System Test Thread: Summary with Statistics on Various Systems

[ayuinca]

Directly looking pattern from streak or chop of Banker and Player, you could not find any bias pattern. Longer the pattern just gives you lower occurence to bet because the probability for longer pattern is lower. The next hand banker or player coming still could not escape the probabilies 50.68% for Banker and 49.32% for Player if there are no tie. In the way, the house still remains. If Bryan can win with this stuff for many years, I bet he not merely play with it.

Once there was an expensive method named 0C40 but later it can be obtained freely from the inventor. Why? I think you know why.

[hlkhoo]

Ayuinca, how do you actually generate those RNG shoes? Is it via some kinda website or software? Do you mind sharing that with me. I want to generate some data and test out some method on my own.
Btw, thanks for your time and effort in helping us testing some system. Really appreciate it. Thanks man.

[Bryan]

ayuinca,

Thanks for taking the time to test. You did it correctly.

[scot]

Ayuinca,
In laymans term please. You are saying one thing and Bryan is saying another. Does the trigger work or not? I PM'd you Bryan and haven't received the trigger yet. Thank you.

[ayuinca]

Ayuinca, how do you actually generate those RNG shoes? Is it via some kinda website or software? Do you mind sharing that with me. I want to generate some data and test out some method on my own.
Btw, thanks for your time and effort in helping us testing some system. Really appreciate it. Thanks man.

Thanks hlkhoo. I have my own program to generate shoes. I can help you to generate RNG shoes for you. You can just tell me how many you want, what output format? e.g. detailed play like P234 B012; or just result shoe BPBPPP...; or need card shoe; regarding rule, burn card or not, 6 or 8 deck, etc.

[hlkhoo]

Hi Ayuinca. can can help me to generate 100 shoes based on the how you generate for Bryan? Thank you.

[ayuinca]

Ayuinca,
In laymans term please. You are saying one thing and Bryan is saying another. Does the trigger work or not? I PM'd you Bryan and haven't received the trigger yet. Thank you.

Sorry, I don't receive PM from you. The result is there you can judge by yourself. Whether put up the rules you have to ask Bryan since I will not copy here without his grant.

[ayuinca]

Hi Ayuinca. can can help me to generate 100 shoes based on the how you generate for Bryan? Thank you.

Find attachment 100 RNG shoes.

[scot]

Ayuinca,
In laymans term please. You are saying one thing and Bryan is saying another. Does the trigger work or not? I PM'd you Bryan and haven't received the trigger yet. Thank you.

ayuinca,

Thanks for taking the time to test. You did it correctly.

Sorry, I don't receive PM from you. The result is there you can judge by yourself. Whether put up the rules you have to ask Bryan since I will not copy here without his grant.

Oh that's OK, you can PM me the trigger. I am pretty sure Bryan is OK with that as he stated if done in Private and I think Bryan has gone on his gambling trip. I just wanted the Trigger via PM in the first place. thanks.

[hlkhoo]

Find attachment 100 RNG shoes.

Thanks man!

[Bryan]

Banker Trigger

Since Bryan has not given me his personal email, I think he does't mind I put sample shoes and the test statistics here because I need him to verify if the play is correct according to his rules; and without the rules I think no one can figure out how is a trigger qualified, and the tirgger in fact still could not work better than the banker probability, i.e. in a long term it has no advantage over the edge.

Here are 5 sets of 10k RNG shoes test:

1) Total Bets: 33881 (3.39/shoe)
Total Shoes Net Win: -534.05 (-0.05/shoe)
Turnover(Net Win/Total Bet): -1.58%

ayuinca,

Actually I was thinking of your test result last night and so, I went through the details card number being played in your sample and I found the differences.

1. You don't burn 1 card before you start playing each game. It is different from what we are playing here in Asia. Before the start of each game, we will burn 1 card.

At the beginning of the shoe, you burn 10 cards which should be (3887621034), how does that come in? Over here, before the start of shoe, the dealer will flip the first card and if it is a 3, only meaning a total of 4 card (plus the 1 card which is 3) will be burn.

Anyway, even when those 10 cards are burned and when play with no burn card before each game, I was not able to get the card point you get in game 10 and from there, I stop. I never goes beyond then.

Maybe you can take a look at the sample shoes again.

Thanks.

[Bryan]

Oh that's OK, you can PM me the trigger. I am pretty sure Bryan is OK with that as he stated if done in Private and I think Bryan has gone on his gambling trip. I just wanted the Trigger via PM in the first place. thanks.

Scot, I have PM you my trigger.

Today is Sun in my country, so, I yet to go for my trip. Tomorrow is the day.

[hlkhoo]

Thanks Bryan. I have received your triggers. When I compared your betting triggers vs Ayuinca sample files, I can't find the mini progression (1-2) which you mentioned in your earlier posts. When you play live these days, do you flat bet only?

[Bryan]

Thanks Bryan. I have received your triggers. When I compared your betting triggers vs Ayuinca sample files, I can't find the mini progression (1-2) which you mentioned in your earlier posts. When you play live these days, do you flat bet only?

hlkhoo,

Can see that you are very observant and detailed. You are the first to ask about my progression which I casually mentioned.

Most of the times, I will go flat bet. But, once I lost 3 units, I will start my 1-2 progression. If total 10 units is lost, I will call it a day.

[ayuinca]

ayuinca,

Actually I was thinking of your test result last night and so, I went through the details card number being played in your sample and I found the differences.

1. You don't burn 1 card before you start playing each game. It is different from what we are playing here in Asia. Before the start of each game, we will burn 1 card.

At the beginning of the shoe, you burn 10 cards which should be (3887621034), how does that come in? Over here, before the start of shoe, the dealer will flip the first card and if it is a 3, only meaning a total of 4 card (plus the 1 card which is 3) will be burn.

Anyway, even when those 10 cards are burned and when play with no burn card before each game, I was not able to get the card point you get in game 10 and from there, I stop. I never goes beyond then.

Maybe you can take a look at the sample shoes again.

Thanks.

First, the burn-card rule. In most of place, the game starts with open one card, then burn 1 to 10 cards according the open card value; so the number of card burnt min. 2 and max. 11. The program just randomly pick a number from 2 to 11 for how many total cards will burn. In this way, that burning 10 cards means to flip one card which is 9 then burn 9 cards; actually 10 cards are burnt. It's analogous to burn 10 cards. I list the whole card-shoe but in fact the first hand start from after the card burnt. I omit all the play details regarding cards dealt. So, in that shoe, the first hand starts from 1872305...

Yes, I realise this will be a bug since I should burn cards according to the first card value instead of burn them randomly. Good point. I will fix it. But now I expect that any change of the rule will change the a single whole shoe result, however for testing purpose, the result after thousands of shoes will not change.

Secondly, about the rule in Asian burning one card before each hand. I have no idea why they do that. I believe that is not necessary; and the total baccarat statistics properties will not change under this rule. If you want to test with such rule I am ready for it.

Please find 'sample1.txt' with revised burn rule. I run other two sets of 10k shoes after the burn rule revised. The results are pretty same.

1.
Total Bets: 33353 (3.34/shoe)
Total Shoes Net Win: -622.25 (-0.06/shoe)
Turnover(Net Win/Total Bet): -1.87%

2.
Total Bets: 33399 (3.34/shoe)
Total Shoes Net Win: -523.95 (-0.05/shoe)
Turnover(Net Win/Total Bet): -1.57%

Note that only revise the first card value burn rule, not very hand burn card rule.

[Ellis]

I agree wholeheartedly. I made the same mistake quite a few years back. The system won 1000 shoes RG then lost the first 3 in the casino. Go figure,

[ayuinca]

I don't know what that means by 'lack of long streak'? RNG never lacks long streaks.

As I have said there are total 416!/[(32!)^9*128!]=1.6494E+376 different Card Shoe in the world, but only 4.7224E+21 Result Shoe. So, there are about 3.5E+354 different(attention to the word 'different') Card Shoe will have one identical Result Shoe. I always believe any RND Card Shoe is one of those 1.6494E+376, why is it not having the same effect as hand shuffle or machine shuffle?

Yes, 3.5E+354 different combinations of Card Shoe, you can't image how big is this figure. So many shoes can have complete one same 72 result hands, what's the point that a RNG not valid to play or test?

Give me sound reason, not your feeling.

[ayuinca]

I agree wholeheartedly. I made the same mistake quite a few years back. The system won 1000 shoes RG then lost the first 3 in the casino. Go figure,

"The system won 1000 shoes RG " what a BS! Even you won 1 millions shoes, it can't tell anything just one event of 3-lost-in-a-row. You are old, man... your words are incoherent.

[ayuinca]

Here is 1541 real shoes test, and zumma 600, 1000. Thank Teo2 for sending me the zumma. Hope he not mind I test with other system, and the gentlemant sent me those 1541 shoes.

1). zumma 600
Total Bets: 1976 (3.29/shoe)
Total Shoes Net Win: 9.10 (0.02/shoe)
Turnover(Net Win/Total Bet): 0.46%

2). zumma 1000
Total Bets: 3463 (3.46/shoe)
Total Shoes Net Win: -44.65 (-0.04/shoe)
Turnover(Net Win/Total Bet): -1.29%

3) 1541 real shoes
Total Bets: 5178 (3.36/shoe)
Total Shoes Net Win: -141.15 (-0.09/shoe)
Turnover(Net Win/Total Bet): -2.73%

[pitty1]

Hi Ayuinca...Hey i'm really having a problem with what to bet on if a PPBB pattern comes up..Would it be better to bet on BBPP or a PPBB pattern?, If u have the time.. can u run this pattern thru the zummas to see if one pattern gives u a better win than the other?(when a PPBB comes up)..thanks much..

[ayuinca]

Hi Ayuinca...Hey i'm really having a problem with what to bet on if a PPBB pattern comes up..Would it be better to bet on BBPP or a PPBB pattern?, If u have the time.. can u run this pattern thru the zummas to see if one pattern gives u a better win than the other?(when a PPBB comes up)..thanks much..

Could you let me know exactly what are you going to do? Do you mean once the last 4 hands are PPBB, you are going to bet BBPP?　Not quite remember your method.

[pitty1]

Hi ayuinca..per shoe..from beginning to end of shoe ..hands are divided into a pattern of 4..i.e....BPBB..BPPB...PBPP.When a PPBB comes up..if betting for a BBPP vs PPBB...which of the 2 patterns have a better chance to hit? a hit will stop betting but let pattern finish at 4 hands per... example a PPBB comes up try a PPBB play bet for a win..and out come would be PBPB..win on 1st bet the P but no bet on rest of pattern..counts as a win ..I would think 2 runs thru the zumma one trying for a win on BBPP play bet everytime a PPBB comes up..and another run with a PPBB play bets when the PPBB pattern comes up..compare the win/loss of the two..thanks much

[sawajiri]

ayuinca,

thanks a lot for testing these various systems against RNG and real shoes,
and posting the results.
i am currently testing a system on my own.
i will let you know if any positive results come up.
i'll send u a PM.

cheers,
sawajiri

[ayuinca]

pitty1, i will come back for that when i have time. should be easy as i need to change a littele of the program.

john, if you have read the relathionship betweeen Card Shoe and Result Shoe you will know. i am not using Excel for testing. i am using c++ programming; i am not generate the result shoes directly but generate the card shoes then accoding to the baccarat rules play the cards to get the result shoes. maybe you can do it with Excel but i don't get it why your Excel don't get streaks.

[ayuinca]

Test just can tell the whole look of a system, surely could not make sure about short term deviation. The term 'short term' is also relative, when you find a short term bias from some shoes played, you just can't be sure how long the bias will last, specifically you don't know since when the bias has faded out. Let say the past 50 shoes give you bias on 1S more than total of >1S 1%, so you guess the bias will continue for the next few shoes but how many? another 50? Ok, you bet 1S for the next 50 shoes and win then. If unlucky you can just be even or lose.

Statistics can't tell short term deviation but neither be good at short term prediction. Short term is relative, if a bias exists I hope such short term could be my life time long but what is the confidence of this will happen? Best scenario maybe 50% winner and 50% loser but am afraid more 50% loser.

In this term I understand what you mean John. Every system can win in a short term and also can loss in another short term so it just makes the player frustrated. If you know the system can work statistically, I believe the player feel better and be more confident even he loses in a short term. A statistical losing system dooms me loser in a long term if I play mechanically according to the system; a statistical winning system may not make me winner but would not doom me loser. Any way I don't care to play completely mechanical (it's really boring to do that) if the system is a winning one; even there is people can win with a losing system with personal style, can everyone?so add in my personal style I can win more, can't I? This is a subject more than just testing.

Personally I won't play a losing system even it can win in a short term, I'd rather play a winning system even it can lose in a short term. So, it's your choice.

[Bryan]

My previous argument about the testing is not good is because I thought it is a plain P/B kind of generation, but, his program testing is not and it is similar to what is being played live with all the cards being lay out.

Having said that, statistics is one thing, playing live is another. If there is a system that can go thru his program a million times and win every shoes, I would said that is the holy grail. Reason because we will never get to play that many shoes in real life.

And I agreed with ayuinca with his view on long term and short term of looking at statistic. Maybe I am lucky, I am still winning after so many years.

[Moonraker]

I should better open a new thread here to post summary with statistics on various systems I read at this forum. I appreciate that people sharing their ideas and works so I am willing to contribute something as return.

Here is my suggestion. If you have any system would like to test please post your system in a new thread and give me a note under this one so that I can receive the notice email since I do not always come to the forum.

I will post some results here about some recent systems time by time. I believe there are ones can overcome the house, yes, mechanical ones. But after overcome the house, how much net win per shoe or turnover is unknown. It's up to your system.

I wish the post can be sticky, and we beat the baccarat in a more mathematical and statistical manner. Thanks.

Hi ayuinca,

With your stats on Andy's Predication system, results of 4,5 or 6 cards in the preceding hand what I think might be an interesting statistic would be what follows a winning hand with 2 cards v a losing hand with 3 cards. for example,

Bank has a 4 and a 3 for 7 beating player who has a 2, ace and 10 totaling 3 hence bank wins 7 to 3 but with only 2 cards against the players 3 cards.

You never know it maybe an interesting stat.

Regards Moonraker

[ayuinca]

Hi ayuinca,

With your stats on Andy's Predication system, results of 4,5 or 6 cards in the preceding hand what I think might be an interesting statistic would be what follows a winning hand with 2 cards v a losing hand with 3 cards. for example,

Bank has a 4 and a 3 for 7 beating player who has a 2, ace and 10 totaling 3 hence bank wins 7 to 3 but with only 2 cards against the players 3 cards.

You never know it maybe an interesting stat.

Regards Moonraker

Andy's one is interesting. It is a good example to demonstrate no matter what bet selection of the system the probabilities for B and P remain same.

You can see that when last hand is 4 or 6 card dealt it bets FOLLOW LAST HAND(FLH); when last hand is 5 card dealt, it bets OPPOSITE LAST HAND(OLH). So, the amount you bet FLH is about as twice as that you bet OLH; however the amounts of + and - are about the same. It deduces that no matter what you bet or what bet selection you play, as long as it is mechanical, the ultimate results win and lose are close to 50-50.

It implies that the house is still there, but I can give you a hint that could decrease half the commission and even more. One is to compare betting nS of Player not go to (n+1)S, each time you win (because you bet Banker) you have to be commissioned. This is directly bet on pattern of nS of Player and Banker. If you bet on pattern of nS of + and - you could save half commission because you do not always bet Banker. The other way to minimize commission is to bet Player only, any time the system calls a Banker just ignore it, only bet when it calls for Player. Think about this, if the system remains win/lose=1 why not make it bet player only? It save you the commission which is a considerable amount in a long term.

[ayuinca]

Ellis' NU System

rules:

1). Starting at 1 under the first circle on each side bet U1D2 M2 straight down both sides on your
card. (Up 1 when you lose and Down 2 when you win except don't go below 1). M2 = Mandatory 2.
Go to 2 regardless of win or loss after that side gets a circle.
2). Table bet the difference on the side with the higher entry.
3). Under a 1, stop after 2 losses, unless it follows 2 on the opposite side, then go to 3. But, if the
3rd bet is a 5, stop at 2.
4). Under a 2, stop after 3 losses.
5). Under a 3, stop after 1 loss.
6). Wait for a side to win then resume where you left off.
7). When a side wins, that side keeps betting.
8). When you lose a 5 entry, that side stops betting until it has a 2 in a row.
9). Always resume at 1 when losing a 2 under a run, unless the other side is suspended (which
means there is not a bet to be made on that side), then it resumes at 3.
10). Get out at +20
11). When you lose 2 bets under a one and then lose one bet under a consecutive 1 (sporadic 1's) it's a good idea to stop. You might be in the dreaded 212s.

10K shoes test result:
Total Net Win: -16560.20
Shoe Range Occurrence(% of total shoes)
------------------------------------------
90~ 95 - ( -%)
85~ 90 - ( -%)
80~ 85 - ( -%)
75~ 80 - ( -%)
70~ 75 - ( -%)
65~ 70 - ( -%)
60~ 65 - ( -%)
55~ 60 - ( -%)
50~ 55 1 ( 0.01%)
45~ 50 14 ( 0.14%)
40~ 45 28 ( 0.28%)
35~ 40 79 ( 0.79%)
30~ 35 145 ( 1.45%)
25~ 30 309 ( 3.09%)
20~ 25 472 ( 4.72%)
15~ 20 690 ( 6.90%)
10~ 15 907 ( 9.07%)
5~ 10 1113 (11.13%)
0~ 5 1106 (11.06%)
-5~ 0 1104 (11.04%)
-10~ -5 1008 (10.08%)
-15~ -10 807 ( 8.07%)
-20~ -15 647 ( 6.47%)
-25~ -20 518 ( 5.18%)
-30~ -25 371 ( 3.71%)
-35~ -30 266 ( 2.66%)
-40~ -35 178 ( 1.78%)
-45~ -40 104 ( 1.04%)
-50~ -45 67 ( 0.67%)
-55~ -50 32 ( 0.32%)
-60~ -55 15 ( 0.15%)
-65~ -60 10 ( 0.10%)
-70~ -65 8 ( 0.08%)
-75~ -70 - ( -%)
-80~ -75 1 ( 0.01%)
-85~ -80 - ( -%)
-90~ -85 - ( -%)
-95~ -90 - ( -%)
-100~ -95 - ( -%)
Positive Shoes: 4864 (48.64%)
------------------------------------------

[Steve]

Not sure if tests have covered this aspect...

It would be interesting to see Bryan's strategy tested on Zumma so that trigger and bet(s) are placed intra-shoe. In other words the trigger will never be the basis for bet placement in the following shoe.

This would ensure accurate bet placement as in live situation, and pick up on any unknown factors.


(Is there a text file version... of Zumma 1000 shoe book.)