CompRookie: Q for Cacarulo/Don

[CompRookie]

Hi, guys, I notice that on p. 394 of Blackjack Attack, the expectations for single-deck games have a number in parentheses for the Total-Dependent lines. Is the number in parentheses the figure for when standing on 77 v T is included? (Does the foonote apply to only the number in parentheses, or to the entire TD-EV line?) I tried reproducing this result, and I am indeed standing on 77 v T, but I am getting the numbers that are not in parentheses. For instance, for 1-deck S17 NDAS TD-EV, I get 0.0023% when I include standing on 77. Where is the 0.003058% number coming from? Are you sure these numbers are right? They do not match Cacarulo's numbers that are on bjmath.

Also, is Cacarulo sure that the 8-deck H17 TD-EV numbers are right? I'm also wondering about the 2-deck H17 TD-EV numbers? I am getting numbers that show a slightly better edge for the player (less negative) in all cases, e.g., -0.3905% for 2-deck H17 DAS TD-EV. Thanks for any insight.

[Don Schlesinger]

> Hi, guys, I notice that on p. 394 of Blackjack Attack,
> the expectations for single-deck games have a number
> in parentheses for the Total-Dependent lines. Is the
> number in parentheses the figure for when standing on
> 77 v T is included? (Does the foonote apply to only
> the number in parentheses, or to the entire TD-EV
> line?) I tried reproducing this result, and I am
> indeed standing on 77 v T, but I am getting the
> numbers that are not in parentheses. For instance, for
> 1-deck S17 NDAS TD-EV, I get 0.0023% when I include
> standing on 77. Where is the 0.003058% number coming
> from? Are you sure these numbers are right? They do
> not match Cacarulo's numbers that are on bjmath.

The extra edge, for the number in parentheses, as the footnote indicates, comes from standing on 7,7 v. T, rather than not distinguishing it from other 14s.

> Also, is Cacarulo sure that the 8-deck H17 TD-EV
> numbers are right? I'm also wondering about the 2-deck
> H17 TD-EV numbers? I am getting numbers that show a
> slightly better edge for the player (less negative) in
> all cases, e.g., -0.3905% for 2-deck H17 DAS TD-EV.
> Thanks for any insight.

Perhaps Cac will answer. I can't give any further insight except to say that all of the numbers have been checked dozens of times, so Cac is only going to tell you that they're all correct.

Don

[kc]

> Hi, guys, I notice that on p. 394 of Blackjack Attack,
> the expectations for single-deck games have a number
> in parentheses for the Total-Dependent lines. Is the
> number in parentheses the figure for when standing on
> 77 v T is included? (Does the foonote apply to only
> the number in parentheses, or to the entire TD-EV
> line?) I tried reproducing this result, and I am
> indeed standing on 77 v T, but I am getting the
> numbers that are not in parentheses. For instance, for
> 1-deck S17 NDAS TD-EV, I get 0.0023% when I include
> standing on 77. Where is the 0.003058% number coming
> from? Are you sure these numbers are right? They do
> not match Cacarulo's numbers that are on bjmath.

> Also, is Cacarulo sure that the 8-deck H17 TD-EV
> numbers are right? I'm also wondering about the 2-deck
> H17 TD-EV numbers? I am getting numbers that show a
> slightly better edge for the player (less negative) in
> all cases, e.g., -0.3905% for 2-deck H17 DAS TD-EV.
> Thanks for any insight.

I get -0.3917% for 2-deck TD, H17, DOA, NS, SPL3, DAS

kc

[CompRookie]

check 1-deck S17 DOA NoRSA RS4 NoDAS (Total-dependent), with and without the standing on 77 v T? (Don't look in BJA3, though!)

Thanks for the other number. That's a big help. I have an idea what is wrong with my numbers.

[kc]

> check 1-deck S17 DOA NoRSA RS4 NoDAS
> (Total-dependent), with and without the standing on 77
> v T? (Don't look in BJA3, though!)

> Thanks for the other number. That's a big help. I have
> an idea what is wrong with my numbers.

I get 0.0023% when hitting 7-7 v 10 to a total of 17 and +0.0031% when standing on 7-7 v 10.

kc

[CompRookie]

a comibinatorial analyzer, right? i am assuming u are not using a simulator.

Thanks for the numbers. You have been quite helpful.

[Don Schlesinger]

> a comibinatorial analyzer, right? i am assuming u are
> not using a simulator.

> Thanks for the numbers. You have been quite helpful.

I guess I'm not understanding. KC's numbers are identical to Cac's,, which I already told you were precise. So, how, exactly, has KC been helpful? I suppose you needed someone else to say the same thing to believe me, huh? :-)

Don

[Magician]

> I guess I'm not understanding. KC's numbers are
> identical to Cac's,, which I already told you were
> precise. So, how, exactly, has KC been helpful? I
> suppose you needed someone else to say the same thing
> to believe me, huh? :-)

I see the smiley at the end but, to some people, having someone prove to them that something is correct is a lot more helpful than having an expert tell them that it is.

[Don Schlesinger]

> I see the smiley at the end but, to some people,
> having someone prove to them that something is
> correct is a lot more helpful than having an expert
> tell them that it is.

K.C. didn't PROVE anything; he simply reiterated the numbers that Cac furnished in BJA3. You don't PROVE results of a CA; you state them! And, while it may be "helpful" to some of you if ten other people state their CA results, and they agree with Cac's, it really doesn't add anything to our knowledge, does it? :-)

Don

[CompRookie]

OK, then, if Cacarulo is infallible, then why do his numbers in BJA3 not match his own previous numbers that are on bjmath to this day? Even a combinatorial analyzer can have shortcuts, misspecifications, or bugs. Having a second, independent researcher state identical results is part of the scientific process, and useful to other researchers. If you could bet your life against $1 million dollars that Cacarulo's numbers, as they appear in BJA3, are correct, I suspect that you would not take the bet.

[Don Schlesinger]

> OK, then, if Cacarulo is infallible, then why do his
> numbers in BJA3 not match his own previous numbers
> that are on bjmath to this day?

Because they weren't generated in the same manner. Each set is correct for what it purported to do.

> Even a combinatorial
> analyzer can have shortcuts, misspecifications, or
> bugs. Having a second, independent researcher state
> identical results is part of the scientific process,
> and useful to other researchers.

Suit yourself. Cac's numbers have been out there and verified by several other researchers, for the past several years. That you aren't aware of this doesn't make it any less true. I already was aware of it and tried to share that information with you. Apparently, you needed to hear it from K.C., so I assume you're happy now.

> If you could bet your
> life against $1 million dollars that Cacarulo's
> numbers, as they appear in BJA3, are correct, I
> suspect that you would not take the bet.

I wouldn't bet my life for any amount of money, and I don't need the million dollars. But, if you'd like to propose a more meaningful bet as to whether the numbers are accurate or not, I'm all ears.

In the meantime, this is a somewhat unproductive and meaningless discussion. I suspect you're just a bit peeved that your figures didn't match Cac's, and you were hoping that he was wrong and you were right. I'm sorry to disappoint you.

You chose a handle that indicates that perhaps you're new to all of this, so I have a suggestion, which you're free to take or leave. Be humble when you question the work of a legend. Consider that you may be the one who is wrong and Cac may be the one who is right.

I challenged Peter Griffin on more than one occasion. Almost all the time, I was wrong. Occasionally, I got lucky (floating advantage), but I always realized that I was a distinct underdog, when we disagreed.

Don

[kc]

> OK, then, if Cacarulo is infallible, then why do his
> numbers in BJA3 not match his own previous numbers
> that are on bjmath to this day? Even a combinatorial
> analyzer can have shortcuts, misspecifications, or
> bugs. Having a second, independent researcher state
> identical results is part of the scientific process,
> and useful to other researchers. If you could bet your
> life against $1 million dollars that Cacarulo's
> numbers, as they appear in BJA3, are correct, I
> suspect that you would not take the bet.

fwiw, it would have been better if your follow up question was posted beneath my original post, but you never know! .

kc

[Cacarulo]

> Hi, guys, I notice that on p. 394 of Blackjack Attack,
> the expectations for single-deck games have a number
> in parentheses for the Total-Dependent lines. Is the
> number in parentheses the figure for when standing on
> 77 v T is included? (Does the foonote apply to only
> the number in parentheses, or to the entire TD-EV
> line?) I tried reproducing this result, and I am
> indeed standing on 77 v T, but I am getting the
> numbers that are not in parentheses. For instance, for
> 1-deck S17 NDAS TD-EV, I get 0.0023% when I include
> standing on 77. Where is the 0.003058% number coming
> from? Are you sure these numbers are right? They do
> not match Cacarulo's numbers that are on bjmath.

Others have already answered your questions except for one:
The EV-tables on bjmath are for "OPTIMAL" play and that's why the EVs are better.

Sincerely,
Cac