with you. You're on a faster roller coaster.

I guess if I had a 5.5/6 game available, I'd check out the FA chapter again. Wouldn't lose too much sleep over it, though. ;-)

ETF

[ML]

The floating advantage tracks the advantage of fewer decks at a zero count. So, the gain is significant but not that great. The normal six deck rules give about -.3 for six decks and about +.12 for one deck so you pick up about .4 as things get deeper. Since .4 is close to one count advantage, a rough method would be to drop your bets about one level from start to end if about one deck is left. So if you were making your top bet at +4 it could probably be made at +3 when under a deck left. But you have to remember, as Griffin points out, the advantage is not linear. Most of the change in advantage occurs below two decks because an inverse function is in place.

For the hell of it, I am going to throw out my speculations on the floating advantage one more time even if the theory has not met with the enthusiastic approval it deserved when posted before.

This speculation is based on the differences in amounts of advantage one count difference makes in situations of spectacularly low or spectacularly high counts. Years ago I ran a series of simulations using stacked single decks. I then compared the expectations between the runs for -10 and -9 etc. and looked at the differences. Looking at the differences show a distinct tendency. As the count increases, there is, of course, a gain in expectation but the amount of gain decreases regularly, using basic strategy which is how the floating advantage is efined. As a practical matter this means that between -9 and -10 there is a decrease of expectation of about one per cent but between +9 and +10 the increase is only about .1 per cent. In other words, a -10 takes away advantage much more than a +10 count gives advantage for basic strategy. But we know we are going to encounter the -10 count as often as we will encounter the +10 count or any minus count as often as we will encounter the equivalent plus count. So, overall, the negatives are going to take away more advantage than the equivalent positives gain.

At about the same time John Auston posted an interesting study which demonstrates my point. He dealt five decks from a six deck stack and tabulated the count in the remaining deck. Of course there was a bell curve of counts and both +10 and -10 happened with some frequency. Of course the last deck dealt could just as easily be considered the first fifty-two cards dealt or the third set of fifty-two cards dealt. Characteristics would be identical. So when we take the expectation of that first hand off a six deck stack which defines expectation for a six deck game, there is some probability we are playing that hand with a stack of 52 cards at the top which have a count of +10 or -10. And if it happens to be a -10 stack we are playing at considerably more disadvantage than we would be our advantage if we were playing from a stack of 52 cards at the top which happened to have a +10 count. Advantages do not balance even if counts do so overall disadvantage worsens as more decks are shuffled together.

We say the floating advantage is at a zero count so it is present when five decks have been dealt and the count is still zero. That is understandable. All the outre possibilities have now been discarded. The first hand in this last deck is being dealt from something similar to a "perfect" deck without the possibility the next fifty-two cards, in fact, are not comprised of 52 cards with a -10 count or a +10 count and thus the overall expectation is better.

IMHO this analysis explains both the floating advantage and the lowered expectations associated with more decks.

Parenthetically, I am about sure basic strategy and indexes should be modified at least in this last deck. If it is correct to double 9 v 2 in single deck, it would also seem to be correct to double 9 v 2 at the one deck level and a zero count.

[ET Fan]

I don't think the floating advantage is exactly the same as the "count of zero phenomenon" Griffin wrote about. Don has mentioned in previous threads that FA occurs at all counts (or maybe it was all positive counts, not sure).

One thing I'm prepared to dispute, though, is that basic should change based on penetration. It mat seem contradictory, but as long as there are enough cards to complete the round, basic doesn't change a whit, even though the correct play at TC = 0 may change.

ETF

> Parenthetically, I am about sure basic
> strategy and indexes should be modified at
> least in this last deck. If it is correct to
> double 9 v 2 in single deck, it would also
> seem to be correct to double 9 v 2 at the
> one deck level and a zero count.

[ML]

True basic does not change. The better way of putting it as you did, the correct play at a zero count with one deck remaining most likely is true IMHO. Also IMO indices would change to mirror SD indices.

[Don Schlesinger]

> I don't think the floating advantage is
> exactly the same as the "count of zero
> phenomenon" Griffin wrote about. Don
> has mentioned in previous threads that FA
> occurs at all counts (or maybe it was all
> positive counts, not sure).

Correct. The floating advantage applies to all counts, not just the count of zero. That's easy to see (at least for the positive counts) in the FA article.

> One thing I'm prepared to dispute, though,
> is that basic should change based on
> penetration. It may seem contradictory, but
> as long as there are enough cards to
> complete the round, basic doesn't change a
> whit,

Also correct.

> even though the correct play at TC = 0
> may change.

That's the interesting part. Once you've counted 5/6 decks, and you have a count of zero, that's no longer the same thing as a basic strategist who joins a game at the 5/6 level. He has no more reason to play the next hand any differently than he would off the top of the pack, and we've been over the reason why a million times.

But, suppose you're counter who, nonetheless, uses only BS to make the plays. when you reach 5/6, and the count is zero, should you now use SD BS or multi-deck BS?

That answer is less obvious, although we have discussed it before, and I lean towards still using multi-deck BS.

Don

[ML]

I was not suggesting the floating advantage does not rise at all counts. What I meant to describe was the fact that if expectation at zero rises, expectation everywhere rises and suggest the rise is about .4 from 6/6 to 1/6.

As to the other question we have differred before and will again I suspect if the topic is brought up. Let me throw out one thought. We both know the remaining zero count one deck is unlikely to be a "perfect" deck but is not the median deck of the possibilities a "perfect" deck and should we not shoot for the median?

[ET Fan]

> But, suppose you're counter who,
> nonetheless, uses only BS to make the plays.
> when you reach 5/6, and the count is zero,
> should you now use SD BS or multi-deck BS?

5/6 with a count of zero is really only comparable to 0/1, if 5/6 is zero before the round is dealt, or would have been zero at 5/6 if you subtract your cards and dealer's upcard after the round is dealt. In this case, the deflection from full deck composition, just from the cards in your hand, and dealer's upcard, is closer to 0/1 basic than 0/6 basic. So my hunch is that 1 deck basic would be more accurate here.

I don't put this forth as mathematical proof. I understand the problem. I just suspect that the large deflection -- which is automatically reflected in 1d basic -- is a greater factor than the imperfect knowledge of composition at 5/6 with a count of zero.

> That answer is less obvious, although we
> have discussed it before, and I lean towards
> still using multi-deck BS.

Well, I do too, but my reason is LAZINESS! ;-) Also, I rationalize that multi-deck BS is risk averse (fewer doubles).

I think MathProf is planning on putting something about a "counter's basic" in his book. I'll bet it would be possible to come up with a one-size-fits-all basic that's "better," from a RA point of view, than any EV maximizing basic that's tailored to the # of decks. Give us lazy people a hook to hang our hat on...

ETF