Baccarat Sputnik's Baccarat Corner

Sputnik's Baccarat Corner

I think the discussion about developing new ideas is missing and rarely you see any discussion with details about how to tackle decision-making.
Rarely do you see a punter expose themself or their ideas! But they let the world know they have a winning strategy.

In this topic, I don't value claims or empty words about success or methods.
Here I will bring what I consider a good and solid way to tackle baccarat.

I will exclusively use stats and data and simulations from random org with my examples.

Cheers Patrik From Sweden

 

Thanks Patrik! I agree nobody shares ideas which is the point of a forum. Thank you for facilitating this!

 

One old and common strategy is playing to catch singles and series of two.
And where you make some decisions making how to handle the larger series that end each sequence.

And it looks like this.
Random org 2023 03 23

Code:

2
2 L
2 L
2

1
2 W
1 W
2 W
2 L
1 W
1 L
2 W
2 L
1 W
2 W
1 W
2 W
2 L
2 L
2
2


1
1 L
1 L

2
1 W
1 L
2 W
1 W
1 L
1 L
1

2
1 W
1 L
2 W
1 W
1 L
1 L

2
2 L
1 W
2 W
2 L
1 W
2 W
1 W
1 L
1 L
1
1

There are many different ways to tackle such periodic events.

a) One is to pinpoint out the series of two within each sequence using some kind of gap methodology.

b) Another is to aim for one win after one fictive win.

c) And the third is to use and develop a march where you play on a rolling basis where you have some clear decision-making for how to handle larger series that ends each periodic sequence.

Personally, I want to narrow down to a minimum amount of losses where two is fine for each negative chunk.
And that is pretty hard to accomplish using a march on a rolling basis.
So I tend to jump in and out rather than follow the flow nonstop to reduce the variance.


Extended variants are to cut the random bits into slices of four to create a periodic gap event.
Here we can apply the same principles of cutting losses short and jumping in and out between the gaps.
And it can look like this.
Random org 2023 03 23

Code:

2
2 L
2 L
2 L

1
2 W
1 W
2 W
2 L
1 W
1 L
2 W
2 L
1 W
2 W
1 W
2 W
2 L
2 L
2 L

2
1 W
1 L
1 L
2 W
1 W
1 L
2 W
1 W
1 L
1 L
1 L

2
1 W
1 L
2 W
1 W
1 L
1 L
2 W
2 L
1 W
2 W
2 L
1 W
2 W
1 W
1 L
1 L
1 L

Cheers Patrik From Sweden

 

Patrik

What is boils down to is that you are looking for a change from single to series and series to single. That seems to make a lot of sense. However, there can be those long strings of singles whereas there aren’t a lot of consecutive series.

 

After looking at your second set of results, are you saying that you would stop after three consecutive losses and then waiting for that fictive win?

 

AsymBacGuy wrote about a method in the past at another forum where you wait for a series of three or higher.
If I remember correctly.
Then you wait for singles or a series of two (fictive win) and then attack for a single/series of two.

This creates a loop of W, LW, LL
And I look at random bits and solutions where I can get either a W, LW, LL

And there is more options, let's say you get a direct win you can continue to strike until the first loss L.
And stop at LW and LL.

And another way is to continue to strike if you make a direct win and stop when you reach two loses LL.
That would in the worst-case scenario result in minus one -1 with the opportunity to jump on board long strikes.

Cheers Patrik From Sweden

 

Yep, the concept could be elaborated further, that is waiting for 'two' fictional winnings to show up, maybe waiting that a kind of 0,25% 'negative deviation' had reached some consecutive values (Marigny De Grilleau), as 3+ streaks do not come around any corner.

For that matter we could even consider 'doubles' vs 3+ streaks after the same negative deviation trigger had shown up.
But now we have to endure a greater impact of variance.

as.

 

Good morning Asymbacguy and all.

I will mention something reverse, the more sequences with singles that end with a series of three.
They can come in several small, medium or large sequences.

Here the likelihood of a series of 5+ is very high and a great opportunity to win two in a row.
That is the probability.

Here is the math.

Singles have a value of 1
The series of two has a value of 0
Series of three has a value of 1

Here you can see the two-sided sharp sword.
Where or if singles dominate and end with a series of three there will be a great imbalance.
And where two points are a small rather than medium or large regression, correction towards the mean.

And it continues with

Series of four has a value of 2
A series of five has a value of 3
Series of six has a value of 4
Series of seven has a value of 5
And so it continues ...

I will update this theory after craking some random bits from random org.
YES, this is one explanation of how to find a loophole that creates boundaries for larger series to happen in the future.
And that is based on math and probability without any crystal ball or guesswork.
You can test and see the benchmark for the values when the occurrence appears in different segments.
And then decide the risk and reward ratio.

Cheers Patrik From Sweden

 

Here are three examples with today's random bits from random org 2023-03-24.
You can see the letter S for singles and the end of a series of three until we reach above 3,0 STD.
And if we win the first bet there is no reason to continue for a second bet.

I will dive deeper into how capturing larger series, is not as simple as following a series of three to become larger.
As you can get a chunk of a series of three following each other to correct the regression towards the mean.
This will be a discussion about a march or algorithms to best capture the growth of larger series.

Sequence 1

Code:

2
2
1 s
2
2
1 s
2 s
1
1
2 s
1 s
2
2
1 s
2
2
2

1 s
2 s
1
1
2
2
1 s
2
2
2

1 s
2
2
2
1
1
2
2
1
1
2
2
1 s
2 s
1
1
2 s
1 s
2
2
1 s
2 s
1 s = 3,44 STD

2
2
2
2 W
2

1
1
2
2
2
2
2

Sequence 2

Code:

1
1
2 s
1 s
2 s
1
1
2 s
1
1
1

2 s
1
1
1

2 s
1 s
2 s
1 s
2
2
1 s
2
2
2

1 s
2 s
1 s
2 s
1 s
2 s
1 s
2
2
1 s = 3,27 STD
2
2
2
2 W

1
1
1
1

Sequence 3

Code:

2 s
1 s
2 s
1
1
2 s
1
1
2 s
1 s
2
2
1 s
2 s
1 s
2 s
1 s
2
2
1 s
2 s
1
1
2
2
1
1
1

2
2
1 s
2 s
1 s
2 s
1
1
2 s
1
1
1 = 3,70 STD
1 W
1

Here you can see a series of four and five plus chopping after 3+ STD.
Hope the example put some light on the subject and theory about catching larger series.

Cheers Patrik From Sweden

 

There are other solutions that wait or find a strong imbalance and you can ask yourself what happens when there is no imbalance (reaching 3 STD).
For example after six or seven singles, then your expectation is at least three series of three before the next six or seven singles will appear.
So if you can crack random bits and find a march to capture that segment less than 3 STD windows you will have a winning method.

March / Algorithms / Indications

The first and most famous solution by Marigny De Grilleau is to attack after +1.
And that makes sense, but what happens if the first bet loses and we get -1?

That is the big question.

Cheers Patrik From Sweden

 

Hi Patrik, thanks for sharing!!

I'm not sure to have understood why you assign 0 to doubles.

Anyway and assuming for simplicity 80-hand shoes and after having assigned the 0 value at doubles, what should be the average final 'count'?

Thanks in advance

as.

 

Good morning AsymBacGuy

First, I want to mention several ways to explore this and will say them below.
Also, talk about the playing model and put some clarity into the subject (why doubles have a value of 0).
I start with the last one.

Marigny De Grilleau talks about - Overrepresented events - & - Underrepresented events -
So when you have a playing method that aims to find larger series you need two opposite candidates that define the values for a 50/50 situation of what we can regard as Underrepresented and Overrerpresented events.

Singles have a value of 1
(is the event we want to be Overrepresented and create the imbalance and create a hole where larger series is missing, Underrepresented).

The series of two has a value of 0
(as they don't represent any value in getting larger series, the underrepresented event - you can see it as a loophole).

A series of three has a value of 1
(here we are close to getting our larger underrepresented events).

Series of four has a value of 2
(here we have a larger underrepresented event and regression towards the mean).

A series of five has the value of 3
(...)

Series of six has a value of 4
And so it continues.

And to put some light on the doubles and aim for larger series the playing model looks like this.

The series of two has a value of 1
Series of three has a value of 0
The series of four has a value of 1

A series of five has a value of 2
Series of six has a value of 3
Series of seven has a value of 4
And so it continues.

So 14 doubles with two series of four would be a 3.0 STD.

To use this in real life.
We just watch singles and series of two in sequences.
Then we want that each sequence ends with a series of three or at most a series of four, as they have the value 1 and 2.
So then is just a matter of counting singles.

There are two simple ways to memorize this.
12 singles versus 1 or 2 series of three (value 1 or 2) is a 2.5 STD or above
14 singles versus 1 or 2 series of three (value 1 or 2) is a 3.0 STD or above
And they can come with a series of four instead of two series of three where the series of four has a value of two.

When there is a stretch with more singles and more series of three you need to watch the chart to get the right value.
And you don't want higher series among the singles as you want the larger series to be underrepresented.
So we look and the windows of events among the random bits.

I see it as a vacuum pressure where the pressure is very high and will break.
Hope this help.

AsymBacGuy I have excel simulation software for this method where you don't need to do anything.
Just put in the random bits and the excel sheet find all sequences with 3.0 STD.
I like you and am willing to share that with you and I know you won't share it with others.

I still have your Gmail, do you want a copy?
Then I send you a copy today.

Cheers Patrik From Sweden

 

Outside the box - not baccarat - but a 50/50 prospect -

I also want to mention that I am into automating my betting using Bots online.
We have two today that can be used for Betfair Exchange BF Bot and Online Casinos with Roulette Extreme RX Bot.
I know one cost 10£ a month with a one-year subscription with BF Bot where you can use the Over/Under 2.5 goal market as a 50/50
2% commission and with Online Casino with La Partage wheel 1,35%.

So I will make some changes to my life by learning how to code RX and it seems okay after reading some parts.
Same with BF Bot.
And you can get coders for both and pay for coding, but I want to learn to master the whole process by myself.

You don't need AI as we can develop our own algorithms and marches.
If anyone wants a copy of the Betfair Exchange and how to use the Over/Under 2.5 goal market as 50/50
I can pinpoint where to download the strategy for free.
Should also mention horse races, lay the favourite, 50/50 and work with BF Bot I have 10 years of data to confirm march or the algorithm.

Cheers Patrik From Sweden

 

I forgot to put some light on this.

Let's say you use singles and a series of two and bet until you hit a series of three or higher.
Then three things can happen W or LW or LL

What you will notice is that the most common is sequences with singles and series of two to chop in chunks after each other for a whole shoe or a whole 300 trails with Roulette.
On rare occurrences, you will get larger series followed by larger series and this will result in LL and LL and LL and LL and so on.
But as you don't attack until you have a fictive win you don't bet and skip those sequences.

What can happen is that you get a fictive win and no more wins, and then you lose.
So the question is if you use a fictive win and after that wait for a fictive loss and then attack to catch doubles.
When you watch the sequences with singles and doubles you see that they chopping in chunks or where there are singles there are doubles.
So you pinpoint out doubles being part of strikes with singles and doubles.

The question is if you aim for one win and jump to the next sequence W.
Or if you make a direct win ride out for several wins like W LW LW W W W LW until you reach two losses LL.
Here you need to decide how to create your march or algorithm.

And also decide to do so after you see a 3+ series and aim to catch sequences with singles and doubles.
Or if you want to cut the random bits into sequences by four and bet between the chunks between series of four.
If you use a series of four then you will have W LW LLW LLL and with a fictive loss, you would have W LW LL.
And the reality is that each segment needs to have a win to avoid LLLL LLLL LLLL LLLL.
Also, those chunks with one fictive win and four losses would be your enemy.
Here you can also decide to aim for strikes W LW W W W LW LW LW W or face a LL or just aim for one win, W LW within each sequence.

Cheers Patrik From Sweden

 

Here is one way out of two when avoiding 3+ series.
The first one is to catch a win after a fictive win and stop after two losses or 3+.
The other way (with my example below) you can clustering the random bits into chunks of three versus singles and doubles.

Here is the result for 300 trails with today's file from random org 2023-03-25
LL W LW W W LL W W W LL W W W W W LL W W W LL W LW W W W LW W W LL W LL W LL W

Here is the output:
Here we avoid the black swan with a long string of losses with no fictive wins.

Code:

1
1 L
2 W
2 L L
2 L L

2
2 L
1 W
2 W
1 W W
2 W
2 L
1 W
2 W
2 L
2 L  

2
2 L
1 W
1 L L
2 W W
1 W
2 W
1 W
2 W
1 W
2 W
1 W
2 W
1 W
2 W
2 L
2 L

1
1 L
1 L

2
2 L
1 W
2 W W
1 W
1 L
2 W
2 L
2 L

1
2 W
1 W W
1 L
2 W
1 W
2 W
2 L
2 L

1
1 L
2 W
2 L L
2 L L

2
2 L
1 W
2 W W
1 W
2 W
2 L
1 W
1 L
2 W
2 L
1 W
1 L
1 L

2
2 L
1 W
2 W W
2 L
2 L

2
1 W
2 W W
1 W
1 L
1 L

1
2 W
2 L L
2 L L

2
2 L
2 L

2
2 L
2 L

2
2 L
1 W
2 W W
1 W
1 L
1 L

2
2 L
1 W
2 W W
1 W
1 L
1 L

1
2 W
1 W W
1 L
1 L

1
1 L
1 L

1
1 L
1 L

2
2 L
2 L

2
1 W
2 W W
1 W
1 L
1 L

1
2 W
1 W W
1 L
1 L

2
2 L
1 W
1 L L
1 L L

2
1 W
2 W W
1 W
1 L
1 L

1
2 W
1 W W
1 L
2 W
2 L
2 L

1
2 W
1 W W
2 W
1 W
1 L
1 L

2
2 L
2 L

1
2 W
2 L L
2 L L

2
1 W
2 W W
2 L
2 L

1
1 L
2 W
2 L L
1 W W
2 W
1 W
1 L
1 L

2
2 L
2 L

2
2 L
2 L

1
1 L
1 L

1
1 L
1 L

1
2 W
1 W W
1 L
1 L

1
1 L
2 W
1 W W
1 L
1 L

1
1 L
2 W
1 W W
2 W
1 W
2 W
1 W
1 L
1 L

2
2 L
2 L

2
1 W
1 L L
2 W W
1 W
1 L
2 W
1 W
1 L
1 L

1
2 W
1 W W
1 L
2 W
2 L
1 W
1 L
2 W
1 W
2 W
2 L
2 L

1
2 W
1 W W
2 W
1 W
1 L
2 W
1 W
1 L
2 W
2 L
1 W
1 L
1 L

2
1 W
1 L L
1 L L

2
2 L
2 L

1
1 L
1 L

1
2 W
1 W W
2 W
2 L
1 W
1 L
2 W
2 L
2 L

1
2 W
2 L L
2 L L

2
2 L
2 L

1
1 L
2 W
1 W W
2 W
2 L
1 W
2 W
2 L
2 L

2
1 W
1 L L
1 L L

1
2 W
1 W W
2 W
2 L
1 W

 

Here is another one using Sputnik's March with inspiration from the member Gizmotron.
He once explains how to bet against doubles using Dozens for Roulette.
So if you get a single dozen you play the other two dozen and I did not have success using it with dozens and Roulette.

But with Sputnik's March, I got pretty interesting results.

Singles is one event
Series of two is one event
Series of three and higher is one event

The March or Algorithm bet against doubles.
And there is several ways to attack.

a) One is to bet after a fictive win to get W or LW or LL
b) The other way is to get a W or an L by betting after a fictive L to catch the middle event of each sequence.
c) The third is to bet once for each sequence on a rolling basis.

Here are the results for today's random org file 2023 03 25

WWLL
WWLWWLL
LL
LWWLWLWWLWWLWLWLWWLL
LL
LWWWWLWLL
WLWLWWLWWWLL
WLWWWLWLWWLL
WWLWWLL
LL
WWWWLWLL
WLWLL
LL
WWWWWLL
LL
LWWWWWWWLL
WLWWLL
WWLL
WWLL
WLWLL
WWLWWLWWWLL
WLW

Alternative (a) results - W W W W LW LW W W LW W W LW W W LW W LW

Alternative (b) results - L W L W L W W W W L W W L L L W W L L W W W

Alternative (c) results - W W LL LW LL LW W W W LL W W LL W LL LW W W W W W W

Here you can see that alternative (a) is superior.
And here below is the random bits for Sputnik's March with today's random org file.
Mention this so you can see there is no curve fitting with real examples.

Cheers Patrik From Sweden

Code:

1
1
2
2
2 W
2
2
1
2 W W
1 L
2 L
2
1
2 W
2 W W
2
2
2
1
1 L
2 W
1 W W
2 L
1 L
2
1
2
1
2
1
2
2
2
1
1 L
1 L
2
2 L
1 W
2 W W
1 L
1 W
2
2 L
2 W
1
2 W
1 L
1 W
2
1 W
2 L
2 W
2
1
1 L
2 W
2 L
2 W
2
2
1
2 W
1 L
2 L
2
1
1 L
2 L
2
1
1
1
2
2 L
1 W
2 W W
2 W
2
2
1
2 W
1 L
1 W
1
1
2
2 L
2 L
2
2
2
2
2
2
2
2
1
2 W
1 L L
1 W W
1
2
2 L
1 W
2 W
1 L
1 W
1
1
2
1 W
1 W
1
1
1
1
1
1
1
2
2 L
2 L
2
1
2 W
1 L L
1 W W
1
1
2
1 W
1 W
1
2
2 L
1 W
1 L
1 W
2
1 W
2 L
1 L
1
1
1
2
1 W
1 W W
2
2 L
2 W
1
2 W
1 L
2 L
1
1
1
2
2 L
2 L
1
2 W
2 W W
2
2
1
2 W
2 W
2
1
1 L
2 W
2 L
1 L
2 W
1 L
1 W
1
2
2 L
2 L
2
2
2
1
1 L
1 L
1
1
1
1
2
1 W
1 W W
1
1
1
2
1 W
1 W
1
1
1
2
1 W
2 L
1 L
2
1
1
1
2
2 L
2 L
2
1
1 L
2 W
1 W W
1 W
2
1 W
1 W
1
1
2
1 W
1 W
2
2 L
1 L
1
2
1 W
2 L L
2 W  W
2
1
2 W
1 L
2 L
1
1
2
1 W
1 W W
2
2 L
1 L
1
1
2
1 W
1 W W
1
2
2 L
2 L
1
1
1
1
2
1 W
2 L L
2 W W
1
1 L
2 L
2
2
1
2 W
2 W W
2
2
2
2
1
1 L
2 W
1 W
2 L
2 W
1
2 W
2 W
2
2
1
1 L
1 L
1
2
1 W
2 L L
2 W W
1

 

Made one more test, because the alternative (a) had such good results, so I crack another 300 trials.
It could be one alternative for flat betting, but using it with a bot you would tackle variance with larger samples, so an extended fibo would do the trick.
Or playing by hand using a stop loss.

Here is yesterday's random org file 2023-03 24

LWWWLL
WLWLL
WWWLWLL
LL
LL
WWWWLWWLWWLL
WLL
LWLL
LWWWWWLWLWLWLL
WWLWWWWWLL
LWWWWLWWLWWWWLWLL
WWLWLL
WLWWLWWLL
LWWLL
WWWWWL
WWWLL
WWLL
WLL
WWWLWWLL
WLWLL
LWWLL
WWWLL

Alternative (a) results - W LW W W LL LL W W W W LW W W W W LL W LW W W

Alternative (b) results - W W W L L W L L W W W W W W L L L L W W W L

Alternative (c) results - LW W W LL LL W W LW LW W LW W W LW W W W W W W LW W

If we look at alternative (a) for 25/3 and 24/3 so did both win flat betting with +7 and +11 units!

Cheers Patrik From Sweden

 

Nice Patrik, lots of stuff to think about here. Thanks.

Ya, I had trouble with the rolliing basis when testing your previous series stuff.

I presently use a 4 column to break up the runs. Seems to help.

ECD tried to do this using designated counts for each series based on averages. It was a good idea. Maybe using an 80 decision Baccarat shoe not enough events?

J

 

Hi Patrik!!!
Thanks for your answers and thanks if you want to send me the mail!!!

as.

 

Hello Jim.

Regarding ECD I am not sure what values he used and I think (at least in theory) a window of events can be anything.
As I mention above, we know that a 3.0 STD window happens at 0,3% and we know one combination is 14 singles and 2 series of three.

So if we reverse the process we should have at least three series of three or above in each following situation betting against 0,3%.
So imagine 7 singles and 1 series of three in any combination, then the expectation is at least two more 3+ within the next 7 events.
I can imagine there are other ways to reverse the situation.

Cheers Patrik