Baccarat Cutting House Edge to -0.56%

Most of you know I have a large live shoe database that I run computer simulations on. I have been playing around with some ideas this week and here is one that cuts the house edge significantly on my data at least. It is pretty simple.

If you see a streak of two, bet with the streak until four (or end of streak), then (if the streak continues) bet against the streak for a max of three bets (or until a win). If the streak continues beyond seven then you would sit out.

X
X
bet
X
X
up to four (or a loss) then bet
Y
Y
Y
up to three max (or a win)

This is a long run test (a total of 91,600 bets were made), so big swings but its gets the edge almost down to playing blackjack with perfect basic strategy. I guess this makes sense since long streaks (while fun) are not very common.

 

Why are you running simulations on accumulated data?

Unless you have an infinite database the only sensible way to run simulations is using probabilities.

Any amount of finite data will only give you best (not so good) guesses biased by the small sample set. And when it comes to multi-deck shoes all samples are SMALL.

You would need BILLIONS to even approach statistical reality.

 

It's true, what you say. I have over 25,000 simulated shoes with a virtual 8 deck dealer but I don't trust it. I get odd results compared to table data so I test with what table data that I have (getting close to 3000 shoes). I wish I had a million live shoes.

 

Well at least you know there is a problem. It's more than we can say about most of civilization. Be well...

You might want to look at poisson distribution & Bayes's Theorem as ways of confidence testing your simulations.

It's like a second level of knowing how sure you should be about your results.

May the force be with you!

 

Thank you. The way I look at it, it won't be worse than random bets, and more likely better, so might as well go with it. Take Care.

 

Money management, managing losses, and bankroll control (like many have said here) are more important anyway.

 

I don't understand how this would reduce the HE.... Because series of two > series of four....

 

Like Frank said earlier it may not over billions of bets, could be a fluke. But in my testing in over 91,000 bets this strategy wins at higher than the normal rate, bringing the house edge down to the figure mentioned.

It is really a four part strategy. (1) Betting with Banker 2 streak up to 4 or loss, (2) Betting with Player 2 streak up to 4 or loss, (3) Betting against Banker 4 streak up to 7 or win, (4) Betting against Player 4 streak up to 7 or win.

The strongest was (4) Betting against Player 4 streak up to 7 or win, it was actually +EV in my testing.

As I mentioned earlier, even if the testing is flawed by not having enough data, it won't do worse than random bets and will most likely do better.

 

Understandable...but I've seen a lot of twos in a shoe, and betting it to continue to four won't be something I do regularly....

 

Yes, having played a lot myself, if there is a string a twos forming, I would have to go with the twos no matter what the testing says. I have made a lot of money on twos. I don't know why people hate them.

 

@fathead - I appreciate your hard work my friend. Keep fighting the good fight.

So when you are playing this, what kind of Money Management are you using? I am finding a lot of people are using Labby for good positive results. I like this idea.

 

I've been testing an anti-streaks style so far shyowin a nice nets profits of about 3 units buckin up against 22, 000+ hand tested shoe. The 1s 2s 3s and4s are the most common event at the baccarats and the 5+ streak by far the least common occurrence. S, it only make sense logical that the shrewd cat wanna make his cake on the most common event and accept getting clipped by the least common event the 5+ streaks, hey hey.

 

soxfan

I have definitely made money playing Anti-5+ so I can see where his theory can work. Good Money Management will certainly solidify that bet selection although I'm not certain of a positive expectation.

 

I have tried many things and still change it sometimes but I always seem to come back to Labby for serious betting. It is probably the best one if you have patience, discipline and proper bankroll. You can get into trouble if you do not work it smartly, though. You must make it a normal practice to split and re-split the Labby into more than one string if it gets too big. If you have to split it (especially into multiple strings) it takes patience to clear all the strings. You should not even feel pressured to do it the same day. If you are tired you may not be thinking clearly. Get some sleep and clear the strings the next day.

One very important habit that has helped me immensely is to stop betting if I hit a losing period, make virtual bets, then start back betting when the winning returns. If you lose 7, 8, 10 bets in a row your string is going to get way out of control. If you start back and start losing immediately, don't be afraid to stop again and make more virtual bets until the picture clears.

Lately I have been using the string 0 20 20 30 30 ... if you start losing right off the bat just the 30 is added each time so it keeps the initial losses low. When I clear the string it is $100 or so. Sometimes I will add $5 to a couple of bets to get back commission. You can make a few hundred dollars per shoe pretty easy if you do not have to split the strings or only do minor splits.

 

Also, this strategy that I tested should just be added as another part of your arsenal at the table. Now when I see a streak of four Players, for example, instead of agonizing about whether I should continue to ride the streak (who knows, it may go to 20 right?) I know the best play is to bet against for 3 bets or win.

 

all the money in the world can't defeat a houses edge! but my system can

 

Baccarat , a game of pure luck and a money management scheme.

 

What are the results for betting chop only and for betting chop+streak.


If you see a chop of one, bet with the chop until four (or end of chop), then (if the chop continues) bet against the chop for a max of three bets (or until a win). If the chop continues beyond seven then you would sit out.

P
B
P
bet
X
X
up to four (or a loss) then bet
y
y
y
up to three max (or a win)

 

No, IT CAN'T!!!!!!

 

Good idea! See, you bring other people into the discussion and new avenues appear.

Overall, -0.89% house edge which is a actually a little better than normal. I think one problem with this is you have to have a two to end the chop, and singles are more common than twos. But good idea.