Summarizing:

- no way you can find a long term profitable betting plan without speculating that outcomes are not perfectly randomly placed as random bac outcomes are unbeatable by a 1 billion degree. 

-  to ascertain outcomes are not properly random produced only place selection and probability after events tools can help you by strict scientifically accurate assessments. Some bac productions are better than others, meaning they involve a higher unrandomness factor.

- best way to take an advantage without suffering the variance impact is by looking just for one unit profit per a given amount of hands.

- no matter how's your strategy and which side you choose to bet, each set of two consecutive wagers must get a way higher 75% probability to win. Considering as Banker side as a steady advantaged option is one of the biggest mistake to make. Asym hands favoviring Banker don't come out so often, especially whether consecutively taken.

-  the game cannot be altered or predicted by human considerations, otherwise it wouldn't exist.

as.

Making things in a more complicated way, we could set up many different r.w.'s utillizing a pace different than 1.
After all the general law of independence of the results should work no matter how deep we want to classify the outcomes, right?

Thus a BPBBPPBPBBBBBBPBPPPPBPBBPPB succession could be

1-2-1-1-2-2-1-2-1-1-1-1-1-1-2-1-2-2-2-2-1-2-1-1-2-2-1 (1 pace) or

1-1-2-1-1-1-1-2-2-2-1-1-2-1 (2 pace) or

1-1-1-1-1-1-2-2-2 (3 pace)

Again summing the two adjacent numbers from left to right we'll get:

1 pace) 3-3-2-3-4-3-3-3-2-2-2-2-2-3-3-3-4-4-4-3-3-3-2-3-4-3 (runs: 12)

2 pace) 2-3-3-2-2-2-3-4-4-3-2-3-3 (runs: eight)

3 pace) 2-2-2-2-2-3-4-4 (runs: 3)

Skipping certain outcomes provides a better evaluation of the place selection impact, that is the main factor by which certain subsequences must be considered as collectives or not.
And naturally in this example the best indicator is the number of runs.

We should convert what others call "stop loss" or stop wins" cutoff points with the simple number of runs, especially if we want to disprove a real randomness.

Without boring to test many shoes, it's intuitive that a kind of asymmetrical force is acting along the way on the vast majority of shoes dealt, our task should be directed to spot the shoes where such asym force will be more likely to act on certain points.

Now let's sat we want to follow two opposite players, one player A wishing to parlay his bet up to 5 steps toward a new same number situation (being 2, 3 or 4) and the other one B wishing to make a progressive plan toward not getting same number clusters (up to 5 steps).

Player A will win anytime 5 or more consecutive homogeneous situations will show up (2-2..-3-3..-4-4.. 3-3, etc) and player B will win anytime a given number won't be clustered up to 5 times.

From a math point of view both players will get the same results getting different W/L frequencies.
In the practice things go quite differently.

as.   

We've seen that every shoe in the universe can be considered just as a 2-3-4 sequence of runs.
In my example I've chosen to consider the simple hand to hand registration, meaning that every resolved hand will be eligible to be listed.
Moreover hands are considered by a simple B=1 and P=2 registration.

Now say we do not want to simply assign the 1 value to B and 2 value to P, instead 1 to an identical situation and 2 to an opposite situation taken at a given mechanically preordered pace.

If the results succession will be really randomly placed, we know this tool won't affect the dispersion values. Technically speaking, we want to disprove the common knowledge that any mechanical preordered plan will be insensitive to every place selection strategy. The only way to prove this game is beatable.

There are infinite ways to set up random walks trying to disprove a perfect randomness, being the runs distribution the common denominator.

Any bac hand/pattern distribution is a complex result made of three finite different forces acting along a slight dependent model:

1- asymmetricity favoring B side

2- very slight propensity to get the opposite result just happened

3- key cards distribution (low cards should be considered as key cards as 8s/9s)

Taking those three factors together some r.w.'s are more inclined to provide a higher number of runs.

as. 

Some studies show this fkng virus tends to spread more in high polluted areas (Wuhan, Northern part of Italy, Madrid, N.Y.C., L.A., etc).
Moreover warm wheather and high humidity seem to lower the COVID-19 virulence (Singapore, Malaysia) no matter how huge is the population density.
Probably Lungyeh could say something about this. 

In Vegas we must hope the warm factor will overcome the humidity one...:-)

as.   

Yeah....

Sadly the mortality rate of this COVID-19 is a lot higher (up to 5%)

as.

No surprise, this virus is extremely dangereous as very often goes unnoticed.

Probably some casinos will increase the number of stadium baccarat where players remain at a decent distance among themselves and from the dealer and no chips are involved.

as.

Next why some random walks applied to baccarat are better than others. The decisive tool to destroy this fkng beautiful game.

as.

Those new derived subsequences are not forming random successions as 2 cannot go to 4 and 4 cannot go to 2 without crossing the 3 step.
Moreover no matter how whimsical is the original BP succession, any shoe will produce a given number of  2-3 / 3-2 or 3-4 / 4-3 steps.

Notice that we shouldn't give a damned fk about the lenght of same level values, let alone the exact or approximated final number of runs. We instead should focus about the actual probability to get one or a couple of runs on different portions of the shoe.

If the original succession is perfectly randomly placed, the subsequent derived collectives cannot give us profitable betting spots as in order to get an advantage we must put in action certain random walks anyway.
I mean that a perfect random original sequence cannot form low dispersion values on derived situations no matter how sophisticated they are intended, what we really need to set up an unbeatable plan.

as.

Suppose we want to classify BP outcomes assigning 1 to any B result and 2 to any P result.
Thus a sequence as BPBBBPPBPBP becomes 1-2-1-1-1-2-2-1-2-1-2

Now let's add the number on the left with the adjacent number placed on the right in a way to build another subsequence.
In our example, we'll get 3-3-2-2-3-4-3-3-3-3

The number of "runs", that is situations where a number stays at the same level are transformed from 7 in the original sequence to 5 in the new one.

Before continuing let's see what happens on strong streaky BP situations as

BBBBBPPPBBPPPPPPPBBBBBBPPPPP =

1-1-1-1-1-2-2-2-1-1-2-2-2-2-2-2-2-1-1-1-1-1-1-2-2-2-2-2

then

2-2-2-2-3-4-4-3-2-3-4-4-4-4-4-4-3-2-2-2-2-2-3-4-4-4-4

here the number of runs is 6 on the original sequence and 11 on the new one.

or a "choppy" sequence as

BPBPBPPBPBBPBPBPBPPBPBPB

1-2-1-2-1-2-2-1-2-1-1-2-1-2-1-2-1-2-2-1-2-1-2-1 =

3-3-3-3-3-4-3-3-3-2-3-3-3-3-3-3-3-4-3-3-3-3

Number of runs shifts from 21 to 7.

Let's try to fictionally build a shoe getting many runs on our new sequence.
Easy to do, we need many different sums coming in fast succession.
Example:

BPPBBPPBBPPBBPPBBPPBB =

1-2-2-1-1-2-2-1-1-2-2-1-1-2-2-1-1  (runs= 9)

3-4-3-2-3-4-3-2-3-4-3-2-3-4-3-2 (runs= 16)

Nothing special so far, it's just another way to consider the hands distribution taken from a simple B/P point of view. A wrong point of view. But...


as.   

Let's see what happens with this fkng Coronavirus disease...it's not the thing US Health System wants to deal with.

as.

Putting things in a simple way, bac is beatable itlr as it's made by continuous asymmetrical propositions, most of them not easily detectable by common standards.
We are here to (partially) demonstrate that such constant asymmetricity (rules, card distribution, key cards concentration/dilution, finiteness of the shoe) will be endorsed by the paramount inference of unrandomness.

More practically speaking, profitable spots arise from a strict scientifical convergence of probability measure where different r.w.'s dictate or not to wager the same B/P result being the  reflex of a I/O situation.

as.

Before going into details of what a multi-level random walk is, let me know how the fkng fk you can lose by MM assessing three simple different one-step r.w.'s working on B double consecutiveness considered at the levels #1, #2 and #3. Where #1 and #2 scenarios take an astounding primary role.

Even if casinos know such B doubles detectable distribution, thus maybe voluntarly fixing outcomes to get a lot of consecutive B doubles, we can easily build many other r.w.'s collecting results by undetectable ways, mainly by coding results as I or O results thus negating a random distribution.

as.

Thanks Lungyeh, I hope to give you very soon a direct demonstration of what I'm talking about.

@Fran7738, you took the point.
I guess many casinos know that bac is beatable, the game is still alive as most players like to gamble.
At the winning rate you've suggested the probability of success is very very very close to 1.

as. 

What's what I name as a multi-level random walk?

It's a mechanical preordered betting scheme made by building one of the several subcollectives derived from the original BP succession. Not necessarily considering each outcome of the original succession.

As long as the attributes to build such subcollectives remain constant, we know that a supposedly random source must produce the same features on every new collective we had built. Regardless of place selection and probability after events features that definitely will confirm or not the real randomness of the sample. 

Next week more about the construction of such r.w.

as.

Another excellent contribute.

as.