Robert V. Lux: Accuracy of BS charts in extreme cases

[Robert V. Lux]

When reading BS index charts, the desicion for some specific situations remain constant regardless of the TC. A good example is 11 vs 4, where one should always double no matter what TC.

Though, there are, at least theoretical situations, where it is incorrect to double 11 vs 4. Assume a SD game. All tens are immediately played. You get an 11 vs the dealer's 4. Now, the best possible card you could get in this case is a 6. The chance that the dealer will bust equals 1/6, right?

This was just an extreme demonstration. Though, do the BS index charts only view specific ranges, such as Wong's -20 to +20 ranges in PB?

If not, what's the logic behind this?

/Robert V. Lux

[Parker]

> When reading BS index charts, the desicion
> for some specific situations remain constant
> regardless of the TC. A good example is 11
> vs 4, where one should always double no
> matter what TC.

> Though, there are, at least theoretical
> situations, where it is incorrect to double
> 11 vs 4. Assume a SD game. All tens are
> immediately played. You get an 11 vs the
> dealer's 4. Now, the best possible card you
> could get in this case is a 6. The chance
> that the dealer will bust equals 1/6, right?

> This was just an extreme demonstration.
> Though, do the BS index charts only view
> specific ranges, such as Wong's -20 to +20
> ranges in PB?

> If not, what's the logic behind this?

Basic strategy makes no assumptions of any sort regarding the count. If it did, it would not be basic strategy.

Basic strategy assumes that you have the knowledge of the three cards in front of you (your hand and the dealers upcard) and that is all. The correct basic strategy is the play with the highest expection, knowing only that information.

Obviously, on a close play such as 16 vs 10, correct basic strategy will actually be "wrong" almost half the time. This is why (along with the fact that the hand occurs frequently) this is such an important index.

Equally obvious, many basic strategy plays will be wrong in extreme counts. Since these counts rarely occur, this is of little consequence.

[Robert V. Lux]

Maybe I did not express myself clearly.

In the 11 vs. 4 situation, shouldn't there be a lower limit, like other double/hit situations, on which you should prefer a hit instead of a double, because a double down is not a methematically correct move?

/Robert V. Lux

[humble]

> Basic strategy makes no assumptions of any
> sort regarding the count. If it did, it
> would not be basic strategy.

Well, I suppose it assumes RC=TC=0 (balanced count) as it assumes cards are drawn from the top of a full pack :-).

[paranoid android]

Yes. SBA tells me the index for 11 vs 4 is -13 for the UBZ2 count (IRC=-2). Sorry, but I'm too lazy at the moment to boot into Windows and have SBA compute the index for hi-lo.

Has SBA ever been compiled to run on the Linux OS (or has anyone had success running it under Wine)?

> Maybe I did not express myself clearly.

> In the 11 vs. 4 situation, shouldn't there
> be a lower limit, like other double/hit
> situations, on which you should prefer a hit
> instead of a double, because a double down
> is not a methematically correct move?

> /Robert V. Lux

[Parker]

> Maybe I did not express myself clearly.

> In the 11 vs. 4 situation, shouldn't there
> be a lower limit, like other double/hit
> situations, on which you should prefer a hit
> instead of a double, because a double down
> is not a methematically correct move?

There would be an index, sure. But that has nothing to do with basic strategy.

[Parker]

> Has SBA ever been compiled to run on the
> Linux OS (or has anyone had success running
> it under Wine)?

Due to the fact that there aren't a whole lot of Linux users, and there aren't a whole lot of card counters, and there really are very few card counters who run Linux, I have had no success convincing the developers of SBA, Casino Verite, BJRM2002, etc., to port their products to Linux.

I must admit that I cannot really fault them for this. They don't sell nearly enough copies of the Windows versions of their products to justify the hours they have spent developing them, let alone develop versions for other OS's.

Still, it wouldn't hurt to drop Karel, Norm, John, etc. an e-mail and express your interest.

I have been playing around with Wine off and on for years, and have yet to be able to get one of these programs to even boot up with it, let alone run properly. However, I readily admit that I am no programmer, and not even a very good Lunux hacker, so I am not saying that it isn't possible.

This really belongs on the "Computing for Counters" page. :-)

[Don Schlesinger]

> Well, I suppose it assumes RC=TC=0 (balanced
> count) as it assumes cards are drawn from
> the top of a full pack :-).

Actually, no. It assumes that the three cards in question are already withdrawn from the full pack, so the RC can be, and often is, something other than zero.

Don

[humble]

> Actually, no. It assumes that the three
> cards in question are already withdrawn from
> the full pack, so the RC can be, and often
> is, something other than zero.

> Don

We could both be right.
It depends on what one considers/defines to be the "experiment" begin: the start of drawing cards (and 3 cards is of course in general not the end) or the first decision point (indeed, after 3 cards).

H.

[paranoid android]

BJRM runs for me under a recent version of Wine on Linux almost perfectly. I haven't had any luck with SBA though. It just tells me it hasn't been properly installed and exits.

> Due to the fact that there aren't a whole
> lot of Linux users, and there aren't a whole
> lot of card counters, and there really are
> very few card counters who run Linux, I have
> had no success convincing the developers of
> SBA, Casino Verite, BJRM2002, etc., to port
> their products to Linux.

> I must admit that I cannot really fault them
> for this. They don't sell nearly enough
> copies of the Windows versions of their
> products to justify the hours they have
> spent developing them, let alone develop
> versions for other OS's.

> Still, it wouldn't hurt to drop Karel, Norm,
> John, etc. an e-mail and express your
> interest.

> I have been playing around with Wine off and
> on for years, and have yet to be able to get
> one of these programs to even boot up with
> it, let alone run properly. However, I
> readily admit that I am no programmer, and
> not even a very good Lunux hacker, so I am
> not saying that it isn't possible.

> This really belongs on the "Computing
> for Counters" page. :-)

[Parker]

> BJRM runs for me under a recent version of
> Wine on Linux almost perfectly. I haven't
> had any luck with SBA though. It just tells
> me it hasn't been properly installed and
> exits.

Most interesting. What Linux distribution and version are you running, what version of Wine, and how much did you have to tweak it to get it to work?

[Groucho_00]

> Maybe I did not express myself clearly.

No, I think your readers understand. And I think that you are confusing the definitions. Basic strategy is designed for players who do not wish to learn a card counting system, but wish to limit their losses at the tables.

Your question describes a counting system. Maybe you should consider learning and using a good system to increase your chances at the tables. There's plenty of good information about systems on this site and several others. The very nature of your question infers an interest in learning more about the game. Buy a book. Download some software. Just do it.

My system's index for 11 vs 4? Double if the true count is -15 or greater, else hit.

[paranoid android]

> Most interesting. What Linux distribution
> and version are you running, what version of
> Wine, and how much did you have to tweak it
> to get it to work?

I use Debian up to date (mostly) with the current unstable distribution (my Linux kernel is 2.4.17). I'm currently running Wine version 0.0.20021125 and I have my Windows programs installed on a Win98 partition (mounted read-only - this could perhaps be a factor in some programs not running under Wine). I didn't really do any tweaking at all to get BJRM working, just used the configuration program that came with Wine accepting the defaults for most things.