Let's take the second pattern happening at every shoe dealt, BP registration still stands.

A
S
S
A
A
A
A
A
A
A
A
A
A
A
A
A
S
A
A
A
A
A
A
A
S
A
A
S
S
A
S
A
A
A
A
A
S
A
A
A
A
A
S
A
A
A
A
A
S
A
S
A
A
A
S
A
A
A
A
S
A
S
A
S
S
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
S
S
S
S
S
A
A
A
A
A
A
A
A
S
A
A
A
A
A
A
S
A
A
S
A
A
S
A
A
S
A
S
A
A
A
A
S
S
A
A
A
A
A
A
A
S
A
A
A
S
A
A
S
A
A
A
A
A
A
S
A
A
S
A
A
A
A
A
A
A
A
S
A
A
A
A
A
A
A
A
A
S
S
A
A
A
S
S
A
S
A
A
S
A
A
A
S
S
S
A
A
A
S
A
A
S
S
A
A
A
A
S
A
A
S
A
S
S
S
A
A
A
A
A
A
A
S
A
S
A
A
S
S
S
A
A
A
S
A
A
A
A
A
A
S
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
S
A
A
A
A
S
S
S
A
A
A
A
A
A
A
S
A
A
S
S
A
A
S
A
A
A
S
S
S
S
A
A
A
S
A
A
A
A
A
S
A
A
A
A
A
A
S
A
S
A
A

as.

Now we're taking care of the very first pattern coming out per every shoe dealt by an A (asymmetrical) or S (symmetrical) shape (according to the rules I've depicted in my post).
To make things easier let's consider just the BP sequence.

A
S
S
S
A
A
A
S
A
A
S
A
A
S
S
A
A
A
S
A
A
S
A
A
S
S
A
A
A
S
A
A
A
A
S
A
A
A
S
S
A
A
A
S
A
S
A
A
S
A
A
S
A
A
A
S
A
A
A
A
A
A
A
A
A
S
A
A
S
A
A
A
A
A
S
A
A
A
A
S
A
A
A
A
A
A
A
A
A
A
A
A
A
S
A
A
A
S
A
A
A
A
A
S
A
A
A
A
A
A
A
S
A
A
S
A
S
A
S
A
A
S
A
A
A
A
S
S
A
A
A
S
A
A
A
A
A
A
A
S
A
S
A
A
A
A
A
S
A
A
A
A
A
A
S
A
A
A
A
S
A
S
A
A
S
A
A
S
A
A
A
S
A
A
S
A
S
S
S
A
A
S
A
A
S
A
A
A
A
A
S
S
A
S
A
S
A
A
A
A
A
A
A
A
S
A
A
A
A
A
A
A
A
A
S
A
A
A
A
A
S
A
A
A
A
A
S
A
S
A
A
A
S
A
A
A
S
A
A
A
A
A
A
A
S
A
A
A
S
A
A
A
A
S
S
S
A
A
A
S
S
A
S
S
A
A
A
A
A
A
A
S
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
S
A
A
A
A
A
A
A
A
A

Any decent bac soul will get the best of such a succession whatever the possible innumerable permutations such short list might produce (That's the beauty and the importance of assessing outcomes by a clustered or isolated parameter).

Anytime you'll register the very first pattern happening at every shoe dealt (providing to be really randomly shuffled), enlist it to a simple A/S table and let me know whether your findings are not correspondent to this brief list.


as.

Quote from: Whatswhats on January 05, 2026, 09:18:22 PMIn my test, probably is better to bet just one time, the second bet, so wait for a virtual loss on a 2 bet succession or in my way that use what b/p pattern succession is I and what is C, can happen that first bet isn't possible (because i use multiple player playing simultaneously with W/L succession so not B/P) and can happen that first bet isn't possible and second yes so I bet only on the second bet, and the entire shoes that I'm tracking is random 4 hands for shoes, so I enter in a shoes and 4 hands that is my amount of hand for I/C, and then I go next shoes to bet etc, so I work with Asym/sym on multiple shoes that together create a single shoes!

I'm not English so probably my explain isn't the best!

I understand what you've sayed, with your multiple registration you'll get some "conflicting" events where one line suggests a B bet and the other one the P bet.
Probably the second bet is best to assess a kind of asym/sym propensity, providing to abandon the plan after two-three losses happening at the same shoe (there's a reason for that I'll address in a future post).

KFB wrote:

However, my personal belief is that one can do both: Use expectations as a baseline or guide, yet, focus on the shoe at hand(currently right in front of us). IMO this is superior vs doing solely one approach.

Where I consider expectations-as-a-baseline a little more is when the shoe has significantly deviated from expectations(2-3 SD),...etc. Especially in the early part of shoe. Dependent on how extreme the limits have been approached.

I concur.

After all most players follow the shoe and all of them are long term losers and trying to exploit the "following the shoe" approach mixed with the expectation guidelines needs a lot of experience and study that surely KFB and Alrelax got in their arsenal.

Independently of the expectation, any moderate/strong deviation is the essence of the asymmetry, yet in turn asymmetry/asymmetry becomes a symmetrical situation, thus only asym/sym and sym/asym patterns remain asymmetrically placed. That's why, IMO, we need a fair amount of hands to possibly resolve the issue.

On the other end and as we've seen several times here, sym/sym patterns are way more likely to stop after short sequences as the game isn't symmetrical and independent by any means.

A question for KFB about this passage:

One of the easiest ways to approach an even-steven game: A) Play at EZ Bac tables, B)Side step 1 out of every 2 3c7 B wins.

Can you please elaborate the B) point?

Thanks in advance!

as.

Suppose your next bet will be Player.
You'll really lose money itlr only if the Player draws and Banker shows a 4 or 5 initial point (actually even if the Banker shows a 3 but by a way lesser degree).
At all the other possibilities, a P bet is not losing money and of course 6s, 7s, 8s and 9s initial points will get a perfect equal probability to show up but the payment will be different, being 1:1 all of the times at P side but less than that at B side at commission games.
On the other end no commission games will get an inferior ROI whenever a B 6 will win (Tiger/Lucky 6 tables) or when a B three card winning 7 will show up.

From a mere asymmetrical/symmetrical patterns point of view and besides the math implications, that means that we should have more reasons to bet P at Tiger/Lucky 6 tables and conversely to bet B at EZ baccarat tables.
That means that a possible symmetrical 'unwanted' pattern should be evaluated by an additional factor, for example about how the previous hand went by the asym/sym math propensity that, of course, is less likely to happen consecutively whether the previous hand was asymmetrical and anyway regardless of the actual outcome.

That doesn't mean that betting P after an asymmetrical hand had shown up will get us an advantage (it would be too an easy task), just that the 'general' asymmetry will take care even of this factor when deciding the side to wager upon.

Our tests have told us that it's a minor mistake to chase asymmetry while betting two consecutive steps at P side than by wagering two consecutive steps at B side. A kind of paradoxical finding we should be aware of.

oOoOo

Back to the S successions considered by isolated (I) and clustered (C) shapes now at their SECOND appearance per every shoe dealt. Again first registration follows a specific random walk then we'll take care of the mere BP successions.

I
C
C
I
C
I
I
C
I
I
I
I
I
C
I
I
I
I
I
I
I
C
I
C
C
I
I
I
I
I
I
I
I
I
I
I
I
I
C
C
I
I
I
I
I
I
C
I
I
C
I
C
I
I
I
I
I
I
I
C
I
I
I
I
I
I
I
I
C
I
C
C
I
I
I
I
C
I
I
I
I
I
I
C
C
I
I
I
C
I
I
I
I
C
I
C
I
I
I
I
I
I
C
I
I
I
I
I
I
C
C
I
I
I
I
I
I
C
C
I
C
I
I
I
I
I
I
I
C
C
C
I
I
I
I
I
I
I
I
I
I
I
I
I
I
C
I
I
I
I
I
I
I
C
I
I
I
I
I
I
I
I
C
I
I
I
C
I
I
C
I
I
I
I
C
I
C
I
I
C
C
I
C
I
I
I
I
C
I
C
I
I
I
I
I
C
I
I
I
I
I
I
I
I
C
C
I
C
I
C
C
I
I
I
I
I
I
C
I
C
I
I
C
I
I
I
I
I
I
I
I
I
C
I
I
C
I
I
C
I
I
I
I
I
I
C
I
I
I
I
I
I
I
I
I
I
C
C
C
C
I
I
C
C

Now the BP registration:

I
I
I
C
I
I
C
I
I
I
C
I
I
I
I
I
I
I
I
I
I
C
I
C
I
I
I
I
I
I
C
I
I
C
I
I
C
I
I
I
I
C
C
C
I
C
I
I
I
C
I
I
I
I
I
C
I
C
I
I
I
I
C
I
I
C
I
I
I
I
C
I
I
I
I
I
I
I
I
I
I
C
I
C
I
I
I
I
I
C
I
I
I
I
C
I
I
I
I
I
C
I
I
I
I
I
I
I
I
I
I
I
I
I
C
I
I
I
I
I
I
C
C
I
I
C
C
I
I
I
C
C
I
I
I
I
I
C
I
I
I
I
I
C
I
I
I
C
C
I
C
I
I
C
I
I
I
I
I
I
I
I
C
I
I
I
I
I
I
I
I
C
C
I
I
I
I
I
I
I
I
I
C
I
I
I
I
I
I
C
C
I
I
I
I
I
I
I
I
I
C
C
I
I
C
C
I
I
I
I
C
C
I
I
I
C
C
I
I
I
I
I
I
C
I
I
I
C
I
I
I
I
I
C
I
C
I
I
C
I
I
C
I
C
C
I
I
I
I
I
I
C
C
I
I
I
I
C
I
I
I
I
C
I
I
I
I
I
I
I
I
C
I
C
I
I
C
I
I
I
I

Maybe there's nothing to guess, just to exploit such probabilities...

as.

Besides of the different approaches anyone would like to adopt, at the end what really counts is the proven math edge our bets will get (or not) over the course of our action.

Since it's impossible to play with a real advantage by simply varying the betting amounts, we're compelled to run long tests and see whether our single bets are really surpassing the well known winning cutoffs of 51.3% at B bets and 50.1% at P bets.
That is the bet selection is of paramount importance.

Regarding those winning probabilities, we know that B bets are affected by the actual rules, so at "Tiger/Lucky 6" tables we need a way higher winning probability than 51.3% to be ahead itlr as we must endure a greater HE (1.46% instead of the common 1.06%); on the other end EZ baccarat tables reduce the HE by a 0.05% (1.01% instead of 1.06%).
Obviously HE at P bets remains at -1.24% at all tables.

If we think our approach will dictate to place half of our bets either at B or P side (even by running complicated other forms of random walks), we see that our B winning probability will be enlarged by few situations (ranges) whereas P winning probability will get "less unfavourable" successions (from a math point of view) as the asymmetrical math probability to have B math advantaged hands is 91.4/8.6.

This consideration could be particularly important anytime we choose to bet very few hands, especially whenever our plan will instruct us to bet two hands in a row.

After all situations really favoring the Banker side are distributed by a 8-9 hands range per each 8-deck shoe (and of course some of those situations-around 10%-resolve in a tie).

More later

as.

A Wonderful New Year to everyone!

as. 

Quote from: Whatswhats on December 29, 2025, 11:26:01 PMHi asym, hoping everything is fine!

I have two question.

1.
First sequence every I/C is a shoes right?

2.  With B/P successions what you mean exactly? Any B/P successions or what?

Hi!
Exactly!

as.

Exactly, but more often than not patterns will prolong or stop at some detectable points especially if we'd run the same situation infinite ways.

For simplicity say we're taking care of the very first S pattern happening per every shoe dealt.
We'll stubbornly bet that that first S pattern will be come out as isolated (A-S) instead of being clustered (S-S...)

Here's a "short" list of real shoe results (I=isolated S pattern, C=clustered S pattern) according to a specific random walk:

I
I
I
I
I
C
I
I
I
C
I
I
I
C
I
I
I
I
I
I
I
I
I
I
I
I
C
I
C
I
I
C
C
I
C
I
I
I
C
C
C
I
I
I
I
I
I
I
I
I
C
I
I
I
I
I
I
I
C
I
I
I
I
I
I
I
I
I
C
I
I
C
I
C
I
I
I
I
I
I
C
I
I
C
I
C
I
I
I
I
C
I
I
I
I
C
I
C
I
I
C
I
I
I
I
I
I
I
I
C
I
I
I
I
I
I
I
C
C
I
I
C
I
C
I
I
I
I
C
I
I
I
C
C
I
I
I
C
I
C
I
I
C
I
I
I
I
C
I
C
I
I
C
I
I
C
I
I
I
I
I
I
C
I
I
I
I
I
I
I
I
C
I
I
I
C
I
I
I
I
I
C
I
I
C
I
C
I
I
I
C
I
I
C
I
I
I
I
I
I
C
I
I
I
I
I
I
C
C
I
I
C
I
I
I
C
I
I
I
C
I
I
I
I
I
C
I
C
I
I
C
I
I
I
I
C
I
I
I
C
C
I
I
I
I
I
C
I
I
I
I
C
I
C
I
C
I
I
I
I
I
I
I
I
I
C
I
C
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
C
C
I
I
I
I
I
I

I'd guess that it wouldn't be so difficult to get an advantage of such succession even if you are changing the sequence by all the possible permutations coming out in your mind.

Make things easier and let's see what'll happen at the same shoes just considered by mere BP pattern sequences:

I
C
C
I
I
C
I
I
C
I
I
I
I
I
I
C
I
C
I
C
I
C
I
C
I
C
I
C
I
I
C
I
C
I
I
I
I
I
I
C
I
I
I
I
I
I
I
C
I
I
I
C
I
C
I
I
I
I
C
I
I
C
I
C
I
I
I
I
I
I
C
I
I
I
I
I
I
I
I
C
I
I
I
C
I
C
I
I
C
I
I
I
I
I
C
I
I
I
C
I
I
C
C
I
I
I
I
I
I
I
I
I
I
C
I
I
C
C
I
I
I
C
I
I
C
I
I
I
I
I
I
I
I
I
C
I
I
I
I
I
I
I
I
I
I
I
I
C
I
I
I
I
I
I
I
C
I
I
I
C
I
I
I
I
I
I
I
C
C
C
I
I
I
I
I
C
I
C
I
I
I
I
C
I
I
C
C
I
I
I
I
I
C
I
I
I
I
C
C
I
I
I
I
I
I
I
I
I
I
I
I
I
C
C
I
I
C
I
I
I
I
I
I
I
C
I
I
I
I
I
I
C
C
I
C
I
I
I
I
C
C
I
I
I
I
C
C
C
C
I
I
I
I
I
I
C
I
I
I
I
I
I
I
I
C
I
I
I
I
I
C
C
I
I
I
I
I
I
I
I
C
C
I
I
C
C
I
I
C
I

Such "simplest" succession seems to be quite similar to the previous one, but in the process of building it we accounted for a way higher number of 2-step winning situations than 1-step situations, so suggesting that BP asym/sym feature needs a relative higher number of hands dealt to privilege the asymmetry.
This finding should suggest us that whenever we take the simple BP succession as the "main line" to rely upon, most asymmetrical impact will show up, on average, at the second step meaning that we'll get more wins at the 2-step than at the 1-step.

as.

Besides of the strategy employed, when we play every hand we'll get an inferior number of long winning streaks than long losing streaks for the simple reason that most patterns get a low level of univocal presentation so making difficult (impossible) to be more right than wrong.
Conversely, playing towards low levels of presentation will suffer from the inevitable strong deviations.
In the meanwhile HE adds more damage to our bankroll.

The same about betting half of the hands dealt or 1/3 (or more) of the hands dealt.

If you stay away from betting you are not losing anything, I don't recall any long term winning player betting more than few or very few hands per shoe.

Make an experiment and register how many times you would have won or lost if you were to bet every hand, then half of the hands dealt and so on. The different W/L lines will esponentially get better results in inverse relationship of the betting frequency (HE discounted).   

The AS/S "scheme" is an example of that.
Despite of being very slight unfavorite to proportionally come out over S patterns by mere quantities, A patterns are generally more manageable to be forecasted just because at baccarat clusters of something are slight more likely to happen for card distributions and rules issues.

On the other end, if S>A, 1-step A winning ranges must be more likely included within short ranges so making more room to the slight likely S patterns happening at the 2-step.
And when a S cluster had shown up in the actual shoe there's a slight probability it will repeat again, so restraining us to wager toward getting the future S as isolated.

oOoOo

Since we are strong believers that the only way to ascertain an advantage is by flat betting, say we want to collect many fictional players working for us and wagering different A/S situations.
The basic assumption we rely upon is the AVERAGE RANDOM card distribution.
Obviously and even though it appears as somewhat complicated, that's just a simplified version of what we're actually using at the tables.

Player #1 will be responsible of the mere 1-step A situation, he will register how many consecutive times he would have won or lost the first attempt and is set up at -3/-4. Meaning that he will advice us to bet in order to reverse ONE TIME that negative range.
It's true that the same specular positive cutoff value (+3/+4) should be more entitled to reverse toward the negative field, but after reading Alrelax valuable suggestions and knowing that actual shoes are not so randomly distributed we preferred to let it go a positive amount without betting or possibly riding the positive 1-step wave (see below)

Player #2 will be responsible of the 2-step A situation, acting by the same guidelines seen above for the Player #1.

Player #3 will take care of the A clustered appearances (so full betting the two steps) knowing that isolated A coming out in a row, despite of being relatively rare, are more likely intertwined by S isolated events.

Player #4 is the one supporting long term data telling us that A clusters longer than three vs A clusters of exact lenght of three (A3+:A3 ratio)  will be distributed by a super high balanced ratio favoring a betting towards A3+ patterns. There are several options to exploit this feature, for sure a strong negative variance is very very very unlikely.

Notice that I haven't included in this simplified plan the S shape appearance, at least by direct terms.
First, it doesn't make much sense to be so interested about a 0.25 probability distribution.
Second and more importantly, actual shuffling machines seem to make more probable S clusters giving us no valuable insights to assess their average impact, mostly as their production is not random.

BTW, there's a reason why Player#1 and #2 won't get precise cutoff values (-3 or -4 and +3 or +4), let's wonder why. 

as.

The basic foundation why we should play baccarat with an edge relies upon the probability that along the course of each shoe some events are more likely than expected, negative and positive variance considered.

Sayed that, we've made interesting studies about different bac productions, anyway knowing that in some instances things could change as casinos (or shuffling machine manufacturers) are not babies in the wood.
Neither we are.

Basically we've concluded that nowadays many shoes dealt lack of the "true random" feature as we compared them with a more reliable randomness source and you know what I'm talking about.
The fact that shoes are not properly random shuffled doesn't necessarily mean that "patterns" are getting more or less detectable forms, just that the cards are not randomly shuffled.

That's why we started to consider shoes by the asym/sym feature getting more constant values at real random shoes but strong volatile situations at supposedly unrandom productions.

So the "average shoe" concept I kept stressing in this thread works only at true random shuffled shoes.
Then what to do when we'd think the shoes we're playing at are unrandomly shuffled?

Well, our basic answer (besides of the important fact that we can't be so wrong while following, at least partially, what the shoe is producing) is to assess the AS/S factor of the patterns that can be devised by infinite ways (sub random walks applied to the original BP sequence).

Rhythm and symmetry are the key words.

For example, ABABAB..sequence is rhythmic and asymmetrical up to the step 3 whereas a succession as AABAABAAB..is rhythmic and asymmetrical.

If (key) cards are asymmetrically distributed and the rules favor an asymmetrical distribution of the results, we could dare to make the conclusion that patterns will be more asymmetrically distributed than symmetrically shaped.

On the other end, rhythmic symmetrical patterns (ABAABB.. or AAABBBAABBA..) are less likely to happen or, at least, getting low sd values to happen.
Once a same pattern had shown up symmetrically (so two times in a row) we let it go until it shifts into a different pattern. Since we have learnt that patterns should be only classified into 1s, 2s and 3/3+s, we know we have to register in the AS/S form the three most common pattern situations.

As a general guideline, isolated S are more likely to come out in a row and when they are not most of the times A clusters will show up.
When a S cluster (more likely in the S-S double form) shows up, our data suggest that the future probability of getting another S cluster vs an isolated S is slight endorsed, that is getting a higher than 1:3 probability to appear (remember we utilize a 0.75%, two-step probability to win).
More generally speaking, an abnormal number of S events happening at one shoe will teach us that asymmetry is somewhat "silenced", a sign that we shouldn't bet (or bet again) at that shoe.

Talking about the 0.75% probability to win, one could argue that there are no differences between a single/double vs 3/3+s streaks plan or the AS/S plan and that's the beauty of picking up the subtleties of this game.
Technically the first plan relies upon the unlikelihood to get many shoe portions of side predominance of 6:1 or 6:2 (or more), the AS/S plan will make a kind of conditional probability to be working as any shoe dealt presenting strong S pattern predominance more often than not will deny a proper A balancement whatever considered.

Here some shoes considered by the AS/S feature and according to the rules described in my pages. Since some casinos will cut off a lot of cards from the play the sequences will be considered by "cores".

1) A-A-S-A-A-A-A-A-A-S-(-1)

That's an 'easy' shoe, all A clusters and all isolated S.

2) S-A-A-A-A-A-A-A-S-A-S-A-A-A-A-A-A-S

Another good shoe except for the isolated A happening at 11th spot of the succession.

3) A-A-A-A-A-S-A-A-A-A-A-A-A-S-S-A-A-A-S-S

Notice the S clusters repeating at the end of the shoe.

4) A-A-A-A-A-S-A-A-A-A-A-S-A-A-A-A-S-A

A perfect shoe

5) S-S-A-A-A-A-A-S-A-A-S-A-A-S-A

The first S cluster makes less appealing the next S situation. In this scenario an isolated S event happened but I wouldn't suggest to bet for A after the second S came out.
Just in case, all A came out in clusters.

6) A-A-A-A-A-A-A-A-A-A-S-A-A

Despite of its univocal shape, this sequence will get us just three wins before vig.

7) A-A-A-A-S-A-A-S-A-A-A-A-A-S-A-S

According to the staregic rules depicted above, here it's impossible to lose.

A-S-A-A-A-A-A-S-S-S-S

A strong S deviated shoe that shows up the importance of not chasing A after a S clustered sequence.

9) S-A-A-A-A-A-A-A-A-S-S-A-S-A-A

A slight negative shoe

10) A-S-S-A-S-S-A-A-A-S-A

A relative unlikely shoe where up to 8 consecutive bets were lost in the initial portion of the shoe unless the first S cluster had refrained us to play for the next S to be isolated.

In total we got +2 units before vig despite of having an unfavourable AS/S 109/115 pattern ratio.

See you in a couple fo days

as.

Considering symmetry (at least by the parameters depicted in this thread) as a kind of losing situation no matter what is the best start to look for bac patterns predictability, IMO.

On the other end, asymmetry moves around more likely values splitted into 1-step and 2-step winning spots.

If A-A is more controllable than A-S and S-A is more controllable than S-S our only 'enemy' are A-S-S patterns.
Taking a deeper step, S-S-A are more controllable than S-S-S-A and so on.

Providing a shoe to be really randomly dealt , isolated S are slight overhwhelming clustered S at any derived road considered, moreover clustered A will be slight superior than isolated A especially when S came out clustered.
It's interesting to notice that by utilizing a 0.75% probability to succeed, overall and itlr A impact tend to be lower than S impact, meaning that A is more likely than S by a kind of "quality" factor.
That means that waiting for a S or S-S sequence to appear is the best way to collect a long term valuable edge, obviously coming out by wagering towards an A event one time.

In turn this edge could be better defined by wagering that 1-step winning situations are more likely to prolong or to stop or simply to show up "out of blue".

Next time I'll present some examples of what I'm talking about.

as.

Good suggestions for sure (and I really mean it) but without a verified edge the player's bankroll will invariably decline no matter how sophisticated is our MM.

Nonetheless it's very likely 99,9% of the bac players aren't going to follow those guidelines as they will try to replenish the buy-in lost in the previous sessions, often by wildly pressing their bets.
Paradoxically doing so they are lowering the math disadvantage by implementing the 'Bold strategy' that teaches us that at any EV- game the best move to adopt is betting large amounts in just one hand or very few hands.

I mean that if we play with an edge the more we play the more we'll win (so the session concept is worthless), but in absence of an edge MM procedures just dilute the eventual loss of the entire bankroll.

Naturally if every player would follow your guidelines casinos will win a lot less money than they actually do.

as.

KFB wrote:

  I find it helpful to view every shoe(& thus every pattern) from a perspective: What is the expected average and how far has this current shoe deviated(- or +). I view a shoe as a combination of patterns that are always trying to approach their respective expectations(& limits).


I agree.

Example.
Say we could bet when a natural will show up, a kind of side bet where any natural is payed less than its math probability to appear (to guarantee the house an edge, of course).
Well, this bet would be so beatable that actually no casino is offering it. And I'm not talking about card counting 8s and 9s, just assessing the average rhythm of presentation with all related statistical features (gaps, card rank combinations forming naturals, symmetry/asymmetry, etc).

It's obvious that the fact that we don't need to guess the exact side where the natural will fall helps a lot.
Nonetheless what really help is that 8s/9s and natural combinations as 7-2, 6-2, 6-3, 5-4 and 4-4 cannot disappear for long and are in relationship of how many NON naturals had happened in the past portions of the shoe.
Then more shoes are played and registered and various probabilities tend to approach more and more the expected ranges of apparition. 

We could make the same considerations about the asymmetrical hands favoring the Banker side, now being a more complicated issue to be devised as their probability drops from 34.2% (naturals) to 8.6% (asym hands).

Going further, and that's what really interests us, we might take care of B/P patterns and now we're moving the problem into more volatile situations that still must obey to math/statistical laws.
For example we take B/P doubles, the easiest pattern formed by two consecutive hands that 91.6% of the times are the product of 50/50 exact situations.

Now differently than the above situations (naturals and asym hands), we should focus our attention about the consecutiveness of such doubles, therefore we must wait for a first B or P double appearance.
Curiously doubles are the prototype of symmetry.

More later

as.

If you'd think that results are somewhat dependent (the decisive factor why we should win at this game itlr), you need to let go a lot of hands without any betting as most of the hands dealt are decided by a chance variable.

For example let's consider the AS/S nature of the outcomes, first by a cumulative (two-step) probability then by distinct steps (first step and second step).
Put those outcomes into three different lines (A=cumulative, B=first step, C=second step) and you'll get more precise situations to start (or not) the betting after having ascertained how long the streaks (or singled patterns) will show up on average and by which level of consecutiveness.

A) cumulative AS/S patterns

We already know that S isolated patterns are the best patterns to look for; once they came out clustered (S-S), we  stop the betting now priviliging the AS clustered patterns.
The same about AS clusters (AS-AS) as opposed to isolated AS events (AS-S).
In fact there's a strong balance between isolated S vs S clusters and/or AS clusters vs isolated AS.

That means that if you want to get a substantial edge over the house, waiting for a moderate/strong imbalancement of S clusters vs S isolated events and/or AS isolated events vs AS clusters will do the job.
That's not a typical example of gamblers's fallacy as cards are more distributed to get asymmetrical results than symmetrical ones, especially if we take care of the infinite random walks originating from the BP original sequence.

B) first AS step

No matter the random walk utilized, long streaks of first AS attempts are quite rare to happen, say that most of the times winning streaks are stopping at 2 or 3 steps. That doesn't mean that after 2 or 3 winning steps we should bet towards S patterns (so switching the trigger), simply that we are not interested to bet anymore.
On the other end, losing first AS attempts will more likely stop after 2 or 3 consecutive situations, meaning that first attempts are not getting long winning streaks so often.

C) second AS step

Once the first step loses, the second AS attempt will intervene.
Again, long streaks of winning or losing streaks are slight less likely to happen, mainly as 2/2 and 3+/3+ streaks cannot be "clustered/clustered" for long and for distinct portions of the shoe without being intertwined by all other patterns.

See you in a couple of days

as. 

Hi Al!
Interesting comments from your part.

In the long run: well, for the vast majority of players the "long run" is way shorter than hundreds or thousands of shoes, not mentioning the complete unintelligent tests made on 100 million shoes produced by "we really do not know from where those F productions came out"...(of course I'm referring to math experts trying to extract an edge at side bets).
Laboratory tests are worthwhile only if they reproduce the exact LIVE situation we'll bet at, otherwise they lack of the irreplaceable replicability factor.

In a word there are subtantial differences among different bac productions wrongly considered as similar or correspondent. The long term B/P ratio (that almost always will approach more and more the 0.5068/0.4932) is not a valid tool to assess that different productions could be considered as similar.

Anyway I agree that each shoe is a world apart, yet knowing that itlr (say within 20-30 shoe ranges) some lines will be more likely than others should help, IMO.

Quit when you are ahead: That's the most st.u.p.id suggestion serious players can hear about games.
The related sentence should sound as "Stay when you're behind", the perfect recipe to destroy every bankroll in the world, naturally while playing without a verified edge.
When things go in our favor, maybe ti's better to stay and stay and stay.
So you are 100% right about that.

Trend following vs presentement recognition
I'll take this important topic in the way we did where the "presentement" word is an intricate parameter putting in relationship long term situations with actual situations: such parameter at selected spots will never ever be 1 (50/50 or, more correctly, 50.68/49.32).
Hard to do, I agree too. At least without the use of algorithms we can use at live tables.

Do not chase
As you sayed, surpassed certain narrow distribution lines (actual or expected), we have no reason to chase that is we have no reasons to get the "expected" when the shoe is "too" biased.
Probably the best rule to follow when playing baccarat.

Good thread.

as.