Baccarat is a game producing constant clustered events in either way as the shoe distribution is affected by a huge degree of concentrated/diluted card factor.
Many times such clusters will act in the form of 'easy detectable' patterns, others are not.

The easiest detectable clustered forms come out from either long B or P (or r/b) streaks or long chops, then other 'next level' situations come around depending upon how many times we would want to challenge an 'isolated' form to show up.
After all even series of isolated spots constitute a cluster.

Since we have no valid reasons to think that every single shoe we're playing at is really randomly shuffled, we must be forced to consider every shoe as a single entity.

For example, the natural B propensity could go down in the toilet as well as the more likely predominance of P side to get long singles and doubles sequences.

Consider the most basic plan of all: singles and streaks.

Most of the times, sections full of clustered singles deny the probability to get streaks and vice versa.
Each clustered events will get a value different from zero unless a single-streak or streak-single situation o ccurs.
This last scenario could easily come out when sequences as BPPBPPPBPPBPPP or BBBBPBBPBBBBPBB show up. But even now we consider it as a cluster.
Notice that whenever a cluster of this last kind happens, the unfavorite side will get only singles otherwise we'll get a streak cluster ending up this clustered particular event.

Now say that we'll mechanically registering how many singles and streaks will come out assigning them a value:

0= single followed by a streak or streak followed by a single, zero clusters on either side;

1= two singles in a row or two streaks in a row

2= three singles in a row or three streaks in a row

3= four or more singles in a row or four or more streaks in a row

At the same time the single-streak or streak-single situation getting a 0 value at the above classification will get different values in relationship of how many back-to-back spots happened:

one 0 = just one single-streak or streak.single situation followed by a same single or streak cluster.

two 0= two single-streak-single or streak-single-streak situations followed by a same single/streak cluster

three 0=three or more single-streak-single-streak or streak-single-streak-single situations followed b a same single/streak cluster.

Example:
the shoe is:

P
B
PP
B
P
B
P
BBB
P
BBBB
P
BB
P
BBBB
PP
B
PP
B
P
BB
P
BB
PP
BBBBBB
PPP
B
P
B
P
B
PPP
B
PP
BBBB
PP
B
P
B
P
BB

Now in numbers the shoe looks like as:

1-0-3-0-0-0-0-0-0-1-0-0-1-0-0-3-3-0-2-3

Since we cannot have numbers different from 0,1,2 or 3 we could have a better picture of what the shoe is really producing.

In this shoe we got a higher than expected single-streak or vice versa spots ('0' values), anyway notice how short went '2' situations and cumulatively how many 0 and 1 situations happened versus superior numbers.
Moroever note the final portion of the shoe formed by a kind of 'concentrated' numbers different from zero.

Anyway we're restricting the field of operation by two tools: any streak is a streak no matter what, and the number 3 incorporates the more likely events after setting up a cutoff at four.

as.

Thanks Happy Christmas and a Wonderful New Year!!!

as.

Btw, there's a reason why a 'clustered' plan of action (at either W or L way) is more suitable than an 'isolated' one.
Whenever we're betting toward a given flow, odds that this flow will stop are inferior than odds that the flow will prolong.

For example, we'll surely face a shoe showing up 7 or 8 or more consecutive doubles (singles ignored) than a shoe producing 7 or 8 or more isolated double spots (that is gapped each time by a 3+ streak).

Anytime you won't get a feeling about what side to bet, consider Player C situation.
He is favorite to get more long losing spots than long winning spots as most of the times the BP texture won't get a WL hopping situation for long or not proportionally placed.

That means that in the majority of the times you are either destined to consecutively win or consecutively lose, giving the best of your fk about math percentages of being WL or LW balanced.

as. 

To better understand how things work at baccarat let's say we set up three different players wagering (fictionally) for us: we simply register their results in term of Ws and Ls distribution. Shoe per shoe.

Player A will bet toward getting positive clusters (trend following, long streaks or long chops, my plans, AD, Banker wagering in possible profitable situations, etc)

Player B thinks player A is an id.iot and takes the right opposite route, that is betting that valuable patterns do not come out or that will break soon or toward short streaks, weak B distribution, my plans are sh.i.t, I like doubles, etc)

Player C (the passive one) takes a middle route by alternatively wagering that player A or B will respectively get very short winning or losing sequences set up at 1 level (chopping level). That is he bets that A and B will get more long WL chopping sequences than a proprotional amount of losing W or L streaks.

Obviously it's virtually impossible that at the end of each shoe all three players will be losers as a perfect balancement world is out of order (see later).
Maybe the negative EV will get a role on that, but it's not the main cause why casinos take a lot of money from bac players.

Nonetheless, the general probability to get player C as final winner is diminished as a perfect hopping mood between players A and B  is slight less likely to happen, otherwise a simple multilayered martingale applied to him would get easy and endless series of winnings (giving a fk about normal variance).
Anyway, even player C is entitled to get his share of wins and, guess what, they must come out clustered at some point. And naturally the losing counterpart (forming Pl A or Pl B clusters) must come out clustered too.

In this way no one single shoe is unplayable as something profitably clustered MUST come out at various degrees of quantity and quality (distribution).
It's important to understand that a 'cluster' is just a back to back scenario so WW or LL belongs to the same cluster category as a WWWWWWW or LLLLL situation.
And by the same way of thought any xWL or xLW spot is an isolated (chop) mood.

Fortunately for us at baccarat clustered events of different nature will prevail over isolated spots.
Tomorrow I'll provide some examples about it. 

as.

Shortly  I'll provide some numbers, without them we're not going anywhere.

as.

Not paying the vig represents a big advantage for us, such shoes are almost symmetrical with their B counterpart.

as.

Good points Indeed.


as.

Thanks klw!!!
Yes...a lot of stuff.....

Take care!

as.

Look at this shoe (1=singles, 2=doubles and 3=3+ streaks)

br: 2,1,1,2,1,1,3,3,1,2,3,1,3,3,2,3,2,3,2,2,2,1,1,3,1,2,2,2,2,2,3,1,1,3,2,1,1

ubp #1: W,W,W,W,L,W,W,L,L,W,W,L,W,L,W,L,W,W,W,W,L,L,W.  w=15, (+15) AND L=8 (-24)

byb: 1,3,1,1,1,1,3,3,1,3,3,1,2,1,1,1,1,1,1,1,1,1,1,2,1,1,3,1,1,1,1,3,3,3,1,1,2,1,1,2

ubp #1: w,w,w,w,w,w,w,w,w,w,l,w,w,w,w,w,w,w,w,w,w,w,w,w,l,w,w,w,w,w,w,w,w,w,l,w,w,w.

w= 35 (+35) l= 3 (-9)

sr: 3,1,2,1,2,1,2,1,2,1,2,2,1,1,1,1,2,2,3,1,3,1,2,3,1,1,2,3,3,3,2,3

ubp #1: l,w,w,w,w,w,w,w,w,w,w,w,w,w,w,w,l,w,w,l,l,w,l,l.   w=18 (+18) l= 6 (-18)

cr: 3,3,1,1,2,1,3,1,3,1,3,2,3,1,1,1,1,2,1,1,3,3,1,2,1,2,1,3,3,3,2,3,1

ubp #1: w,l,w,l,w,w,w,w,l,w,w,w,l,w,w,l,w,w,l,w,w,w,l,w,w,w,l.  w=19 (+19)  l=8 (-24)

Say we are a team formed by 4 persons (each betting ubp#1 at one of the four roads).
For simplicity we ignore the vig.

At the end of the shoe:

br player loses 9 units = -9

byb player wins 26 units = +26

sr player breaks even = 0

cr player loses 5 units = -5

In total our team won 26-14 units, that is a +12 profit (before vig).

There are one million of post hoc considerations to be made, say we're just focusing about the W isolated spots (IS) and W clusters (CL).

br:  CL, CL, CL, IS, IS, CL

byb: CL, CL, CL, CL

sr: CL, CL, IS

cr: IS, IS, CL, CL, CL, CL, CL, CL.

In total 16 W Clusters and 5 Isolated Ws.

Now the reverse situation about losses, that is the number of L isolated spots and L clusters.

br: IS, CL, IS, IS, IS, CL

byb: IS, IS, IS

sr: IS, IS, CL, CL

cr: IS, IS, IS, IS, IS, IS, IS.

In total 16 Isolated L spots and 4 L clusters.

If our team would like to bet toward W clusters and L isolated spots we'll get 16+16 winning spots (+32) and (5+4) losing spots (-27). That is a +5 unit profit (minus vig)

Notice that the original 'full betting' plan produced two losing players (br and cr players), one breaking even (sr player) and one strong winner (byb player).

The clustered/isolated sub plan formed br player losing 4 units, byb player winning 7 units, sr player losing 5 units and cr player winning 7 units. That is a cumulative +5 unit profit.

Globally the player who contributed most to our team was byb player, cr player won slightly, sr lost 5 units and br got a tremendous -13 unit loss.

Let's take another shoe.

br: 1,3,2,3,1,1,2,1,3,1,1,3,2,1,1,1,1,1,3,2,1,2,1,2,1,3,1,1,1,3,1,3

ubp#1: L,W,L,W,L,W,W,W,L,W,W,W,W,L,L,W,W,W,W,L,W,W,W,W,W,W   W=19 (+19)  L=7 (-21)

byb: 1,3,1,1,1,1,3,1,3,3,2,1,1,1,1,1,2,3,1,3,1,1,3,3,2,2,1,2,3

ubp#1: w,w,w,w,w,w,w,w,l,w,w,w,w,w,l,w,w,w,w,w,l,w,l.   w=19  (+19)  l=4 (-12) 

sr: 2,2,3,2,1,3,1,1,2,1,3,3,1,3,3,3,1,3,1,1,1,1,1,2,1,3.

ubp#1: l,w,w,l,w,l,w,w,w,w,w,w,w,w,w,w,w,l,w,l  w=15 (+15)  l=4 (-12)

cr: 1,1,3,1,1,1,1,3,2,3,3,1,2,2,1,3,3,2,3,3,2,1,2,2,1,1,1

ubp#1: w,w,w,w,w,l,l,w,w,l,w,l,w,w,w,w,w.   w=13  (+13)  l=4 (4 (-12)

So let's see if our team survived this shoe and who contributed most to save the joint:

br player kept losing (we should discard him/her from the team as being too unlucky  ) : -2 units

byb player won 7 units (+7)

sr player won 3 units (+3)

cr player won 1 unit (+1)

Overall our team won 9 units (before vig).

Let's see again the sub IS/CL plan, first in W then in L spots (reversely taken of course).

Wins:

br player: IS, IS, CL, CL, CL, CL

byb player: Cl, CL, CL, IS

sr player: CL, IS, CL, IS

cr player: CL, CL, IS, CL

Losses:

br player: IS, IS, IS, IS, CL, IS

byb player: IS, IS, IS

sr player: IS, IS, IS, IS

cr player: CL, IS, IS

Globally we got 26 wins (12+14) and 8 losses (6+2) that is a +26 - 24 = +2 unit profit (before vig)

In this second shoe original plan made three winners and just one loser whereas the sub plan made two breaking even players, one winning 3 units and one losing 1 unit.

Some comments.

It's evident that in those two shoes the first (original) plan got us more gross profits than the sub plan (+21 vs +7).
Quite interesting is the fact that sub plan got 1/3 of the original plan profits.

Anyway we shouldn't forget the vig burden, especially if we have to play a kind of mini progression.

Moreover variance is always around the corner and it's not that easy to set up a plan like this.
I mean that many situations might easily dictate to bet opposite sides to get a given searched result and the trick, already investigated, to wait for multiple roads to converge toward an expected spot has shown to be of no value.
Naturally 'conflicting' bettable sides are less likely to happen when adopting the sub plan proposed here.

Another possible trick that may come to mind is to select the player seemingly performing best, abandoning (at least for some betting spots) the losing or 'limping' players.
But this is an equivalent move as adopting a simple trend following strategy: we do not know when to enter and when to quit, maybe jumping from one player to another without a statistical reason to do that.

Finally, yes, some shoes are unplayable as they do not give the proper room to get a fair amount of pattern situations. Among the worst shoes to play our strategies at there are those producing a lot of ties, so when we suspect the actual shoe will be 'tie' rich we'll simply stand up or starting to bet them. 

as.

Thanks klw! I appreciate your words and the time you spend here reading my section!

Imo at baccarat there's no such a thing as quitting when ahead or behind, setting up a goal (in either W or L way) or stuff like that.
Our plan must get an edge in the long term by assessing the game features both in theory and, more importantly, in practice.

I can't give a lesser damn whether we're (temporarily) behind after playing 3 or 4 shoes, such thing happens, albeit rarely (it doesn't happen to those forums' geniuses claiming that they win multiple units at every shoe dealt).

Each W/L pattern will come out with the same probability, so we should focus our attention about their distribution shoe per shoe. Not every shoe is featuring exploitable situations, as everything depends about how symmetrically or asymmetrically are placed the cards at every shoe dealt.
It's true that many results come from 'whimsical' spots, but those will show up with the same profitable or unprofitable probability: after all what it's math advantaged remains advantaged.

Best example to make is by assessing the 1-2 (or 2-1) and 1-3 (or 3-1) BP flows coming out at a given shoe (unb plan #1).
There are strong reasons why we had discarded 2-3 spots from our play as they act accordingly to the statistical tools already touched here.

Most of the times such situations tend to come out clustered in a way or another and this is not mainly caused by the 0.75 probability to happen.
Not given precise values are going to get a steady advantage (unless our progression is utilized) as this edge depends about how symmetrically or asymmetrically are placed the cards on the actual shoe we're playing at.
In addition and even though we have assigned them the same 3+ value for simplicity, not all 3+ streaks are equal either in quantity and in quality.

as.

Here the progression scheme

as. 

In essence and no matter the strategy employed, per each shoe dealt we'll get a number of pseudo equlibrium spots and a number of deviations going toward the natural RMS value.
Obviously it's highly preferable to play toward the latter scenario not because we have in mind a kind of 'sky's the limit' attitude but simply as baccarat is a word of multiple asymmetrical propositions.

Adopting a mini positive (multilayered) progression will help us to define if our bets are falling into the 'pseudo equilibrium' territory or a more hopeful mild, moderate or strong deviation.

Naturally if WW is the searched scenario and LL is not that terrible (proportionally speaking), our enemies are isolated Ws and, at a different extent, isolated Ls.

Example.
The succession is WLLLWLLLLWLLLLLLLLWLLL (W/L ratio 4/18)
The RMS value is abundantly passed (3.7-sigma), but we were betting the opposite situations.
Our progression (assuming to stay at the same first level) applied to this sequence is:
+1, -2, -1, -1, +1, -2, -1, -1, -1, +1, -2, -1, -1, -1, -1, -1, -1, -1, +1, -2, -1, -1. total = -18 units (plus vig).

The specular natural counterpart is LWWWLWWWWLWWWWWWWWLWWW (W/L ratio 18/4)
Now we'll get: -1, +1, +2, +1, -2, +1, +2, +1, +2, -1, +1, +2, +1, +2, +1, +2, +1, +2, -1, +1, +2, +1.
total = + 21 units (minus vig)

It's evident that this 22-hand succession from a statistical point of view is perfectly balanced but from an economic point of view is asymmetrical to our favor.

Someone could argue that in the first heavy losing sequence we might wait for a W to come out before betting without immolating money on those long L series (so losing less than 18 units). And the same about the second sequence by betting more than a mechanical 1-2 progression on that strong winning succession (so winning more than 21 units).
But this kind of reasonment is failed by an obvious post hoc thought.
Winning and losing both belong to the same process, we know sooner or later things will take a more natural flow but we do not know when (partial imperfect information) and how much (true imperfect information).

Of course a steady pseudo equilibrium status will more likely elicit a deviation of some kind, after all the original succession could be splitted into infinite ways where the equilibrium is the exception and the deviation is the rule.

Progression steps

Our initial betting unit is 1 (there are reasons to set up this initial bet at 5, we'll talk about this issue in the future).
We'll always parlay a first win just one time forever and ever, no matter whether we win the first step (WL) or the second step (WW) our bet will be 1 until a deficit of 8 units will be reached.
If we are in the positive territory we'll stop the betting until the RMS value is reached, waiting if the RMS value is erased then starting to bet again.
We'll bet 1 unit whenever our losing amount is included into the 1-8 units range.

When we are behind of 8 units we'll pass to the 2nd betting level made of four stages, that is betting 2 units with the same parlay rules.
We stay at this 2 level if our deficit is reduced to 7 units, now we take back the first 1 unit betting level step.

If our deficit raises to 16 units, we'll start to bet 3 units made of four steps, again with the same parlay rules.
If we are able to come back at a 14 unit deficit we'll go back to the second step (that is betting a 2 unit); if we're unfortunate to further increase our loss (28 unit loss) we'll step forward the 4-unit level (again made of four steps) and so on by increasing one unit for four times long up to the final 10-unit betting level. Up to the point where we'll get one unit profit.

Step one and two follow the same rules, but next step levels (3,4,5,6,7,8,9 and 10) must stay at the same level after a parlay win.
After two consecutive parlay wins made at levels 3-10, we'll go back to the immediate inferior betting level; for example after having won two times a parlay bet of 4 (4+8 and 4+8 = 24 units) we'll step back to 3 level.

I'll provide a scheme about what to do in relation of the actual deficit, but it's quite simple to understand that to go back to the inferior betting level we need to end up as winners at a given betting level.

Technically this progression 'challenges' the results not to give at least a RMS value for long and, frankly, this situation it's almost impossible to happen even if we're voluntarily playing to lose.

But let's falsify such hypothesis by setting up a strategy oriented to get more pseudo equilibrium spots than deviated spots those ones reaching at some time a normal natural deviation. The progression utilized remains the same.
In other words we're hoping that WL and LW spots are more likely to happen than WW or LL spots.
A paraphrase of D'Alambert progression that was proven to be a sure recipe for disaster.

Instead of 4-hand propositions let's take just 3-hand ways to consider outcomes:

WWW = +1, +2, +1 (+4)
WWL = +1, +2, -1  (+2)
WLW = +1, -2, +1  (0)
WLL = +1, -2, -1  (-2)

LLL = -1, -1, -1  (-3)
LLW = -1, -1, +1 (-1)
LWL = -1, +1, -2  (-2)
LWW = -1, +1, +2 (+2)

In eight 3-hand attempts, odds to break even are 1:8.
In eight 3-hand attempts, odds dictate that the maximum win (1: is +4 and the maximum loss (1: is -3.
In eight 3-hand attempts, half of the time we'll get either 2 units won or 2 units loss.

Just in one scenario we'll be behind of 1 unit.

Therefore we know that after three hands played, our bankroll movement is 6:2 favorite to get either an increase or a decrease superior than 1.
Not a good thing to know when applying a kind of constant 'balancement' play.

as.

Thanks a lot for your replies Al and KFB!!

Back to the progression

Say we are splitting our strategic plan into four infinite WL situations by adopting the 1-2 positive limited progression (anytime we win we'll leave the original bet plus the bet won; any bet lost at either two-step stage remains at the same level without any betting increase). For simplicity we do not take into account the vig.
Further progression rules will be posted later.

No matter how concentrated or diluted our wagers are, 4 hands bet can produce just 16 WL situations:

1) WWWW (+1, +2, +1, +2) = +6

2) WWWL (+1, +2, +1, -2) = +2

3) WWLW (+1, +2, -1, +1) = +3

4) WWLL (+1, +2, -1, -1) = +1

5) WLLL (+1, -2, -1, -1) = -3

6) WLLW (+1, -2, -1, +1) = -1

7) WLWW (+1, -2, +1, +2) = +2

WLWL (+1, -2, +1, -2) = -2

Specular 'losing' counterparts are:

9) LLLL (-1, -1, -1, -1) = -4

10) LLLW (-1, -1, -1, +1) = -2

11) LLWL (-1, -1, +1, -2) = -3

12) LLWW (-1, -1, +1, +2) = +1

13) LWWW (-1, +1, +2, +1) = +3

14) LWWL (-1, +1, +2, -1) = +1

15) LWLL (-1, +1, -2, -1) = -3

16) LWLW (-1, +1, -2, +1) = -1

Of course the total final amount of hands #1-#8 is +8 and at hands #9-#16 is -8.
For that matter 1-8 hands range provides 5/3 W/L cumulative ratio and the same is oppositely true about 9-16 hands range (5/3 L/W cumulative ratio).

Without a possible bet selection advantage, the probability to get each of those different outcomes is fkng symmetrical but there are some differences even in this perfect balanced scenario:

a) any pattern starting with a W will get an average final total amount of +14 at W spots and an average final total amount of -6 at L spots (+14 - 6 = +8)
On the other end, any pattern starting with a L will get an average final total amount of -13 at L spots and an average final total amount of +5 at W spots (-13 +5 = -8).

b) whenever a specular extreme pattern as WWWW or LLLL comes out (getting the same probability to appear), we'll get different final results, in fact:

WWWW = +6
LLLL = -4

the same about the back to back appearance of those 'extreme' results, so:

WWWW/WWWW = +12

LLLL/LLLL = -8

and so on.

In other words, whenever our first bet of our 4-step wagering 'unlimited' strategy is a W, we'll get 5/3 odds to end up with a final win ranging from +6 to +1 and a final loss ranging from -1 to -3.

Conversely, whenever our first bet is a L, we'll get 5/3 odds to conclude the 4-step wagering with a cumulative loss ranging from -4 to -1 and a final win from +1 to +3.

If we'd symmetrically put the same W/L opposite eight scenarios together, we'll see that cumulatively we are going to get just two situations having a +2 global result, one getting a -1 result and another one getting a -3 result. The remaining four situations are going to get a 'break even' spot.

Hence itlr we know we must face half of the outcomes to be W/L 'balanced' and the rest to be deviated in a way or another.

Back to the WL scenarios getting a transitory strong 'balancement' (temporarily denying a root mean square normal value):

WLLW = -1

WLWL = -2

LWWL = +1

LWLW = -1

WWLL = +1

LLWW = +1

Cumulatively those 'perfect balanced' W/L spots are going to get us a -1 global result perfoming a 6/16 probability to happen (p=0.375).

That means that the remaining part of results (p=0.625) are affected by a kind of normal 'bias' toward either side of the bell curve.
Naturally a perfect independent proposition cannot get us any valuable hint to bet this or that, therefore these assumptions cannot apply at roulette, for example.

But since at baccarat it's highly improbable (say impossible) to arrange key cards proportionally along any shoe dealt and by different paces both being insensitive of the place selection and probability after effects statistical tools (those ones confirming or not the real random nature of the shoe), we know that sooner or later the root mean square value will get its rights to happen.

For that matter I do not know a single long term winning player capable to quit a session as a winner unless he/she is able to catch the spots where W is more followed by another W than by trying to stop a L spot.

Later about the progression topic.

as.

Natural deviations at any shoe dealt

If you toss a fair coin for long, sooner or later you'll encounter 20 or more heads or tails in a row and/or every other strong predominance following the expected sd values.
For example a streak of 20 at either side will come out, on average, two times over more than 1 million trials.
It's a normal happening everyone knows here.

The same could happen at baccarat, long streaks happen of course more probably at B side.

What it's interesting is that all intermediate situations are taking a univocal expected distribution at a coin flip proposition but slight different lines at baccarat.

Due to the asymmetrical card distribution typical of baccarat shoes, you can take for grant that most part of our bets will be either hugely right or hugely wrong as two-card initial situations will mathematically dictate the final result by a very strong edge.
A thing completely different than a coin flip proposition where each bet made will get a symmetrical probability to be right or wrong.

No matter the strategy employed, this feature endorses the probability that things will tend to come out wrong or right more in clusters than in isolated fashions. Shoe per shoe.

It's a sure fact that casinos will make tons of money from baccarat tables as players cannot realize how many bets are placed into the 'negative territory' (almost always by improperly increasing the wagers) and how many bets are placed into the 'positive territory' by low or too low amounts.

The math edge is just a 'booster' for casinos and not the main cause for they are collecting millions from baccarat.

From a technical point of view, we can summarize and simplifying the issue by expecting more W clusters and L clusters than long WL or LW spots as the root mean square deviation must happen shortly (especially when strong WL and/or LW states seem to negate this natural appearance).

Of course taking B and P as the main registration to base our strategy upon and under normal circumstances is among the silliest things to do that casinos particularly like.
'Smart strategies' as betting B after PPP or PPPP are just a waste of time and an insult to common intelligence that were tested and demonstrated to be as totally worthless.
(Those who are disputing this assumption are invited to confirm their hypothesis by betting B after three or four random hands are dealt regardless of the BP outcome, and results will get the same negative EV).

Our progression, part 1

There are no other ways around to confirm that baccarat is beatable by adopting a strategy different than a flat betting scheme. Maybe a well calibrated progression could dilute the risk of ruin but sooner or later this ruin will happen.
If not, it means that along with a careful MM, a strong valid bet selection is in order.
And if this is true, why not concentrate our betting by wagering very few spots supposedly getting a kind of EV+ by large sums?

Back to the progression.

Bankroll is formed by 224 bets splitted into eight and four multilayered betting levels.
Minimum bet is 1 unit and highest bet is 10 units.

So we can safely join a Vegas HS table by setting up a $2000 minimum bet, knowing at the worst scenario we'll wager $20.000. Bankroll is $448.000.

Actually it's very very rare to wager more than 4x standard units, meaning we'll get yet 24 bets behind (see later).

The progression is able to endure a 43 negative WL gap, meaning we need to meet a 44 L/W gap to lose the entire bankroll.

Naturally vig will erode winning spot amounts, but progressive bets stay low no matter what.

The progression is so solid that most peers stressed us to accelerate it but getting a firm 'no' from our part as we know that s.h.it could happen.

Progression basis concept

If things tend to come out more clustered than isolated and taking into account the root mean square value, we better bet progressively toward winning clusters than isolated clusters at the same time fearing most losing clusters that represent the same opposite side of the medal.
If our effort is restricted to the least when losing and slightly enlarged when winning, we should get an edge as the W and L situations are specular itlr (before vig).

Therefore our scheme is splitted into 10 different bet unit levels (totaling 44 betting spots) by wagering a two-step positive progression. That is after a win we let the entire amount to run for one time more. 

First eight levels stay at 1 unit betting level until a higher than 8 unit loss come out, now we'll step forward a 2 unit bet.

From the 2 unit bet, each level will stay four times instead of eight and so on.

The amount of the actual bet is dictated by the total number of units lost:

from 1 to 8 unit lost = bet 1 unit

from 10 to 16 unit lost = bet  2 units

from 19 to 28 unit lost = bet 3 units

from 32 to 44 unit lost = bet 4 units

from 40 to 64 unit lost = bet 5 units

from 70 to 88 unit lost = bet 6 units

from 95 to 116 unit lost = bet 7 units

from 124 to 148 unit lost = bet 8 units

from 157 to 184 unit lost = bet 9 units

from 194 to 214 unit lost = bet 10 units

There are additional 'rules' we'll see in the second part, the important thing to see is that each every losing bet (regardless of the first or second step result) remains at the same level (and set up by the current total losing unit amount) whereas every winning bet must be always 'parlayed' one time even if we're in the positive or neutral field.

So if we are working at the 1 unit level we'll get:

LL  = -1, -1 = -2
LW = -1, +1 = 0 (before vig)
WL = +1, -2 = -1 (before vig)
WW = +1, +2 = +3 (before vig)

Itlr such totals will be equal so getting us a loss for the vig, but the slow progression betting will make more and more probable the easiest winning situations we are really looking for, that is any WW sequence that tend to neutralize (many times slowly) many previous losing spots.

Acting in this way we are deadly sure to play a perfect symmetrical winning/losing proposition not giving a fk about the urge to break even shortly and at the same time capable to get the best of it when things come in our favor for long (think about a WWWWWWWW sequence getting us a +12 unit profit before vig)

After years of reading gambling forums, you should not find yourselves in the position to be behind 44 bets, it's a kind of impossible task.

as.

Hi KFB!!

That in my opinion is one of the main differences between Bac and many comparable games  that instead utilize dice vs cards(primarily games such as craps). Yet gaming authors and casino employees incessantly blather on "its just a fitty-fitty coin flip game", suggesting that it doesn't matter how we wager/ "no advantage to be found here".


Exactly, I agree too, such people keep thinking the game as a coin flip or 0.5068/0.4932 propositions not considering at all the source of the results taken distinctively

Of course itlr distinct ditributions will converge toward the math expected values, but a precise card distribution will get way more restricted values toward one 'status' (pseudo equilibrium) or the other one (light, moderate or strong deviation) and the final math assessment must be calculated by the root mean square.

Thus if we're going to apply a kind of 'balancement' strategy we must hope that the actual shoe is full of pseudo equilibrium spots, and not that at some point of the shoe things will change (toward deviations of some entity). That is the shoe we're playing at hopefully must contain root mean square values lower than expected.

Conversely, a kind of 'deviation' strategy must hope that the root mean square tool will get its expected value at different degrees and we just need a number different to zero to get a sort of advantage.

Again let's consider how many times a given A or B (or W and L) opposing scenario will get long 'hopping' sequences vs homogeneous sequences and we know that the latter case is well more likely.

Raise the general probability to succeed and we see that such 'hopping' balanced scenario will be more and more unlikely.
After all, the root mean square value (and values acting toward it) must happen at some point for three diifferent reasons:

- it's a normal situation happening even at pure independent propositions;

- the game is produced by a sure asymmetrical card distribution;

- the A/B or whatever taken x/y propositions are asymmetrical by definition.

In some sense any single fkng shoe dealt is affected by a number of 'pseudo equilibrium' spots and 'deviation' spots where the latter tool tends to be more prominent (or not quite balanced by the opposite situation) than the other one.

Say that cards are voluntarily arranged toward a strong balanced status, meaning that 'trend following players' (constituting the vast majority of bac bettors) will get no hint to bet this or that.

Really?

Card distributions making strong balanced distributions (that is not reaching the mean square value) for long cannot happen at all four common registrations (BR, byb, sr and cr), let alone about other random walk registrations.

The number of -1/0, 0/-1 and +1/0 and 0/+1 spots cannot last for long as they MUST jump to -1/-2 and +1/+2 spots. And so on.
If the 0 value would be reached several times in short sequences, baccarat wouldn't exist.
And btw it would be a strong falsification of the sure fact (not hypothesis) that cards are asymmetrically distributed per each shoe.

Whenever a shoe is forming at a given subsuccession a higher than expected number of 'pseudo equilibrium' spots (huge number of very low deviations going around a 0 point), odds are that subsequent shoes will approach more and more the root mean square (RMS) value, meaning that the same succession will get more deviations than 'pseudo equilibrium' spots, and it's now that we'll get an edge over the house.

Globally, our edge is estimated to be 3% over the house AFTER VIG, a huge edge I mean.

This edge comes from a simple RMS natural happening:

- from 7 to 12 hands played the RMS is 3
- from 13 to 20 hands played the RMS is 4
- from 21 to 30 hands played the RMS is 5
- from 31 to 42 hands played the RMS is 6
- from 43 to 56 hand played the RMS is 7

and so on...

Even though a strict flat betting scheme will get the best of it, we can easily set up a kind of 'bringing down the house' strategy by adopting a progression that cannot be wrong by any means as the probability to lose the entire bankroll is almost not existent.

We'll see it in a couple of days.

as.