I'm strongly convinced that playing toward an 'average' card distribution (or not) will get the job, yet I'm way more convinced that shoes rich of ties are unplayable (at least according to my methods).

We've made the same adjustment KFB was talking about: looking at the side (or patterns) happening most after a tie.
Unfortunately this approach didn't get us any advantage over the house (no hurt either).

Probably many ties showing up work as a kind of multiple 'resetting' patterns, so we need some more hands to detect the most likely flow of the outcomes.
But if ties gaps are short, patterns are shorter and more 'whimsical'.

Baccarat is a game of numbers, ties are not numbers. So we can occasionally afford them up to a point.

as.
 

HI ASB

U say:
"...
it's quite unlikely to get a back to back 6-card hand, then it's even more unlikely to get a two cluster of 6-card hands. And so on.
Itlr 6-card hands are way more likely to come out as 'singled' patterns.

and

In addition, 6-card hands deny the Banker advantage unless the third card is a 6 or a 7 and B has a 6 two-card point.

Shoes rich of 6-card hands are more difficult to be detected as key cards distribution do not make their more likely job..."
-------------------------------------------------------------------------
I agree with first part re unlikely scenario for bk2bk 6-card totals. So, if we follow that logic post-6card hand that now there is a greater probability for a 4-card or 5-card hand showing next (hopefully for B bettor it's the latter) ---

Q: Is there anything within that most recent 6-card hand that would suggest which of the two hand totals(4 or 5) is more likely as a f(x) of that 6-card hands' makeup. I believe 5-card hand is slightly more likely itlr (can't recall the exact % but believe its ~~31%.)

(Q1 is there an indicator suggesting a 4-card is more likely to show vs a 5-card hand next.)?
(Q2: Does your logic above also suggest that a 4-card hand is similarly less probable to show Bk-to-Bk? What about 5-card hand ??)

Thx in advance ASB or anyone else,

Hi KFB!

Q: Is there anything within that most recent 6-card hand that would suggest which of the two hand totals(4 or 5) is more likely as a f(x) of that 6-card hands' makeup. I believe 5-card hand is slightly more likely itlr (can't recall the exact % but believe its ~~31%.)


6-card hands probability is 31.6%.

5-card hands probability is 30.5%

4-card hands probability is the remaining 37.8%. (ok, the total is 99.9% but it wasn't me to make the calculations) :-)

(Q1 is there an indicator suggesting a 4-card is more likely to show vs a 5-card hand next.)?


Well, there's a 7.3% general math propensity toward 4-card hands than 5-card hands formation. Obviously the main factor orienting 4-card hands is the naturals apparition (34.2% vs 3.6%).
And again more obvious is the fact that shoes rich of 8s and 9s make more probable this possibility.

(Q2: Does your logic above also suggest that a 4-card hand is similarly less probable to show Bk-to-Bk? What about 5-card hand ??)
 

Naturally it's way more likely to get a back to back 37.8% math probability (4-card hands) than a 30.5% math probability (5-card hands), anyway both fight against a greater world.

Those are just 'general' math values that must be filtered with the actual shoe conditions considered by patterns and key cards concentration/dilution.

Thus itlr even 4, 5 or 6 cards successions impact the patterns formation. We do not know which side will win but we could assess the 'ranges' of intervention.

More later

as.

A succession of events will be really random whenever we can assign to it the attribute of a 'collective' (RVM), so in our example about cards employed to form a hand any 4, 5 or 6 succession happening at every shoe dealt must produce 'random' so unbeatable sequences.

4-card situations are surely symmetrical, a fair portion of 6-card situations are surely symmetrical. But 5-card events are strongly asymmetrical by definition as either for the situation or for the bac rules, one side is heavily favored to win the hand.

Now, a finite and slight dependent actual card distribution cannot be considered as an endless random production, meaning there will be spots endorsing asymmetrical spots by a value different than an expected number. In a word, 5-card situations will privilege one side or at least more likely 'ranges' of apparition. Therefore bac shoes are not a collective, then beatable.

ITLR and you can bet everything you get on your name, it's the number of 5-card situations guessed right or not that make you winners or losers, shoe per shoe.

In fact, 4-card and a huge portion of 6-card situations are symmetrical, so unbeatable by definition.
Sometimes you'll guess right and other times you'll be wrong. At the end the sum will be 0 (before vig).

Say you know for sure that the next hand will be a 5-card hand, the only situation to be hugely right or hugely wrong.
Obviously you'll be inclined to bet B as there are more winning 5-card hands at B side than P side.
True, yet a fair amount of 5-card hands go toward Player side, think about standing or natural P points or asymmetrical spots where third cards help the P with Banker standing.

Do not forget that B winning hands are payed 0.95:1 and that a fair amount of 5-card hands (nearly 40%) are strongly favorite (at various degrees) to win the P side at the start.

It's like we're playing a game where the most part of results (4-card and 6-card hands) are belonging to an undetectable world and the remaining portion (5-card hands) is splitted between B and P.
And, imo, it's just the attitude to 'guess' which side will be more kissed in such 5-card hands occurences that itlr will make us winners or losers.
The major hint to look for is, again, the 'clustering effect' as more propositions are considered at the same production, lesser is the probability to get 'equaling' results up to some cutoff points.

as.

Clusters

Say A is a betting approach toward clusters and B the 'anti clusters' counterpart.

In no way at a random production A>B, let alone B>A. In fact even an excess of anti clusters B constitutes a cluster of some kind.

Things tend to differ when we have reasons to think that the production is not really random.
Now clusters are more likely to happen, but again they could manifest by 'excesses' of A or B taken at different portions of the shoe.

The dilemma is deciding when to take the A or B route and how long.

Of course there are several 'clusters' to look for, not necessarily considered by common BP (or r/b) patterns

Example.

Whenever we bet Banker we hope Player will draw first, as B is generally advantaged no matter what.
Obviously when Player is standing, B side is underdog to win at various levels.

It could happen that a P drawing will win many hands in a row and, conversely, that a P standing will lose some hands in a row, but the rule is that situations when P draws make B advantaged and when P stands makes B underdog.

Anyway, situations when P draws may be classified under different classes (none, singled, two in a row, etc) and such feature is way unbalanced along any shoe we're playing at.
The same about P standing points (nearly 40% of total hands), now with the important factor that we'll expect a greater number of none or singled situations than clustered situations at different levels.

In reality the actual card distribution tends to deny 'expected' values up to some cutoff points.
For example, in ten hands considered, the P standing/P drawing ratio will very rarely reach the 4/6 value.

Thus we may infer that P drawing and P standing situations more likely move around 'clusters'.

That doesn't mean that guessing the P drawing or standing nature it's a condition to win, but it's a good start.
Let's name it as a 'first clustering factor' happening along any shoe dealt.

Next step (twilight zone) is to assess how many times a so called unfavorite side will win (we do not care about the times when an advantaged side will get the best of it as this is the common course of action).
Now we should assess how many third cards will make P side to win (and vice versa) and how many standing points will succumb to opposite better standing points.

We'll see that later.

as.

Once you have determined the actual P drawing or P standing gaps (in relationship of their expected probability), well, sky's the limit in the sense you'll crush every live bac table in the world.
Providing to assign a proper value to the 'twilight zone', that is the events affecting the math oriented situations for actual card distributions tending to surpass given cutoff points (so 'gaps').

Most of the times the twilight zone is relatively insensitive to actual occurences deviating from the norm, yet they could get you a harsh damage at your bankroll especially when you like to place a lot of bets.

Example.

For whatever reason, you think P side will be more probable than B side. In math terms that means P side will get a standing point by a proportion greater than 40%.
You bet Player getting a standing 7. Nice job so far.
Unfortunately Banker shows a natural and you lose. Even worse is when B side has any point different than 6,7,8 and 9 and will catch as third card a card surpassing your 7.

Now, how many times such instance could happen?
Maybe once, maybe two. After this 'cutoff point' we're not interested to chase a math more likely situation as the actual card distribution put a strong stop on it.
The reason is because the number of standing 7s at P side is limited and the shoe is a finite and dependent world.

It's the same reason working at asymmetrical spots when the third card instructs the Banker to stand while Player has a winning hand.

Maybe in the future Banker will win by standing points but the asymmetrical spots are somewhat consumed as they are limited in their appearance.
With the decisive difference that P standing points get a 40% probability to happen whereas asymmetrical spots have a 8.6% probability to happen, that is 4.65 times more likely to show up.

I've been repeating this important concept many times in my pages:

To win itlr you must take the math advantaged side giving a fk about transitory results, those are there just to illude recreational players (99.9% of bac players, maybe more than that).
If you'd bet P side and P side is drawing, you know your bet is more likely to lose than win, no matter the final fkng result.
Conversely if your bet is placed at B side and Player is standing, you are losing money. 

Say that you'll bet Banker only after a single P standing (Ps) point will happen.
Since Ps < Pd (P drawing spots), you might conclude to get a kind of edge as Ps + Ps < Ps + Pd.
Unluckily, some card distributions make Ps + Ps > Ps + Pd.
So we move to the further step.

That is: Ps + Ps + Ps < Ps + Ps + Pd.

Now the likelihood to be wrong (that is to get a third Ps) is way more limited but it still happen.

No need to look further, our Ps expected propensity went wrong at two consecutive levels, no matter how were the actual results.

The reason why we stop the Pd propensity after two steps is because shoe is limited and dependent, of course we might prolong the Pd propensity up to three consecutive steps but it takes too much time to look at those occurences.

Nonetheless, Ps spots are way more likely to show up as singled or two-paced, even if real outcomes are hurting us.

Try to bet Banker after any Ps spot shows up, then after a couple of Ps spots show up.
Obviously you can't be more wrong than the expected EV-.

as.

Many thx Asymbacguy for your elaborate answers to my Q up above.

Many gold nuggets:

Well, there's a 7.3% general math propensity toward 4-card hands than 5-card hands formation. Obviously the main factor orienting 4-card hands is the naturals apparition (34.2% vs 3.6%).
And again more obvious is the fact that shoes rich of 8s and 9s make more probable this possibility.
...

yet a fair amount of 5-card hands go toward Player side, think about standing or natural P points or asymmetrical spots where third cards help the P with Banker standing.

Do not forget that B winning hands are payed 0.95:1 and that a fair amount of 5-card hands (nearly 40%) are strongly favorite (at various degrees) to win the P side at the start.

...

It's the same reason working at asymmetrical spots when the third card instructs the Banker to stand while Player has a winning hand.

Maybe in the future Banker will win by standing points but the asymmetrical spots are somewhat consumed as they are limited in their appearance.
With the decisive difference that P standing points get a 40% probability to happen whereas asymmetrical spots have a 8.6% probability to happen, that is 4.65 times more likely to show up.


    I agree 100% on this last one. Im always astonished at how many players will state or suggest their "go to" method is to always bet Banker. Its funny that in past couple months I've had two acquaintances that play poker and related games(with above-avg skill) on a semi-professional level come up to me and state: I've got this friend or brother,...etc that makes a killing playing Bac and he always bets Banker continuously and everytime, ("and he wins big,makes a living,makes huge money...etc), as if they are trying to impress me knowing that I mostly play Bac.
They lose me as soon as they state : "he only bets Banker everytime".


Continued Success,

Hi KFB!
I can't agree more on your words!

If people claim to win constantly by always (or only) wagering Banker side, they should win constantly by always wagering Player side: the difference is just a miserable worse 0.18% ROI.
In fact ask them to let you know what's their Banker winning percentage: to get a long term advantage the wp must be 51.3% or higher.
At Player side it should be 50.1% or higher.

Obviously in the short term such values could be misinterpreted as a kind of 'magic skills', in reality it's just a chance factor.

Neither a statistical long term study made on the slight math propensity to get more B rich patterns than P rich patterns will help them (or anybody).

The only way to win constantly at this game is trying to catch the 'actual' card distribution features that are surely dictated by several levels of asymmetry (about this topic in general I recommend the reading of N. Taleb books).

This has almost nothing to share with common strategic lines as 'following trends', unless we have strictly determined what a 'trend' really is and what are the limits of intervention along any shoe dealt.

as.

Random walk steps applied to a bac card distribution

In a random and symmetrical proposition (e.g. a coin flip succession), random walk steps are undetectable by definition. Several roulette studies (ignoring zero/es) confirmed that no matter after which trigger point we'll decide to bet (example after a 3 or greater sigma happening at one side), every next spin will be 50/50.
Obviously.

At baccarat things work in a similar way only apparently.

First, the propositon is asymmetrical by the rules (B>P);

Secondly, there's a more important asymmetrical factor regarding the actual card distribution, being finite and slight dependent;

Third, we have strong reasons to think that bac shoes are not perfect randomly shuffled.

Putting things into a semplified scheme:

- Coin flip successions: symmetrical + symmetrical = symmetrical

- Bac shoe successions: asymmetrical + asymmetrical + asymmetrical = asymmetrical

Itlr both propositions will approach more and more the math expected values (50% for coin flips and 50.68%/49.32% for BP hands) but surely by different random walk steps.

Nonetheless at baccarat there are some patterns more likely to roam around the 0 point, that is getting a lesser number of bell curve thick 'tails'.
Those that can be a 'heaven' or a 'hell'.

It could be surprising that three levels of asymmetry are roaming more probably around a 0 neutral point (providing the patterns to look for), but that's it.

A possible explanation is that whereas a math asymmetry (B>P) works as a costant, the remaining two factors tend to overwhelm (or conversely to endorse) the first propensity.
In practical terms everything stands as a more likely 'clustered' probability working at different steps.

Naturally such steps are whimsically placed along shoes, anyway 'heaven' and 'hell' will get more detectable spots than an unbeatable random and independent symmetrical proposition. For their more likelihood to go toward left or right up to cutoff points and always considering a 0 'target'.

More on that next week.

as.

For sure to win itlr at baccarat we need to 'catch' the 'best' propensity coming out from the actual card distribution.
We do not need astounding propensity values to be ahead of the math negative edge, everything moves around tiny percentages that in the long run will add up.
Definitely whenever those great propensity levels come around we better take advantage of them. Yet they are not so likely to show up.

That's why statistics will help us to define the terms of intervention as huge propensity values are not coming around the corner.

A given card distribution eliciting a univocal propensity happening for the entire shoe is out of order, it's way way more probable to get several 'propensity' levels.

Card matchings forming B/P or r/b results act by several levels quite different than a 50/50 independent model.

Thus we may introduce the term of 'shoe multiple propensity levels', meaning that cards may or may not endorse the formation of some patterns.

So propensity P could be splitted into subclasses of P1, P2, P3 and so on.

Later.

as.

If shoes will produce inconsistent patterns for long, that is featuring propensity values not surpassing some cutoff points, the game would be easily beatable by a simple MM procedure.

Unfortunately the number of shoes NOT featuring inconsistent patterns are a large minority, so we're somewhat forced to 'guess' when different levels of propensity will surpass or not such cutoff points.

Therefore to hope to win itlr P (P1 + P2 * P3, etc) must be cumulatively > R (random world). In other words if P=R we cannot have a single possibility to win itlr.

Obviously we can confide that the R is just a virtual entity to face, as cards cannot be properly and randomly shuffled per every shoe dealt.
So R is not a perfect R, then P must be larger than R at least at some portions of the shoe capable to erase a P=R effect.

Yes, even R will produce natural Ps, but in the long run those Ps are surely inferior to the number of Ps following an actual unrandom card distribution. And needless to say, right guesses made when R seems to overwhelm Ps are just symmetrically placed. That is unbeatable by definition.

That's why Alrelax stressed about the importance to adhere at most of what the actual shoe is presenting. That is not hoping to get infinite Ps, but to select the situations where P should be greater than R as it's a natural occurence at unrandom shoes.

Put things into a simplified way.

We think that a kind of P propensity will happen after a given event(s) happened.
Of course whether the production is really random, the number of right 'guesses' will be equal to the number of 'wrong' guesses. Unbeatable propositions as P=R.

Actually a bac shoe is oriented to form many P flows, even multiple low level Ps will produce a pattern. Easily beatable by a MM approach. 
Do not be fooled about the supposedly 'randomness' of the shoes, it's a fkng idiocy stated by mathematicians that like to mix different asymmetrical situations into a whole.

Baccarat is a game of clusters getting different levels of appearance.
Each level follows a general probability to happen that must be compared to the actual probability.

as. 

Remember that at baccarat you can choose the side to wager any amount you wish and anytime you wish.
And maximum limits are quite huge.

Moreover, bac players are considered as pure losers even while betting thousands.
At bj tables, people utilizing a $20-$80 betting spread seem to be a treat for the house. LOL.

as. 

As you said, "Baccarat is a game of clusters getting different levels of appearance. Each level follows a general probability to happen that must be compared to the actual probability."

Which can happen at any time, 'due' or not.  As well, might and very well be extremely large at times bringing about words describing such clusters as 'rare' and 'unbelievable'. 

And, what sends me ROMAL with real tears from my eyes  caused by laughing so hard, are all the forum bac/gambling mathematicians, when they carry over their stats shoe after shoe after shoe, stating how their numbers are correct because they run continuously and cannot be limited to each shoe, etc.

Al, I've carefully read your post about randomness at bac, I'll present my opinions on the subject here.

Baccarat randomness

You sayed well: probability theory and randomness are very intricate fields, many times producing strong disagreements among statistics and math experts.

Here we are talking about a gambling game and, more importantly, about its possible beatability.
Whereas it's debatable a sure and firm definition of 'randomness', we are certain that a pure random EV- game can't be beaten.
I agree with you that 'reading randomness' is a foolish attempt to take the best of bac and having a 0 possibility of success.
But since I know some successful and consistent bac winners (including myself, you and probably a couple of other forums writers), I infer that the most likely (!) explanation is that the game is not so randomly distributed thus giving very few players the hints to beat it itlr.

In fact, if a player tells he/she's beating baccarat and at the same time stating that the baccarat production is 'random', well this person should be able to win even at single 0 roulettes where for sure the production is random.
Now I haven't known a single person capable to win at unbiased wheels but at baccarat such people exist.
Of course another possible reason is that roulette outcomes are totally independent and bac is a slight card dependent game, but again that confirms that some portions of the shoe are not random.

Uncertainty vs randomness

Casinos' profits prosper about a indeniable math edge and for the 'uncertainty' of the outcomes that has nothing to share with pure randomness.

For example, even at bj where math edges shift from the house to the players, the 'uncertain' world poses a serious treat about such math edge. As the 'randomness' condition may not be satisfied.
That happens as we just make an 'estimation' of the high cards-aces/low cards ratio; in poor words our expected profitable high cards portions could be confined right at the unplayable end of the shoe.
That is when a favourable math edge is verified, we need shoes to be shuffled by a fair level of randomness.

At baccarat the opposite is true. IMO.
Since we can't rely upon a math edge, we must 'hope' shoes are not perfect randomly shuffled and this thing can more likely happen at some portions of the shoe as almost all cards are used to be played.

The fact that we are 'uncertain' about the next hand(s) destiny doesn't necessarily mean hands are 'randomly' shuffled as the uncertain world can be measured whereas a random world can't besides the common 'imperfect' and general values not helping us in any way. (Obviously this last part of the statement implies a kind of perfect random shuffle for every shoe dealt)

How to detect a possible non randomness at baccarat

If some people win consistently at this game just two things must happen: a) bac productions are non random; b) baccarat successions produce flaws capable to erase and invert the negative math edge.
Since the scientific world has denied the latter possibility (at least at a significant level, see Jacobson studies about a perfect PC card counting technique made on BP hands), the only possible explanation is about non randomness whether properly exploited.

First, how to classify any events succession as 'really' random?

General probability theory doen's help much in this way for its intrinsic limits and foremost for a lack of deep studies made on the subject. Always oriented to get a kind of advantage by knowing certain cards concentration/dilution helping one side.

But, more importantly, is the succession we're facing really randomly produced?

Jaynes, Keller, Diaconis and others made important studies about this issue, yet the best definition of randomness I like is the RVM's one.
Simplifying, no matter which point of the succession you take, a random production will get the same limiting values of relative frequency at any different point considered.
And at baccarat this thing doesn't exist.

Say I'll ask you: what's the probability that a toss or several tosses of a specific dice will show a '6' face?
Obviously the 'theorical' best answer is 1/6, that is 16.666%.
Say that after 100 tosses the frequency of '6' is 20%. You could think that this is a natural effect of positive variance getting normal sd values (even quite deviated toward one side).
If after 1000 tosses the frequency still stands around 19% or 20% you might think that the dice is  biased or that the 'shooter' doesn't 'randomly' toss the dice in some way.
After 10.000 tosses you'll get a clearer picture of what the 'actual' probability is about the dice getting a '6' face.

But regarding the actual dice probability to show a '6' there are more important properties to look for other that the final occurence: for example what's the most likely number coming out after any number different than 6.
Or splitting the outcomes into odd and even categories, or classifying the average gaps between '6's.

At baccarat we might think that the production is somewhat biased, but in the same way a possible dice may be biased or unrandomly tossed under multiple statistical parameters (and most of the times is not), we need a proper amount of hands to exploit such flaws. Shoe per shoe.

as.

It is truly random, BUT.........The Key Lies in the Following:

.........."So, if it is impossible to predict randomness, what can we do to predict winning hands? I would strongly say, that we should expect some sequences to be perfectly parallel as well as, those that look non-random or simply lopsided. Do not limit yourself to the consistency or the inconsistency as the indication of what to wager. Rather, be alert to the sequences producing parallel or lopsided outcomes and siding with those."