BlackJack Simulator/Analyst

#1
1. Internet Blackjack Basic strategy are all same, but I search
The internet no blackjack Analyst to run thousands hands to compare
Analyst report, I was a mathematics teacher and retired software engineer, I wrote A CGI scripts by user set users 2 cards and dealer face card start with
Basic strategy found it is good but not best, so I generate a advance
Strategy go all test against basic strategy to make a little improvement.
2. Basic strategy estimated edge 0.46% advance strategy has edge about 0.32% Click last line web site to test your own Strategy include high-low card count
on program to test the difference for free.
(Dead link: http://bj21.tearsjoys.org/)
 

Sonny

Well-Known Member
#2
cjjssc said:
...I search The internet no blackjack Analyst to run thousands hands to compare Analyst report
There are many software simulation programs available. Programs like Blackjack 6-7-8, Smart Cards, Sage BJ, Blackjack Risk Manager, Universal Blackjack Engine, Professional Blackjack Analyzer, and the entire Casino Verite software suite (Casino Verite, CVCX, CVData, CVShuffle, etc.) all handle this type of thing. The CV products are the industry standard and by far the fastest and most flexible. There are also pure strategy programs like BeeJack 2021, and a few programs by members of this website that handle perfect playing decisions based on combinatorial analysis of the exact deck composition. Looking at some of those programs could save you a lot of time.

-Sonny-
 

blackjack avenger

Well-Known Member
#4
Possible Answer

jack said:
Does anyone know if there are any software programs, thats tells you the efficiency of a hand, using a paticular count?

For example XvsX, using Zen @.80%, While using [email protected]% efficiency, for XvX.
The count with the higher playing efficiency should play indices better. Provided you know a lot of them. The fewer the decks or flatter the spread the more this matters. If you are playing many decks and wonging/spreading the counts with the higher betting correlation dominate and indices are not as important

I probably did not help at all. :joker::whip:
 
#5
Share my BlackJack Simulator/Analyst web site

Sonny:
Thanks for your information, but all the software are not free.
I am retired software and created the web site just for fun.
It may not have good pictures but doing the same thing as you have
to pay software, I like to share my:) web site to you
(Dead link: http://bj21.tearsjoys.org/)
I am good player too, I like the information about bj21 tournaments
information on S California Casino
cjjssc

Sonny said:
There are many software simulation programs available. Programs like Blackjack 6-7-8, Smart Cards, Sage BJ, Blackjack Risk Manager, Universal Blackjack Engine, Professional Blackjack Analyzer, and the entire Casino Verite software suite (Casino Verite, CVCX, CVData, CVShuffle, etc.) all handle this type of thing. The CV products are the industry standard and by far the fastest and most flexible. There are also pure strategy programs like BeeJack 2021, and a few programs by members of this website that handle perfect playing decisions based on combinatorial analysis of the exact deck composition. Looking at some of those programs could save you a lot of time.

-Sonny-
 

jack.jackson

Well-Known Member
#6
blackjack avenger said:
The count with the higher playing efficiency should play indices better. Provided you know a lot of them. The fewer the decks or flatter the spread the more this matters. If you are playing many decks and wonging/spreading the counts with the higher betting correlation dominate and indices are not as important

I probably did not help at all. :joker::whip:

Thanks Avenger, but unfortunatley thats not quite what im looking for. Im looking for something more along the lines, for the efficiency of hands, with paticular counts. I was told from ETfan that Griffin talks about this in theory of blackjack and that he gives you the formula on how to do it. I just wish there was an easier way, such as a analyzer. Talk to you later;)
 

KenSmith

Administrator
Staff member
#7
cjjssc said:
Basic strategy found it is good but not best, so I generate a advance Strategy go all test against basic strategy to make a little improvement.
2. Basic strategy estimated edge 0.46% advance strategy has edge about 0.32%
According to your website, you seem to disagree with the following plays in basic strategy, for 6D, S17, DAS, No Surrender, Peek:

You say Hit A2v5 instead of double.
You say Hit A3v5 instead of double.
You say Hit A4v4 instead of double.
You say Hit A5v4 instead of double.

For these hands, there's no doubt about the mathematically best strategy. The Expected value of these hands can be found here:
http://www.bjmath.com/bjmath/ev/6ds17.htm (Archive copy)

A2v5: Hit=0.137333, Double=0.139852. Double is better.
A3v5: Hit=0.115726, Double=0.138428. Double is better.
A4v4: Hit=0.059395, Double=0.062389. Double is better.
A5v4: Hit=0.039363, Double=0.058627. Double is better.

If you are getting different results, you either are running too few hands to see the correct statistical answer, or you have a bug in your code.
 
#8
Interesting strategy difference

Dear KenSmith
Thanks for your information about the interesting strategy difference
A2v5: Hit=0.137333, Double=0.139852. Double is better.
A3v5: Hit=0.115726, Double=0.138428. Double is better.
A4v4: Hit=0.059395, Double=0.062389. Double is better.
A5v4: Hit=0.039363, Double=0.058627. Double is better.
All this data provided by courtesy of Cacarulo
I believe this is software developer worked for Casino, the probabilities value
Is Casino want people to believe.
You are administrator and I am a retired programmer thinking may different,
I developed my blackjack analyst, set player cards A, 2 dealer face card 5 run 10000 hand bet for $10
Player 1-4 hit and player 5-7 double to bet on $20 for one card.
This test show player 1-4 win more than player 5-7.

10900 10780 12170 12988 10220 6940 6600 dealer: -70598 round: 10001

That is why
I say Hit A2v5 instead of double.
I say Hit A3v5 instead of double.
I say Hit A4v4 instead of double.
I say Hit A5v4 instead of double.
I developed this test in my web site and You are welcome to test by yourself
(Dead link: http://bj21.tearsjoys.org/)
My test can display each hand from seat 1 to seat 7, if any one can found a bug in my program
I will buy a big gift for him or her.:)
 
#9
cjjssc said:
Dear KenSmith
Thanks for your information about the interesting strategy difference
A2v5: Hit=0.137333, Double=0.139852. Double is better.
A3v5: Hit=0.115726, Double=0.138428. Double is better.
A4v4: Hit=0.059395, Double=0.062389. Double is better.
A5v4: Hit=0.039363, Double=0.058627. Double is better.
All this data provided by courtesy of Cacarulo
I believe this is software developer worked for Casino, the probabilities value
Is Casino want people to believe.
Aha! Its a goddamn BJ conspiracy! zg
 

Sonny

Well-Known Member
#10
cjjssc said:
I developed my blackjack analyst, set player cards A, 2 dealer face card 5 run 10000 hand bet for $10
That's the problem. Your sample size isn't big enough. The standard error is much too high. Try running that play for a few billion hands and see what the results say.

To see how high the standard error is, run your 10,000 hand program several times and compare the results. Each run will give very different results. You can take the average of each sim and use that as your mean to find the standard deviation of each individual sim. You will see just how high the variance is for such a small sample.

You can also double-check your results against some of the free programs on our site, like this one:

http://www.blackjackinfo.com/bb/showthread.php?t=9752

It uses exact combinatorial analysis so it is completely accurate.

-Sonny-
 
#11
Sonny said:
That's the problem. Your sample size isn't big enough. The standard error is much too high. Try running that play for a few billion hands and see what the results say.

To see how high the standard error is, run your 10,000 hand program several times and compare the results. Each run will give very different results. You can take the average of each sim and use that as your mean to find the standard deviation of each individual sim. You will see just how high the variance is for such a small sample.

You can also double-check your results against some of the free programs on our site, like this one:

http://www.blackjackinfo.com/bb/showthread.php?t=9752

It uses exact combinatorial analysis so it is completely accurate.

-Sonny-
He's a programmer mathametician - he know about limited sampling - its a conspracy, I tell ya! zg
 
#12
Copy test result screenshot to me, tell me I am wrong.

Hi Sonny,
I went to your web site download the software, but it can not compare different strategy to show
The result, as my web site can display the suffered 1 to 8 decks card sequence and send to
7 seats with 7 users different strategy.
Your screenshot the probability value by each move, it may run from simulator but can not
convinced me how you get the probability value for sample I tried to do same test to
A2v5 show stand -11.88 double 21.23 but no value on hit.
You said “That's the problem. Your sample size isn't big enough”
I wrote Activeperl program on PC command prompt, run sample size to 100000 hands I get the result as
My test say Hit A2v5 instead of double.
My test say Hit A3v5 instead of double.
My test say Hit A4v4 instead of double.
My test say Hit A5v4 instead of double.
If any one have simulator/analyst software, can you run the same test as I did and
Copy your result screenshot to us, tell me I am wrong.
The only way to find the fact and truth is testing and testing.
Do not believe any body sell you software it may be is the casino want you to believe,
My web site is total free directed test on internet just for fun :
(Dead link: http://bj21.tearsjoys.org/)
You can test for fun, It may not fast enough to run 10000 hands display every 1000 hand to
Verify your strategies, about 2 min run time on internet.
 
#13
cjjssc said:
Hi Sonny,
I went to your web site download the software, but it can not compare different strategy to show
The result, as my web site can display the suffered 1 to 8 decks card sequence and send to
7 seats with 7 users different strategy.
Your screenshot the probability value by each move, it may run from simulator but can not
convinced me how you get the probability value for sample I tried to do same test to
A2v5 show stand -11.88 double 21.23 but no value on hit.
You said “That's the problem. Your sample size isn't big enough”
I wrote Activeperl program on PC command prompt, run sample size to 100000 hands I get the result as
My test say Hit A2v5 instead of double.
My test say Hit A3v5 instead of double.
My test say Hit A4v4 instead of double.
My test say Hit A5v4 instead of double.
If any one have simulator/analyst software, can you run the same test as I did and
Copy your result screenshot to us, tell me I am wrong.
The only way to find the fact and truth is testing and testing.
Do not believe any body sell you software it may be is the casino want you to believe,
My web site is total free directed test on internet just for fun :
(Dead link: http://bj21.tearsjoys.org/)
You can test for fun, It may not fast enough to run 10000 hands display every 1000 hand to
Verify your strategies, about 2 min run time on internet.
You need to run 10M hand each and then if the answers are still off, according to some, we shall seek the truth.
If I were to gamble on this, I would bet that you are wrong, but I would hope that you are right. zg
 

KenSmith

Administrator
Staff member
#14
cjjssc,
Perhaps you should post the source code for your shuffle, and let us know how you are obtaining your random numbers.

The correct answer here is not in doubt, and has been verified by many people for many years. Your answers are not the correct answers.
 

QFIT

Well-Known Member
#15
cjjssc said:
Dear KenSmith
Thanks for your information about the interesting strategy difference
A2v5: Hit=0.137333, Double=0.139852. Double is better.
A3v5: Hit=0.115726, Double=0.138428. Double is better.
A4v4: Hit=0.059395, Double=0.062389. Double is better.
A5v4: Hit=0.039363, Double=0.058627. Double is better.
All this data provided by courtesy of Cacarulo
I believe this is software developer worked for Casino, the probabilities value
Is Casino want people to believe.
You are administrator and I am a retired programmer thinking may different,
I developed my blackjack analyst, set player cards A, 2 dealer face card 5 run 10000 hand bet for $10
Player 1-4 hit and player 5-7 double to bet on $20 for one card.
This test show player 1-4 win more than player 5-7.

10900 10780 12170 12988 10220 6940 6600 dealer: -70598 round: 10001

That is why
I say Hit A2v5 instead of double.
I say Hit A3v5 instead of double.
I say Hit A4v4 instead of double.
I say Hit A5v4 instead of double.
I developed this test in my web site and You are welcome to test by yourself
(Dead link: http://bj21.tearsjoys.org/)
My test can display each hand from seat 1 to seat 7, if any one can found a bug in my program
I will buy a big gift for him or her.:)
Cacarulo most certainly does not now nor has ever worked for a casino and would never quote purposely incorrect numbers to harm players.

Your results are incorrect.
 
#16
Actually, I would guess that his results are too consistent to be chalked off to limited sampling.
Herr Professor, there's a bug in your algorithm. zg
 
#17
Post the source code for my shuffle

from KenSmith to cjjssc,
Perhaps you should post the source code for your shuffle, and let us know how you are obtaining your random numbers.

Hi KenSmith,

Yes I like to post my source code for my shuffle, let me explain a little.
Random number sample :
$I = int ( rand($cards - $a) );
The random number call rand is a decimal number between 0 to 1, int is pickup the integer function.
$cards is the number from 1 decks to 8 decks ( from 52 cards to 416 cards)
It is picked up position $I, relocate to new table position start with 1 (or 0 as the Perl CGI index)
The old table will be picked up one card, a sample on deck 52 card position 10 is card 3,
Put it on new table position 1 (or index 0), and the old table will shift the empty position and old
Table will be only 51 card left new table has one card only.
Looping 52 times the new table will be all shuffled, as the dealer does it.
If you are a programmer know Perl CGI, you will understand my source code :
It will be much easy click my web site and test it by yourself
(Dead link: http://bj21.tearsjoys.org/)


$deck = "A A A A 2 2 2 2 3 3 3 3 4 4 4 4 5 5 5 5 6 6 6 6 7 7 7 7 8 8 8 8 9 9 9 9 10 10 10 J J J J Q Q Q Q K K K K";
$deck2 = $deck." ".$deck;
$deck8 = $deck2." ".$deck2." ".$deck2." ".$deck2;
if ( !$rq{deck} )
{ $rq{deck} = 2;}
$cards = $rq{deck} * 52;
$leng = $rq{deck} * 108;
$leng = $leng - 1;
$sufat = $rq{deck} * 13;
$deck6 = substr($deck8, 0, $leng);
@c = split (/ /, $deck6);
# This &suffer sub-routine shuffle cards
&suffer; $cnt = 0; $cur_h_l = 0;
# deliver first 2 cards to each
$a = 0; $b = 1;
@player = ""; $rnd=1;
if ( $rq{round} > 10001) { $rq{disccnt} = 1000; }
elsif ( $rq{round} > 1001) { $rq{disccnt} = 100; }
if ( $rq{bet} ) {$bet = $rq{bet};}
else { $bet = 10; }
if ( $rq{round} ) { $totalh = $rq{round}; }
else { $totalh = 5; }
while ($rnd <= $totalh)
{ &round;
if ( $rq{discnt} > 0 )
{ $last = $rnd % $rq{discnt}; }
else
{ $last = 0; }
$rnd++;
}
print "<tr><td>$ftotal[0] <td>$ftotal[1] <td>$ftotal[2] <td>$ftotal[3] <td>$ftotal[4] <td>$ftotal[5]<td>$ftotal[6]<td> dealer: $ftotal[7] <td>
round: $rnd \n";

#close OUT;
print "</tabel>";
sub suffer
{
@dk6 = split (/ /, $deck6);
$a = 0; @s="";
$cards = $rq{deck} * 52;
while ($a < $cards)
{
$I = int ( rand($cards - $a) );
$s[$a] = $dk6[$I];
$dk6[$I] = " ";
while ($I < $cards && $dk6[$I] eq " ")
{ $dk6[$I] = $dk6[$I+1];
$dk6[$I+1] = " ";
$I++;
}
$a++;
}
@card = ""; $i=0;
if ( $last == 0 && $rnd == 0 )
{
$x=0; $y=0; $z = 8; @card="";
while ($y < $z && $y < $cards)
{
while ($x < $z) {$card[$y] .= $s[$x++];}
$y++; $z += 8;
}
print "<tr><td>shuffle<td>$rq{deck} decks:";
while ( $card[$i] )
{ print "<td> $card[$i++]"; }
print "<tr><td>seat 1 score<td>2 score<td>3 score<td>4 score
<td>5 score<td>6 score<td>7 score <td>Dealer score<td>up cnt h-l true";
print "<tr><td>total=cards<td>t=cards<td>3 t=cards<td>4 t=cards
<td>5 t=cards<td>6 t=cards<td>7 t=cards<td>Dealer t=cards Cards<td>card index cnt";

}

}
 

QFIT

Well-Known Member
#18
cjjssc said:
from KenSmith to cjjssc,
Perhaps you should post the source code for your shuffle, and let us know how you are obtaining your random numbers.

Hi KenSmith,

Yes I like to post my source code for my shuffle, let me explain a little.
Random number sample :...
1. When I run a new strategy, I run hundreds of billions of hands -- not 10,000.
2. Your RNG is useless.
3. Without looking at the content of your program, it is obvious that it is flawed since it is a tiny, tiny fraction of the code needed.
4. Cacarulo does not use sampling. He looks at every single possible deck composition.
5. You owe an apology to Cacarulo.

Sorry to be so severe. But you posted that people have purposely attempted to harm players because of errors that you have made. BJ has been investigated by many, many people for decades. Consider the possibility that so many mathematicians could be correct.
 
#20
Casino not welcome list

Hi QFIT
I agree your advice as
When I run a new strategy, I run hundreds of billions of hands -- not 10,000.
I do not have high speed computer.

I like to post my other experience: Casino not welcome list
:confused:
I was winning on blackjack table, casino manger from my back
hand in her business card and said “You can not play blackjack”,
I asked her “Why?” She said “I do not know why, But my manager told me to do it,
You are welcome to play any other game”.
I paid my room on that casino, next day I tried to play again; I thought nobody remember me,
But after playing for half hour, another manager said “are you name xxxx ? did we tell you
not to play blackjack?”.
I was asked “not to play Blackjack” 4-5 times from that casino, I complained to better business bureaus
But they said “Casino have right to refuse you without any reason”.
The reason why by myself:
1. I have the advanced strategy as I developed blackjack simulator/Analyst software to modify basic strategy post
On your web.
2. I am a good card counter as I developed Hi-Lo eye scan practice up to 6-8 decks, If any one interesting,
I will post with a sample explain in detail.
3. I was a VIP member on that Casino; I can mark thousand dollars on table, every time when I show my card,
Casino easy to trace how I play, I could not mark or even play anymore:confused:
 
Top