Baccarat Evolution gaming live online

I have over 25,000 shoes sent to me from a programmer dealt from a virtual deck. The programmer did everything right, 8 decks, shuffle, burn cards, card to end the shoe but when I test ideas that win consistently on my 4062 live shoes, they tend to fail badly on the virtual data.

An example is the 70% P bet B system (starting at hands 14 through 32). It wins easily on my live shoes but fails pretty much immediately on the virtual shoes. The Banker/Player ratio is fine but something seems off. Maybe the extremes are different? I will post an example shoe below of the virtual data (includes ties).

shoe,hand,play,natural,pCards,bCards,pCardVals,bCardVals,pHandVal,bHandVal
1,1,"B",1,"J6","8J",06,80,6,8
1,2,"P",0,"6T","K22",60,022,6,4
1,3,"B",0,"7K","JK9",70,009,7,9
1,4,"B",0,"7Q","TT8",70,008,7,8
1,5,"B",0,"687","A57",687,157,1,3
1,6,"P",0,"A49","659",149,659,4,0
1,7,"P",1,"53","83",53,83,8,1
1,8,"P",1,"A8","32",18,32,9,5
1,9,"T",0,"Q3J","A2Q",030,120,3,3
1,10,"B",0,"29A","6T",291,60,2,6
1,11,"P",0,"7T","762",70,762,7,5
1,12,"P",1,"K9","4K",09,40,9,4
1,13,"B",1,"52","Q8",52,08,7,8
1,14,"B",1,"9A","J9",91,09,0,9
1,15,"B",0,"6K","TT9",60,009,6,9
1,16,"B",1,"KK","45",00,45,0,9
1,17,"B",1,"84","8Q",84,80,2,8
1,18,"P",1,"54","K5",54,05,9,5
1,19,"P",1,"K8","4T",08,40,8,4
1,20,"P",0,"58K","769",580,769,3,2
1,21,"P",0,"KJ7","6T4",007,604,7,0
1,22,"B",1,"87","8J",87,80,5,8
1,23,"P",0,"Q52","A3A",052,131,7,5
1,24,"B",1,"A4","63",14,63,5,9
1,25,"P",1,"26","KQ",26,00,8,0
1,26,"T",0,"42","88",42,88,6,6
1,27,"B",1,"43","53",43,53,7,8
1,28,"B",1,"A9","K9",19,09,0,9
1,29,"P",1,"T8","A2",08,12,8,3
1,30,"P",1,"8K","87",80,87,8,5
1,31,"P",1,"99","2Q",99,20,8,2
1,32,"B",0,"2JK","96",200,96,2,5
1,33,"B",0,"T2Q","J7",020,07,2,7
1,34,"B",1,"TA","35",01,35,1,8
1,35,"T",0,"K7","QJ7",07,007,7,7
1,36,"B",1,"A9","K8",19,08,0,8
1,37,"B",1,"78","72",78,72,5,9
1,38,"B",0,"563","J5",563,05,4,5
1,39,"P",1,"62","43",62,43,8,7
1,40,"B",1,"76","63",76,63,3,9
1,41,"P",1,"J9","AT",09,10,9,1
1,42,"P",0,"Q4K","QAQ",040,010,4,1
1,43,"P",0,"4Q3","QJ5",403,005,7,5
1,44,"P",1,"27","QJ",27,00,9,0
1,45,"P",1,"J8","AQ",08,10,8,1
1,46,"B",0,"KK2","K5",002,05,2,5
1,47,"B",0,"399","945",399,945,1,8
1,48,"T",0,"T6","T6",06,06,6,6
1,49,"P",0,"24","85T",24,850,6,3
1,50,"P",0,"JA4","498",014,498,5,1
1,51,"P",0,"322","6J",322,60,7,6
1,52,"P",0,"K54","J6",054,06,9,6
1,53,"P",1,"A8","57",18,57,9,2
1,54,"P",1,"8J","37",80,37,8,0
1,55,"P",1,"9J","AJ",90,10,9,1
1,56,"P",1,"54","3T",54,30,9,3
1,57,"P",0,"K25","767",025,767,7,0
1,58,"T",1,"9T","36",90,36,9,9
1,59,"P",0,"KT5","TTT",005,000,5,0
1,60,"P",1,"9Q","65",90,65,9,1
1,61,"P",1,"A8","J5",18,05,9,5
1,62,"B",1,"Q3","53",03,53,3,8
1,63,"B",0,"28K","397",280,397,0,9
1,64,"T",0,"KQA","65J",001,650,1,1
1,65,"B",0,"2J9","644",209,644,1,4
1,66,"B",0,"T49","52",049,52,3,7
1,67,"B",0,"2AJ","89",210,89,3,7
1,68,"P",1,"A7","AQ",17,10,8,1
1,69,"P",1,"J9","33",09,33,9,6
1,70,"P",1,"45","AQ",45,10,9,1
1,71,"P",0,"964","387",964,387,9,8
1,72,"T",0,"77Q","4Q",770,40,4,4
1,73,"B",0,"A37","QT6",137,006,1,6
1,74,"P",0,"2QT","47J",200,470,2,1
1,75,"P",0,"T6","2T8",06,208,6,0
1,76,"P",0,"T7","JJ4",07,004,7,4
1,77,"B",0,"73A","654",731,654,1,5
1,78,"P",1,"A7","82",17,82,8,0
1,79,"B",1,"QT","26",00,26,0,8
1,80,"P",1,"K8","57",08,57,8,2
1,81,"P",0,"QT9","3QT",009,300,9,3
1,82,"T",0,"842","3AT",842,310,4,4
1,83,"P",0,"33","QKJ",33,000,6,0

 

Yes, I guess that would be true in a regulated environment. I think you have that in NJ now. There may be a subtle difference between the Live and Simulated shoes that the software provider may not be aware of, or that would make no difference in an Actuary's mind. In any case, I prefer Live shoes.

 

x

 

It sees then all you got to do is reverse the BS and win on the RNG, no? Or is it just variance? Looking at the SD should give a clue.

What is that selection? Wait for 3P bet B? Once again, just reverse and win, no?

Personal references aside, the idea of "eyeballing" shoes to draw a conclusion akin to drinking bleach to rid Covid. Surely there must be a way to compare other than P B outcomes. Run length ratiios? This is above my pay grade but it shouldn't be for you guys who can program.

A friend sent me a BS whereby he won well for 100 shoes. I picked 100 shoes from my data and lost miserably using the same selection. Personally I haven't played that many RNG shoes but haven't had an issue with the ones I've played EVEN THOUGH my FEELING was they were streakier.

Then we have the possibility of different casinos shuffle producing different data streams. In the end it is all moot when playing a guessing game with MM, doesn't it?

 

This is a system I have talked about before (70% P bet B), not sure what thread. It was sent to me years ago with limited testing (less than 200 shoes if i remember). Surprisingly it did generate a flat bet profit under testing and has held up as I have accumulated more live shoes. I have modified the parameters slightly.

Here is the theory. We know that Banker wins 50.7% of decisions and Player wins 49.3% of decisions. For this reason, most shoes are Banker heavy or at least Banker decisions outnumber Player. If we find a shoe where Player decisions outnumber Banker decisions by a significant margin relatively early in the shoe, we flat bet Banker for the remainder of the shoe with the idea that Banker will eventually even things up somewhat, bringing the decisions closer to 50:50 or even pushing Banker back into the lead.

The exact instructions are as follows. Beginning at hand 14 (not including ties) and ending at hand 32, if we see that Player makes up 70% or more of the non-tie decisions, we start flat betting Banker for the rest of the shoe. This is +EV flat betting on my data. You can also use a -7 unit stop loss (including commission paid) to improve results and decrease drawdown. The downside of this method is that only 10% of shoes qualify. It would only be practical if you have a B&M or online casino with a large number of tables playing at once so you can pick and choose your shoe as opportunities arise.

This is something that I look at each session. If Player crosses over 70% level and is in the middle of a Player streak I wait until the Player streak is over before favoring the Banker side. I have created a reference sheet to help me which I will attach below.

 

Out of curiosity, why start at hand 14 and not wait later in the shoe? I've read something similar where they waited until about a third of the shoe was complete and used the same strategy.

 

What a fallacy. RO(FLMAO.

 

Waiting for an imbalance and betting regression to the mean is no fallacy.

Sidebar: The downfall of JAE’S Oscar Grind is that he wasn't getting a true imbalance due to chops putting him in correction territory.

 

You could wait until one-third of the shoe and still show profits.

When I run it under simulation on my 4062 live shoes, starting at hand 14 and ending at hand 32 achieves maximum profit. You could start as early as hand 12 and still do great. Starting at hand 11 or earlier decreases profits significantly. The damage is done more by starting too early. In an eight deck game you have about 72 banker/player decisions so one-third would be hand 24. I ran this for you and if Player is at 70% or higher at hand 24 this is profitable as well, although you do cut total profits for the test by more than 50%.

 

I write this here for everybody to becareful when you decide to play any game on this provider especially live roulette tables.
What happened to you i can assure you it will happen again and nobody will give your money back.
I will give a very simple example you can try if you have time.
Go and watch speed auto roulette(black).

Sometimes numbers come out and they do not pay you if bet also they do not put it in the history of the game so it never happened for them.
Even if you have a video that proves otherwise what do you think will happen?
Nothing because all these gambling commissions out there are taking money from these companies also governments making tones of money
from taxes so nobody cares.

Also their licence is under Malta s gambling commission which means even if you want to appeal you need a lot of money to spend.
Who will do that?
Answer almost nobody and they know that.

If somebody disagree feel free to go on youtube and check all these fake spins with magnets.
Personally i would never play on their tables.

 

95% of the online casinos in my state use Evolution and I play Baccarat or roulette every day and have never had a problem. Evolution is one of the oldest and most respected live platforms in the world.

 

International on line casinos . Play at your own risk.

 

Bbbwwaaaahhh, so now the evenbob-spike playin baccarats too??!! I gotta say that you are just like yer limey, arch-nemesis/adversary! Just two pea in the pods, a pair of attentions-seeking bulls-shits artists, fulls-stop, hey hey!!!!

 

The site that I found something similar had: play at least 10% of the shoe. Once EITHER player or bank reach 58% play the opposite until it goes down to 50%.

 

I have tried to find some type of imbalance on the Banker side that works but have been unable to. 58% is not enough in my opinion. I ran it again just now to refresh my memory and it lost badly. Not recommended.

 

Have you ever looked into the amount of singles as compared to verticals in a shoe?

 

I don't know if you remember the '24 Karat Baccarat' system but it dealt with this comparison. I am not sure if I every actually tried to code and test extreme columns vs singles, maybe I should when I have some time.

There was another system called 'Probability Baccarat' that dealt with these to some extent, single chops / multi-chops and single columns, multi-columns, I think.

 

@fathead Hi fathead. I just want to revisit this distinction between RNG and live baccarat shoes.

I think you have said that the ratio of P and B is consistent with the expected value. If the ratio of run lengths are also consistent with the expected value then there should be no difference between live and RNG. This should be pretty easily discerned. I'm surprised no one has done this.

@Jake0628 right as long as they're using complete decks it should play out like a regular game as long as the cards are "randomized." If not it should show up in expected values being skewered.

 

Random number generators are really 'pseudorandom'. They never really achieve the same randomness of a roulette wheel, for example.

Many years ago I wrote some simple code using a random number generator to simulate Baccarat hands. The Banker/Player ratio looked good and at the time I thought it would suffice for testing. I later found out it was useless. Out of 2500 shoes, do you know how many shoes had an imbalance of 70% Players as compared to Bankers sometime early to mid shoe? Zero. Obviously the extremes that you see in the live shoes were not present.

If you do complex coding with 8 decks of 52 cards you can get closer but you still need some sort of RNG to shuffle the virtual cards. It is not quite the same, in my opinion, although it may be good enough for most methods that bettors use.

 

So true!