Thanks both KFB and klw for your interest!

I've started to write a Smoluchoswki derived strategy but I've seen it's a too complicated strategy to post here and for sure we can get the same profits by adopting easier concepts.

What I name as 'runs' is the number of same situations shifting from one state to another.
For example a BBPBPPPPPPPBBPBBBBPBP pattern is made of 9 runs.

The same about a byb sequence as bbbrrbrbrbrrrrbbbrrbrbbrrr: it contains 15 runs.

Itlr the number of runs follows the expected general probability providing a random production.

To 'bring down the house' we need to know either the general probability to work but more importantly an actual evaluation about how good or bad are shuffled the cards.

For example, it's literally impossible to get all steady states happening at a given shoe and the same is true about unsteady states.
Imo it's a strong mistake hoping to get too many steady states for long and of course is a worse mistake to get unsteady states coming out for long as cards are not randomly shuffled, thus what happened is slight more likely to come out again than to get an opposite situation.

Rain is very unlikely in Las Vegas, yet on my very first trip there I've got 5 straight raining days in a row.
After all sh.it happens in clusters.

See you tomorrow

as.

Marian von Smoluschowski was a Polish Physics Professor that made several interesting studies about Brownian motion and many other topics, including the several mentioned here 'probability after effects' (I voluntariily changed the second word with a 'events' word).
Simplifying a lot, a given independent 0-1 binomial succession having a 0.5 probability to appear (the event come out=1 or not=0) considered at different consecutive steps will form another succession by adding the previous number with the very next number, now the new succession loses the 'random' factor as some values cannot be reached.
   
Example:

B=1 and P=0

Say the original sequence looks as 1-1-1-0-1-0-0-1-0-1-1-0-1-0-1-0-0-0-1, if we want to add two consecutive 'values' we'll get:

2-2-1-1-1-0-1-1-1-2-1-1-1-1-0-0-0-1.

Naturally the value 2 may go back to 1 or remaining at 2, but it can't never directly go to 0; the same about 0: no way it can jump to 2 without before going to 1.

If the original sequence is really random and independent (and not affected by the natural asymmetry working at baccarat), the derived subsuccession (albeit being unrandom), will present 'unbeatable' values typical of a coin flip proposition.

Nonethless, a simple 'runs test' made on reliable samples will deny this possibility, in the sense that the original succession is more or less affected by a kind of unrandomness. Of course at values higher than the expected natural 'asymmetry' working at B favor.
Otherwise we're just considering 'natural' situations and not 'actual' situations.

In the above example and considering the Smoluchoswki subsuccession, we got 7 'runs' whereas at the original sequence the runs number was 12.

Now let's consider a strong polarized streaky shoe where the original sequence looks like as:

1-1-1-1-1-1-1-1-1-1-1-0-0-0-0-0-0-0-0-1-1-1-1-0-0-0-0-0-1-1-1-1-0-0-0-0-1-1...

the Sm. succession will be:

2-2-2-2-2-2-2-2-2-2-1-0-0-0-0-0-0-0-1-2-2-2-1-0-0-0-0-1-2-2-2-1-0-0-0-1-2....

original sequence got 7 runs and the Sm subsequence got 12 runs.

The exact counterpart of what happened at the first shoe.

Now let's compare how many shoes will get the original/derived suuccesions getting more runs at derived roads than at original sequence.

To falsify this hypothesis let's take three different real and very polarized shoes portions:

1) 0-0-1-1-0-0-1-1-0-0-0-1-1-1-0-0-1-1-1-0-1-1-0-0-1-1-0-0-1-1-0-0-0-0-0-0-0-1-1-1-1

that is a Sm derived sequence as

0-1-0-1-2-1-0-1-2-1-0-0-1-2-2-1-0-1-2-2-1-1-2-1-0-1-2-1-0-1-2-1-0-0-0-0-0-0-1-2-2-2...

original sequence got 16 runs and Sm subsequence got 31 runs.

2) 1-0-1-0-1-0-1-0-1-0-1-0-1-1-1-0-1-0-1-0-1-0-1-1-0-0-0-1-1-1-1-1-0-1-0-1-0-1-0-1...

SM derived sequence is:

1-1-1-1-1-1-1-1-1-1-1-1-1-2-2-1-1-1-1-1-1-1-1-2-1-0-0-1-2-2-2-2-1-1-1-1-1-1-1-1...

original sequence got 29 runs and Sm sequence just 9 runs.

3) 0-0-0-1-1-1-0-0-0-0-0-1-1-1-1-1-0-0-0-0-0-0-0-0-1-1-1-1-1-1-1-0-0-0-1-1-1-1-1...

original sequence got 8 runs and Sm subsequence got:

0-0-1-2-2-1-0-0-0-1-2-2-2-2-1-0-0-0-0-0-0-0-1-2-2-2-2-2-2-2-1-0-0-1-2-2-2-2...

15 runs.

Notice that regardless of the actual distribution,  the overall number of 'shifting' numbers at Sm subsuccessions will be superior to the number of 'same value' clusters, now matter how's the fk cards are arranged.

In the four shoe fragments displayed, and assigning a W at shifting clusters and a L at losing situations we'll get:

1- L L W L W L

2- L W L W L W L W

3- L L L W W L W L W L L

4- L W L W L W L W L W L W L W L

Notice again that Ws came out clustered just one time, so no 'positive variance' in the common intended sense had acted along the way, yet Ls are more likely to be followed by a W than by a L, another way to consider results.

Now we have the tools to set up a 'bringing down the house' strategy by adopting a careful multilayered progressive plan that makes baccarat as the best game to take casinos' money.

See you in a couple of days.

as.

Hi klw!
Your observation is a good one, imo 

Shoes are just numbers and numbers are the reflex of patterns lenght


There are innumerable possible BP (and derived sequences) combinations at baccarat, yet we could take care just of the most likely patterns coming up along the way, getting rid of the most unlikely ones.
Of course we do not know the precise numbers distribution but we know that something must happen in a way or another at different degrees of probability

For example, if we'd bet toward getting two singles in a row at the start of the shoe, we know the general probability to win is 0.25, the same about getting two doubles in a row but now we need a double to come out as the very first pattern. In the sense that after a BB or PP sequence (that is a BBP or PPB) we'll bet respectively toward another P or B then playing the opposite side to limit that consecutive second double.
Again the probability is 0.25.

When we'd consider triples, we need a triple to come out and we know that any 3+ streak belongs to this category.
Thus BBB or BBBBBBBBB, or PPP or PPPPP is a triple.
Now to get another back to back (consecutive) triple of any kind we must bet two times the same side after the triple trigger has shown up.
So after BBB(P) we'll bet two times P, or after BBBBBBBBB(P) we'll bet two times P and the same about P triples. And guess what, the probability remains 0.25.

Oppositely thinking the probability not to get two consecutive same patterns in a row is 0.75 and of course it's the same fkng general probability to get two consecutive bets winning (or losing now by a 0.25 value).

But a baccarat shoe is not a 'sky's the limit' proposition, if it were we couldn't get a single possibility to win itlr.

I mean that those probabilities are dynamically placed along each shoe either for 'natural' reasons but more importantly for other features where a possible 'bad shuffling' takes a primary role.

More on that later

as.

Hi klw and ty!

Yep. the game is quite complicated to learn (imo), patterns are important but rank cards distribution matters...
You won't find many REAL LIVE shoes samples that consider ranks; best way is to deal shoes by yourself then registering everything.

Let's go back to my unb #2.

The basic strategy is quite simple: we'll only take care of B patterns in form of singles, doubles and triples (3+s) distribution.
So we'll start the betting (better sayed, the registration) after a new B hand comes out.
People who have read that thread know that in some way B doubles clusters are considered as a constant 'enemy', so we'll classify the B1 (singles) and B3+s distribution happening along each shoe dealt in terms of 'profitable' sections.

For obvious reasons, the probability to get the whole shoe providing just 1s and 3s at B side is well higher than the probability to get only 1s and 2s, not mentioning the very distant probability to get all the shoe showing 2s and 3s without any single in between.

Mathematically speaking, B1<B2 but B2<B3, overall the 1-3 vs 2 probability is very close to the expected 0.75 value.
Hence to get a long term profitable plan either we should raise the 1 percentage or the 3+ percentage.

In some way we should evaluate how many times 'coin flip' situations must shift toward one side or the another and how huge is the actual finite asymmetrical strenght favoring B side.
Moreover we must take into account the very slight propensity to get the opposite outcome already happened (good when wagering toward B singles and bad when wagering toward B 3's).

Let's randomly take a 10-shoes sample from my live shoes datasets:

1,1,3,2,2,1,3,1,2,1,1,3,2,3,3

1,1,1,1,2,2,2,1,2,3,2,3,3,1,1,1,2,1

3,1,1,1,2,1,1,1,2,3,1,2,2,1,1,2,1,1,2,3

1,1,1,3,1,1,1,2,1,3,1,1,2,3,3,1,1

1,2,1,1,2,2,1,1,1,1,2,3,1,3,2,1,1,1,3,2,2,1,1

1,2,1,1,1,1,3,1,3,1,2,2,2,1,2,2,1,3,1,1

2,2,2,1,1,2,1,1,2,2,1,2,1,2,2,2,3,1,1,1

1,1,1,1,1,1,3,3,1,1,3,3,1,1,1,1,1,2,1,1,3,2,1

1,1,2,2,1,3,3,2,3,1,1,1,1,1,1,1,1,2,2

3,1,2,3,3,3,1,1,2,1,2,1,2,3,1,3

Taken as 1s and 3s as wins and as 2s as losses and adopting a 1-2 mini progression (W=+1 and L= -3) we'll get an overall - 12 units loss (vig ignored for simplicity) so no short natural positive variance was involved here.

Nevertheless, a careful 'sections' W/L distribution will help us to define that an asymmetrical (yet mathematically proportional) betting made on a sure asymmetrical and dependent production cannot reach the unbeatable limits typical of pure symmetrical and independent situations.

Even at shoe #7 (10 2s and 10 1s/3s, a statistical abnormality) W streaks must come out, after all just one B 3+s streak had come out there (test your shoes and see how's unlikely is this happening).

It's like that profitable spots are surely happening along the way and of course they're not coming up when opposite situations seem to show up clustered at various degrees.

More on that next week

as.

10 shoes randomly taken and considered by 2,3 and 4s patterns.

1- 4 4 2 3 2 4 2 3 4 4 2 3 3 2 4 4
2- 2 2 2 2 3 3 2 4 2 4 2 2 2 4 2 3 3 4
3- 2 3 2 4 2 4 2 3 3 2 2 2 2 3 2 2 4
4- 4 3 3 4 2 2 4 2 4 3 3 3 2 2
5- 4 4 2 2 4 2 3 2 4 4 4 2 2 2 4 (3)
6- 2 2 2 2 4 2 3 3 4 2 3 3 3 2 4 3 4 2
7- 2 3 2 2 4 2 33 3 3 1 1 4 2 4 2 2 4 2
8- 4 3 3 3 3 2 2 2 2 4 2 4 2 2 2 2 2 2 (3)
9- 3 2 4 2 4 4 2 2 3 3 4 2 4
T- 2 2 4 4 2 4 2 3 3 2 2 2 2 2 2 2 4 4 2 2

Notice shoes #5, #6 and #7 all present the same back to back results (in bold) in the identical five positions and I've chosen randomly ten shoes from my datasets. Moreover at shoes #6 and #7, even columns six produced another same back to back outcome.

Sh.i.t happens, we know, but crossing the probability to lose 12 bets in a row....

So let's see about the unb plan #1.2

1- L W L W L L L W W L W

2- W W L W W W W W W L W L

3- W L W W W L W W W W W W W L

4- W W W W L W

5- W W W L W L W W W W W W *

6- W L W L L W W W L L

7- W W L W L W W W W W L W W W W W W

8- W W W L W W W W W W W W *

9- L W W W W W L W L W

T- W W W W L W W W W W W W L W W W

W= 91 (+ 91 units before vig)
L= 29  (- 87 units)
*= 2  (- 2 units)

Nothing to be appalled at, yet....

as.

Quote from: klw on September 30, 2021, 02:22:05 PM
Hi AsymBacGuy  -- Great writings as usual and I am very grateful for one for all this information.

If you have the time is there any chance you could do an example of the above system so I can " get it " 100% ?

Cheers.

Hi klw and thanks!

Any shoe dealt is formed by a number combination, singles are 1, doubles are 2, etc.
In another way of thought, each shoe presents a 'general' speed having accelerating spots (singles and doubles) or slowing down spots (triples and superior streaks).
We know that the average speed is influenced most by lower classes of 'speeds' as they are the more likely occurences.
Even knowing that in the long term accelerating spots vs slowing down spots ratio will be close to 1.

Yet the speed cannot be uniform along any shoe as cards are asymmetrically placed.
Moreover, different speeds acting at different shoes and considered by multiple levels cannot get the same positional strenght for long, especially if we'd consider back to back shoes getting the same production.

Example. 4= 4 or superior streaks)

Shoe #1 (Palace Station, LV)

PPBPPBPPPBBBBBPBPPPBPPBPBBPPBBBPPPBBPPPBBPBPPPPBBBPPBBPPBPPPPPPBP

is a 2-1-2-1-3-4-1-1-3-1-2-1-1-2-2-3-3-1-3-1-3-2-1-1-4-3-2-2-2-1-4-1 number succession.

compare this shoe with the actual next shoe dealt at PS:

BBPPBBPBPPPBBPBPBBPBBBBPBBPPPPBPBPPBPPPBPPBPPBPBBBPPBBBPB

that is

2-2-2-1-1-3-2-1-1-1-2-1-4-1-2-4-1-1-1-2-1-3-1-2-1-2-1-1-3-2-3-1

Get rid of singles (1) and see what those two shoes look as after positionally coupling the patterns higher than 1: (two bets in a row after any 3 or 4 pattern coming on the first shoe)

2-2 (-1)
2-2 (-1)
3-2 (+1)
4-3 (+1)
3-2 (+1)
2-2 (-1)
2-4 (+1)
2-2 (-1)
3-4 (+1)
3-2 (+1)
3-3 (-3)
3-2 (+1)
2-2 (-1)
4-3 (+1)
3-2 (+1)
2-3 (+1)
2-x
2-x
4-x

Naturally it will be more probable to lose after any 2 spot (implying just one bet to get a W or a L) than after any 3 (two bets needed half ot the times to get a W) or 4 (two bets needed half of the times to get a W).

Since 3s and 4s are less likely to coming out clustered and since singles tend to confuse the positional back to back situations, we can safely assume that only a whimsical and very unlikely confrontation will get us many losses in a row, that is every column will get more often than not a different pattern already happened or randomly taken for that matter.

But more simply let's see the distinct 2,3 and 4 distribution on each shoe.
As explained on my unb plan #1, instead of considering 1s, 2s and 3+s now will take care of 2s, 3s and 4+s, so betting toward 2-3 and 2-4 clusters after those two distinct events happened at least one time and stopping the bet whenever 3-4 or 4-3 patterns show up.

First shoe: 2, 2, 3, 4, 3, 2, 2, 2, 3, 3, 3, 3, 2, 4, 3, 2, 2, 2, 4.

Second shoe: 2, 2, 2, 3, 2, 2, 4, 2, 4, 2, 3, 2, 2, 3, 2, 3.

First shoe: L, W, W, W, W, W, W, W, L, L, W, W, L.

Second shoe: W, W, L, W, W, L, W, W, W, W, W.

Notice that the cumulative number of W and L is equal, that is 6 losses and 18 wins (minus vig).
It seems that the first betting 'comparison back to back shoe' plan would get us an edge.

So let's start a 'random' test to check out what will be the most profitable course of action to be taken.

as.

Quote from: alrelax on October 01, 2021, 12:02:24 PM
Each section must be viewed and partially wagered independently without total reliance upon the previous sections.  Unless there are long streaks and a large total unbalance of winnings hands in your current section.

That's right, yet the problem still stands when to consider the starting and ending section points.
Obviously it would be a blasphemic strategical plan to stop to bet after a single win when a long winning streak appears while we would prolong the betting after the same counterpart single loss (instead of waiting another possible favourable section).

Since any shoe dealt will present a finite number of sections, raising the theorical probability of success to values higher than 50% will help us to define better the WL 'chopping' or 'clustered' lines happening at most part of the shoes.

as.

One more randomly taken 10-shoes sample regarding 2-3 and 2-4 unb plan #1.2
(*= 1 unit losing hand, a transitory triples not getting the room to be classified)

1- W W W L L W W W W W *

2- W W W L W L W W W L W W W W

3- W W W L L L W W L W L W W W L W L

4- W W W W W W W W W L W W L W W W

5- W W W L W W W W L W W W W W W W W W

6- W L W W W W W W W W L

7- W W W W W W W L W W W L W W W W W

8- L W W W W W W W L W W W L W W

9- W W W W W W W W W L W W W W W W W

10- W W W W L W L L W W L *

W= 119 , that is +119 units before vig

L= 28 plus two 1-unit losses (*) that is -86 units

In this sample I've caught a W/L ratio fairly shifted toward the left.

Now you probably should get a better idea of what 'bringing down the house' means when talking about baccarat.

as.

Thanks a lot Al!

Hi KFB!

Splitting the actual results by two bets terms in order to get a win could help us to define better what we're trying to accomplish.

Let's speak about patterns.
Say we set up our plan by first wagering toward singles: the first winning attempt is made at singles, if we win that's it. What happens next doesn't interest us, we collect the money, period.
If we lose, we'll get two same hands in a row and now we must decide if we want to bet toward doubles (betting the opposite hand) or triples (betting the same hand). This is the second part of our plan.

What's important is that this second bet cannot overcome mathematically the first losing attempt even if all second bets will win and even though the second bet is higher than the first bet (a necessary condition to get a win after two bets when the first is lost).
In fact if we knew that second attempts will get more wins than losses (or must get a kind of 'more likely' probability after a first loss), why not waiting a fictional first loss and then starting to bet?

Naturally singles fight against streaks, doubles vs triples and so on.
Therefore the only option to study is about their general distribution compared with the actual distribution.

Even though general probabilities dictate that everything is possible, heterogeneous distributions cannot happen for long along each portion of the shoe. As card ranks are asymmetrically distributed along the shoe.

Hence from one part we must spot the situations to get a first win, from the other end we must take care of the actual distribution when seems to produce 'clustering' events of different shape.
The term 'clustering' is a direct reflex of the asymmetrical cards distribution, but for the nature of the game something being surely asymmetrical seems to produce 'symmetrical' probability patterns.

For obvious reasons each shoe produces a card distribution following, itlr, the bell curve shape.

Our tests have shown that best profitable shoes are those placing themselves near the top of the bell curve or quite distant from it.
Near the top shoes are ridicously beatable, distant from the top shoes could be either a heaven or a nightmare. Of course the 'nightmare' may be transformed into a 'heaven' by adapting at most our strategy to the actual shoe.

In my unb plan #1 I've considered to register and play 1-2 and 1-3 patterns, stopping whenever a 2-3 or 3-2 pattern will come out. Thus shoes particularly rich of singles will be more likely to show up long winning clusters.
Nothing prevent us to add a superior patterns level, that is betting toward 2-3 and 2-4, now shoes particularly full of doubles will be more likely to win (now singles become not influential). Here the 3-4 or 4-3 pattern is a stopping sign.

And so on with the 3-4 and 3-6 betting plan, etc.

It's like that whenever the trigger pattern value come out combined with the next value or the +2 value will produce lines affected by low degree of variance.

Let's take randomly 10 shoes from my datasets by considering 2-3 and 2-4 patterns (say unb plan #1.2).
W= +1 unit and L= -3 units.

1- W L L W W W W W W W W L W

2- W L W W L W W W W W W L W W

3- W W W W L

4- W L W W W W  L L

5- W W W L W W W W L L W W

6- W W W W W L L W W L W W W W W

7- W W W W L W L W W W W L W

8- L W L L W W W L W

9- W L W W W W W W W W W L W L W

10- W W L W W L W W L W W L W

Total W= 87 (+ 87 before vig)
Total L= 30. (- 90)

Fortunately (lol) at this sample I didn't catch up a favourable W/L ratio, we've lost money and furthermore burdened by the vig, nonetheless (not displayed here) we've won more bets at the 2 level (that is the first step of the two-step procedure) than by backing up the first loss by a second wager.
Not mentioning that distributions of winning or losing spots could help us to choose when and how to bet.

Even not taking into account specific patterns as I did, every shoe results splitted in two betting sections are made by a number of first wins and second wins (with the same losing counterpart); in order to win itlr, we need to be more right than 50% (51.3% or more when betting B) at the very first hand wagered. Everything different from this will produce an inevitable loss and lack of control over the outcomes.

Alrelax is completely right about his 'sections' way of thinking the game.
Some sections of the shoe are playable, most part of them aren't.
I would add that some shoes are unplayable no matter what.

Finally it's interesting to note that most successful bac players rarely take a seat at the table.
They prefer to bet by standing behind other players.
They do not give a lesser s.h.i.t about the privilege of peeling cards or getting comps.
Moreover, it's easier to quit a table either as a winner or a loser by not seating at it.
If we are not seated at a table at the eyes of the personnel we simply don't exist, yet our bets are.

as.

An old gambling quote states that players are no guaranteed to forecast sure winning bets along the way.

A banal assertion that when applied at baccarat needs some specific considerations to be made.

Of course probability (to win or to lose) cannot be 1 and neither 0 as such those extremes fit the concept of 'certainty' (the event will surely happen or the event can't happen).

Nevertheless long term 'human' observations made on reliable samples tend to negate that 'set in stone' assumption. In the sense that 1 and 0 can reach those extremes in many ways, thus assuring or not the opposite certainties.

Besides considering exceptionally rare and only theorically possible situations, any shoe dealt will produce a BP succession and infinite derived sub sequences getting a finite number of patterns of different quality and probability to happen.

Assuming a perfect equal betting placements between B and P side, to get a long term edge we need to 'be right' by at least a 50.7% winning probability. Giving a lesser fk about the vig, that is before vig.

Every other series of 'back to back bets' must follow proportionally this value, hence and for example to get two bets in a row as profitable itlr (whatever diluted they are) on average we should win 75.7% of the times.
And so on by taking into account proportional values working at three or more bets in a row and so on (but we recommend to stop at two value).

Tricks to reach or to enlarge at our favor the 75.7% winning ratio after two hands wagered.

Easiest way to get multiple wins in a row is about lowering the first losing attempt. We can't put much pressure about the second bet, I mean. We must proportionally share the same 'probability' weight on both bets.
Actually and after studying long statistical samples, each 'two-step' bet has shown to get a larger overall impact over that 75.7% winning ratio than what the second bet can do.
Contrary to common belief it's like that being right at the start of those two bets will be more profitable than trying to win the second bet after a first loss.

Of course losing two bets in a row must be considered as a sort of imminent disaster as every two-step bet getting a 0% winning value will get a huge impact over the future two-step bet situations.

The beauty of this game is that 'two in a row loss' spots whatever considered, are way more limited than what the fkng math values dictate.
Along each shoe dealt, the probability to be right by wagering two hands in a row cannot be 0 if we would consider clustered mixed events made of singles or doubles or triples as a starting betting point.

To make a vulgar example about a 'bringing down the house' strategy, consider how many times consecutive 3+ streaks or doubles at any fkng random walk you wish to classify will come out clustered or isolated and you'll see that low values are way more likely to be followed by other low values and/or that high clusters will be more likely followed by low degree clusters of the same shape.

Say that at small road we're betting that consecutive doubles won't make three or more doubles in a row, stopping the betting after a win (on either two spots) or getting a two-step loss (that is another consecutive double will come out).
That's our random walk we want to put in action.
For the absolute 'lack' of dependency stated by mathematicians, we are supposed to get the same number and sd values of two double isolated spots than superior than two double spots.

This is a complete wrong fkng bighorn.sh.it.

According to our live shoes sample, at sr the probability to get consecutive doubles at a value higher than two is way more limited than what a mere coin flip proposition dictates.
So astute players trying to get the best of it by exploiting this feature are going to get a substantial long term edge over casinos.

After all most bac pros do not give a fk about trends, just about probabilities happening on selected circumstances.

as.

One area Im always striving to improve is recognizing there is nearly always a little bias showing. Just hold up awhile as most shoes will have at least one or two  groups of 9-12 outcomes where a significant  biased opportunity is available. Readily available for a huge cash extraction. 


+1

Hi KFB and thanks again for your interest at reading my pages.

Imo baccarat is a game of multiple observations made on several different aspects.

For example, how many times an asymmetrical hand comes twice in a row, so building the Banker math advantage on the second hand?
General probability tells us is 8.6%, hence it's way more likely to cross a symmetrical hand of course NOT favoring Player but even less Banker as on the latter scenario a winning bet is payed 0.95:1 and not 1:1.

Therefore when we place a series of bets at Banker side we must encounter a more restricted range of asym hands than average to get a math advantage.
And there are only two ways to be really advantaged when wagering Banker:

- Spot the situations where an asym hand will come out shortly.

- Hope that the actual shoe we're playing at provides a larger than expected number of asym hands.

Since the average asym advantage of B bets vs P bets is 15.86% but winning bets are decurted by a 5% vig, it's easy to see that there are very few spots to really get a math advantage (naturally 5% muts be considered as 2.5% per bet as half of the bets at symmetrical spots are losing).
Do the math and realize how many asym hands must show up to get a ROI (return on investment) advantage when betting Banker.

In reality casinos particularly love 'only Banker' bettors because they extract money either at losing bets  but even at winning bets. As long as the shoes they are providing not contain a substantial higher than average amount of asym hands, they collect some money.

Another important feature of the game that almost nobody would take care of is the quality of winning hands.

The largest part of winning results is formed by strong or moderate card situations, think about naturals and standing points or two-card higher initial points.
Nonetheless some rare shoes will elicit the formation of long clusters of winning hands (whatever shaped) by many 6-card hand situations or by a unfavorite side hitting multiple 'miracle' cards for long.

And many other situations in between.

Sadly to say, most 'betting many hands' strategies getting a profit at the end of the shoe would meet some unfavorite situations transforming a more likely losing hand into a winning hand.
It's what I name as a 'poker tournament' effect: eventual poker tournaments winners not only managed to win their advantaged hands but even some underdog confrontations (not mentioning a fair amount of so called 'coin flips').

This effect is partially erased by operating a strong and diluted bet selection.

Say that to be constant winners at baccarat we need to play a lot of shoes thus collecting imperfect informations about how the game works and for each of them to make multiple observations that are not worth per each shoe dealt.

In fact, probabilities continuously hop from one side to another, no one future hand will get a 0.5 probability to happen. And to get a long term edge we just need to bet B with at least a 51.3% probability and P with at least a 50.1% probability.
It can be done.

as.

Consider those 6-deck live shoes:

1)

PP
BBB
PP
B
P
B
PPPPPP
BBB
P
B
PP
BBB
PP
B
P
BBB
PP
BBB
P
B
PP
BB
P
BBBB
PPP
BBB

Doubles went as singled, singled, singled, singled, singled, doubled.

If we were to place a bet after any double toward not getting another double, we would have got several wins in a row, that is 5 wins in a row stopped by the final doubled sequence.
Interestingly, at big eye boy road we got just two 3+ streaks, the remaining hands produced singles and doubles.
Even at beb road doubles went as single, double, double, single, single, single, 4 consecutive doubles.

Even though some spots are colliding (that is forecasting an opposite event) playing toward getting an isolated double will get us a possible statistical edge.

Think about NOT getting three doubles in a row, now the colliding force will smooth up to the point that clusters of two doubles in a row will get a very low variance probability.

2)

BBBB
PP
BB
PPP
B
P
BBBBB
PPP
B
P
B
PPPP
B
P
B
P
BB
PP
BBBBB
P
BBBB
P
B

Now at big road doubles went as doubled, doubled.
At big eye boy road we got 3 doubles in a row, two doubles in a row and one isolated double.

At main road (big road) and considering doubles as enemy, every column produced a win but at columns #2 and #3 and #17 and #18.
At big eye boy road 'losing columns' appeared at hands #9 #10 and #11 and #13 and #14.

Notice that at both roads (big road and byb road) consecutive doubles are more likely stopping at 2 level, in the sense that the probability to get two consecutive doubles vs superior doubles clusters is way more restricted than what a strict independent world dictates.
In addition and more often than not, long clustered doubles will elicit the formation of several 'no doubles patterns' as the probability to get simmetrical 'double' (thus consecutive) patterns  will be lowered.

Practical reflexes on actual betting

A fair amount of shoes will present a total isolated 'doubles' happening, a careful key card distribution study will be an additional tool to ascertain this.
That means that those shoes producing all isolated doubles cannot give us more than two losses in a row.

The more doubles are clustered, higher will be the probability to spot the situations where 1s and 3s prevail.

Doubles are more likely to show up when the actual shoe provides few singles and many streaks.
It's like that at the actual shoe we're playing at, consecutive streaks burn the half probability to isolate doubles at the first attempt.
So avoid to bet when the results seem to be 'streaky'.

Doubles are working as triples, with the important effect that they are slight more likely to show up.
That means that they are more likely to show up clustered, so enlarging the probability to spot the situastions where they won't act along a shoe.

as.

Yep, this is a classical example where general probabilities seem to be disregarded and actual probabilities are so polarized that every player should quit the table as winner or huge winner.
Jackpots happen and we should try to be there when they happen.

Imo almost every shoe presents one or more spots where a range of hands is more probable than the counterpart and we know there are infinite ways to dissect the outcomes.
Of course so called 'profitable' spots will fight against 'undetectable' spots, thus the only way to play a EV+ game is to get rid of those undetectable spots the more than we can. In the sense that we can't transform the undetectable world into a profitable world by the use of progressions or other 'human' tools.

I wish to present another way to consider a shoe by classifying as 'enemy' the most likely baccarat two-side occurences at Big Road: the doubles. A contradiction in terms, right?

Doubles as enemy

Our plan will be to bet toward singles and 3+s at any side, so doubles will produce two losses in a row, the remaining patterns will be winners (by a light two-step progression) or break-even or so spots.

Of course along every shoe doubles will distribute by different degrees, since we'll start our classification by waiting a first double appearance we'll get:

- single double (that is a double followed by a single or a 3+ streak)

- two consecutive doubles (xBBPPB or xPPBBP) followed by a single or a 3+ streak

- three consecutive doubles

- more than three consecutive doubles

Since doubles must follow a general probability to show up, we may infer that the more doubles are clustered, higher will be the probability to spot 'doubles-free' portions of the shoe.
Naturally it could happen that shoes particularly rich of doubles won't be a realiable source of profitable betting, but there's always a counterpart situation where doubles are so rare (thus singled distributed or by forming rare two-consecutive patterns) that most part of shoe is bettable toward singles and triples.

More on that later

as.

Hi klw!

They aren't.

I suggest to consider only real live shoes, the remaining part is just worthless rubbish.

as. 

Hi klw!

Itlr and discounting ties B probability is 0.5068 and P probability is 0.4932. So it's not a coin flip proposition, moreover is a light dependent process.
Simplyfing on average we'll get one more B hand than P hand per every shoe dealt.

'Itlr' and 'on average' are the worst player's words when talking about gambling.
On the contrary those two words are the best casinos' allies in order to make a lot of money.

So, yes, that average more B hand per shoe should fall after a B single (thus forming a B streak) or to stop any P pattern coming out.
If that more B hand would fall at any already formed Banker streak it simply prolong it just one step, so not changing the B streak/B single ratio or shapes.

Very frequently shoes will show long series of B singles without any B streak or few B streaks in between. Thus the potential B math propensity goes right down the toilet.
More importantly, any B winning bet is payed 0.95:1 and not 1:1 as any P winning bet, easy to see that betting B in those circumstances 'no matter what' constitutes a big mistake.

In the long run the mathematical mistake between B and P bets is just a 0.18% value favoring B, yet in the short run the payements remain hugely shifted toward one side.

Think of those shoes where the final number of P hands overcome the number of B hands or, conversely, of those shoes where B streaks were very prevalent but unnecessarily short payed more often than not.

Cheers

as.