Thanks Al, the same from my part (along with a couple of hs players I'm used to play with).

It would be a great idea to meet us with Sputnik and Lungyeh too, providing they want to make a visit in US.

as.

I hope you'll keep posting, you are one of the best gambling researchers I know.
And for that matter I'm sure you have devised a plan capable to overcome 5 sigma negative deviations, the problem remains about the practical aspects of it.

Hope to read you soon.

as.

Since I can't touch the SM machines topic, let's compare baccarat with roulette.

At roulette every spin will provide symmetrical probabilities, since the probability of each number or groups of numbers remains the same (1/37, 2/37, etc).
Say the whole model we are playing into is symmetrical by any means.

At baccarat every BP hand will be formed by two distinct and very different probabilities: 50%/50% and 57.93%/42.07%. Those different probabilities alone makes baccarat an asymmetrical game.

Of course every fkng shoe dealt will present different values of such asymmetricity, either in terms of numbers and, more importantly, in term of distributions.

Everybody reading my pages (btw, thanks to you) knows that the asymmetrical 57.93/42.07% value should come out on average 8.4% of the total hands dealt.
A probability value very similar to betting 3 numbers at a single zero roulette (8.1%).

Every player having a decent familiarity of both roulette and baccarat would expect that a similar probability (3 numbers vs asym hand) will produce similar dispersion values taken on the same 75 hands sample.

It seems this is not the case.

Easy to argue that a shoe formed by a finite number of cards burnt hand after hand is quite different from a so called perfect symmetrical world happening at roulette.

More importantly is to notice that when a 3 numbers group hit at roulette the winning probability is 100%, whereas at baccarat we are still fighting with a well lower 15.86% edge.
On the other hand, every other spin not hitting our 3 numbers provides a 100% losing event whereas at baccarat we still get a "fair" 50% (taxed) probability to win.

Itlr, a perfect math plan should be oriented either to bet P trying to escape the 42.07% unfavourable winning probability or, it's way better, to catch the 57.93% winning probability when betting B.

In truth a wonderful virtual player capable to always bet P without crossing one time a single asym hand will get very tiny profits (p=50%, yet certain card distributions happening on symmetrical situations help the Player side thus enlarging a bit the P probability). But there's a more excellent player, that is whoever is capable to bet B as he/she assessed that an asym hand will come out more likely within a more restricted range than what math dictates.

Some very experienced players (Alrelax and Sputnik surely belong to this list) have raised the ability to catch or abandon the situations where B or P winning probability ranges are more or less restricted than what the old 50.68/49.32 ratio dictates.

But the common denominator we have to put in first place is that shoes are not randomly shuffled (say it's physically impossible to arrange cards by so called perfect random models).

There are many ways to detect this, I prefer to choose a strict objective betting placement following the best "randomness" definition ever made by some statistics experts.

as. 

Imo there are no other ways to beat the game unless we have proved that bac is working by more or less unrandom standards.

Of course we can't rule out the possibility that an "usual" unrandom world sometimes could take the resemblance of an unbeatable random model, that's why we prefer to discard shoes not fitting our plan at the start instead of trying to get a kind of "more likely world" in the subsequent portions of the shoe.

More on that later.

as.

Baccarat is one of the purest form of gambling, no wonder it has acquired an increasing popularity over the years.
After all players must guess a pre-ordered succession of events and getting the luxury to choose what, when and how much to bet.
No one other gambling game provide such features.

But to be consistent winners we must assess by the greatest possible precision what's our real probability to win or lose.
Since a baccarat shoe is composed by a finite number of cards where many of them are "key cards" we should estimate what are the real probabilities to get an event or the opposite.

We all know that B probability to win on each spot is either 50% or 57.93%, whereas P probability to win remains at 50% (actually some card distributions favor P side more than that).
Itlr, that is after having mixed several outcomes (maybe springing form different sources) the average BP probability comes closer and closer to the 50.68/49.32 ratio.

A total different issue regards the probability of success (POS), that is the probability to win after a given succession of bets.

Whereas the probability to win or lose on each side remains constant and mostly unguessable, shoes present a variety of POS that equals to 1, that is the certainty that at least one searched event will appear.
Of course the possible unfortunate counterpart is zero, that is that the event searched won't appear at least one time in our shoe or after a short sequences of consecutive shoes.

Easy examples where POS=1 (probability equals to certainty) are:

- shoes producing at least three streaks

- shoes producing at least one P or B double (unless long streaks happened on either side)

- shoes producing at least one asymmetrical formation along the way

and so on

Of course such strong features generally won't be of practical use without the use of an impossible progression, unless being mildly moderated and multilayered conceived (Albalaha could instruct us about this).

Forgetting the single shoe probability which could be easily affected by a kind of so called "randomness", POS may be endorsed by waiting the appearance of huge unlikely situations.
The more we wait for the "unlikely" events, greater will be our POS.
A thing that cannot work at other independent models as roulette, for example.

Say we are putting outcomes vertically in a grid made of columns of 10 spots each (a 10-hand bead plate not considering ties). Now we want to form a new registration of I and O results regarding the left position of the new outcome.
At the eyes of the experienced player it will appear very soon that such new random walk isn't affected by a an indipendent and unguessable model, as a place selection procedure will demonstrate that most shoes won't follow a 50.68/49.32 ratio by any means.

Some spots are slight more likely than others, some ranges of apparition are more likely than others.

as.

I completely agree and those suggestions are the reason why casinos will make a lot of more money than expected by math.

as.

We've been taught for years that B probability is 50.68% and P=49.32% but probably just a couple of persons explained mathematically by combinatorial analysis why those percentages were obtained.
A shortcut would be to consider a very long sample of observations and, voila', those percentages tend to coincide with those values.
Therefore theory and practice meet.

But it's interesting to notice that such probabilities are the reflex of dynamic probabilities since B probability varies with big jumps from 50% to 57.93%, values that P side must accept passively.
Moreover the game is, yes, intended as partially dependent anyway at a degree not substantially altering the features of a perfect independent world happening at a fair roulette, for example.
Finally and fortunately nobody investigated seriously whether certain outcomes come from a real random production, an essential requisite to make unbeatable a slight taxed game offered at casinos.

Actually and by utilizing a very strict definition of randomness, no one live shoe is randomly produced even though for practical purposes not every shoe will be playable (at least by the"human" possibilities tested so far).
That's because is very difficult (not to say impossible) to arrange cards in a way that certain events cannot be perfectly independent to others and, of course, the word we must take care most of is dispersion.

The fact that after 10.000 BP resolved hands dealt on average 5068 are banker outcomes and 4932 are player results doesn't necessarily mean that every possible distribution will follow the dispersion values known regarding other propositions.
Neither should be considered an insurmountable obstacle the tiny tax applied at baccarat.

As previously sayed in my posts, it could be that what I label as "random defects" are just  instrinsic flaws of the game not investigated by so called baccarat experts, mainly oriented by nature to find math advantages (card counting techniques).

At any rate we think that dispersion values cannot be practically limited when apllied at a random situation even if the game is asymmetrically governed and acting under slight dependent processes. Thus a kind of unrandomness must act in some way.

For a moment let's say the first initial collection of BP results appears as really random. Therefore unbeatable. No problem with that.
That is per every class of W situations we'll get a proportional class of L events with huge degrees of variance.

In order to confirm that outcomes are random, we'll make certain sub collections derived from the primitive simple BP succession every bac player in the universe relies upon.
If the first collection is really random then every each sub collection must be random, otherwise it's negated the perfect randomness condition.

For example, say we build our personal derived road, that is a random walk in such a way:
Anytime a winning natural point comes out on a given side, we'll register the outcome of the next hand as I (identical) or O (opposite) in relation to the side which previously won by the natural point.
Therefore per each shoe we'll get a I and O succession having an average 34.2% probability to appear, meaning that on average such new road will get around 26 decisions.
No surprises, the average number of I and O after this new collection will be as expected but what differs on most part of shoes dealt is the distribution of patterns that could alter on our favor the probability of success.

It's astounding to see that shoes coming from the same shuffle procedures acting on the same shoe will provide the best opportunities to grasp a possible unrandom world that, I repeat, shouldn't be considered other than from a strict dispersion point of view.

as.

But what's so difficult about winning 2 times in a visit ?

Thanks Lung for your reply, among your interesting points I highlighted this passage.

It's so difficult to win in two visits in a row as people treat baccarat as a kind of lottery where each ticket they are buying offers (slight) unfair odds.
A lot of ding-dong? Hit the jackpot. A lot of singles and doubles or consecutive streaks? Another jackpot.
Strong imbalances between B and P? Again it's a jackpot as well as every kind of repetitive patterns.

Now, are there reasons to think that along the way we'll hit such lotteries more often than not?
Yes, such (small) jackpots come out with a decent frequency but not enough to balance and invert the constant house edge. No matter how sophisticated is our progression plan or MM.

Sayed that, I'm not ruling out the possibility that some acute players tend to get a clearer picture of the whole situation without the knowledge of possible randomness defects or whatever could alter an unbeatable random model. Still the common trait of such players is to play very few hands.

We ought to remember that without math advantages, it's impossible to beat any EV- game whether considered randomly distributed.
Therefore our only option to beat it is to consider and study why, when and how could be unramdomly placed.

No luck intervenes on our side.

as. 

Imo bac is beatable as the "general" probability doesn't correspond to the "actual" probability.

According to the general probability, itlr each spot will follow a 50.68%/49.32% BP probability, thus no one betting method could find spots where 50.68/49.32 ratio will be higher (or lower) than expected. In a word that the statistical deviations will follow such values, practically meaning that the model we are playing into is randomly placed and very very very very slight dependent at most.

Actually efforts made to find profitable spots were made ONLY by math procedures.

Easy to see such procedures contain a big mistake as they were tested on pc generated shoes where randomness supposedly prevails.
Moreover, they couldn't take into account the probability of success of certain events considered by ranges, as they kept for grant that whenever A>B any other subsequent situation will follow a costant asymmetrical line and it's not the case at baccarat as a single 8 or 9 falling on one side will dictate mostly the outcome.

as.

Actually along those suggestions you touched a key valuable aspect that it's very difficult to falsify mathematically and that's our fortune.

as.

Quote from: alrelax on December 13, 2019, 11:34:49 PM
Excellent. I have actually started an outline and I've identified no less than 10 advantages that I have used successfully, at times, over the past years .

More on that at a later date thanks for putting in the input.

Thanks Al!
I'll wait further comments from you about that.

as.

In reality no one long term winning player wants to inform the enemy about the details by which this game could be beaten. Casinos prosper about the ignorance of their bettors and not only about their fkng math edge.
And truth to be told, baccarat is still alive as the vast majority of asian players rely on luck about their bets destiny, say the persistence of certain trends showing up along the way.

I can't rule out the possibility that other researchers had scientifically theorized the unrandomness of baccarat, yet if we keep thinking the game as a randomly produced game we're going to nowhere.
Probably we'll get better odds to cross a turtle roaming on the Mohave desert than trying to win a game we think to be randomly placed.

Actually and even taking for grant that the game is really random (a horrendous mistake), we can build certain betting lines that will minimize the variance factor working into an asymmetrical proposition.
Next post will be about those methods.

as. 

The best baccarat player in the world is not the person who is capable to win the larger amount of units but whoever keeps his/her losses very close to the negative edge value (say an average -1.15% of total money wagered on BP hands).
At the end of the year we should try to recollect how money we have bet and how much we have lost at tables. If the sum is around 1.15% of the money bet we are really in good shape.
That is the number of W should be almost equal to the number of L, only vig caused our debacle.

Unfortunately it's quite likely we had lost more than that, maybe we have added some side bets here and there or that we have used a bad money management. Of course no MM could provide us a winning method thus we should accept the idea that the game cannot be beaten other than by a proper bet selection.

Imo there are only two ways to get a winning bet selection working itlr:

- flat betting strategy where number of W exceeds the number of L and the vig impact

- short multilayered progressions oriented to get a key W or Ws happening on restricted sequences considered as profitable

Alas, those strategies cannot win when applied at random EV- games. And for that matter they can't win at EV=0 games either.

Many baccarat books or, worse, internet system sellers, keep stressing us about the importance to "quit when ahead". They want to teach us how to win and then they put in emphasis such silly phrase.
If I'm winning and I have to quit to preserve my bucks, why not starting to play a kind of an opposite strategy?
And when do I know I should quit because I've reached the apex of my winnings?


Gambling is a game of streaks, at baccarat say a game of "gaps" between two opposite situations that not necessarily must be B or P hands.
It's just the natural streak appearance that destroy every system. Providing the game is randomly placed.
Then our task should be directed to spot situations where a lesser number of streaks should be more probable than normal thus increasing the likelihood to get a more expected outcome. And it can't be that unless a kind of unrandomness or super complex dependency is acting.

But even if you take for grant that bac shoes are not randomly formed, you can't forget that we're speaking about an edge quite high but limited to very few spots and not to every shoe dealt.

Let's make an example of one of many singular random walks we could put in action fictionally and oriented to disprove the concept that bac shoes are collectives.

Say we want to set up a short "road" where we'll classify outcomes as A or B depending whether after a winning natural point happening on a given side the two next hands bet on the same side are producing at least one win. If we win in two attempts we mark 1, otherwise we mark 2.
Thus our trigger to start or follow up the classification is the winning natural happening on either side.

Example: B9 winning point, we'll bet two times B; if we win we mark 1 otherwise we mark 2.
Whenever naturals do not show up or by gaps higher than 2 we do not mark anything.
In reality this is an irregular random walk in the sense that two-step betting action will get an obvious nearly 75% EXPECTED probability to win whether a kind of progression is applied.
And naturally it's not about the general more likelihood to get 1 or 2, just the distribution of such 1s and 2s. That is that that 75% value is more or less deviated toward one side. 

No matter how whimisically are the actual results, this new 1/2 line most of the times won't follow the natural probability distribution, especially from a place selection point of view.
Not everytime but most of the times.

as.

To get a better idea of how baccarat really works, I introduce my concept of "random walk".

A player whimsically betting here or there, following trends, following lucky or unlucky players, playing drunk or perfectly sober, wagering by the influence of Alprazolam or THC or whatever, constitutes a random walk.
The same about big road and the four derived roads, now being mechanically ordered. 

Naturally even a player wagering only one hand per every 2, 5 or 100 shoes is a random walk.
A random walk is just an infinite sequence of W and L successions having almost (as B bets >P bets) the same probability to show up.

For example, if the shoe provides really random outcomes, W/L dispositions follow the old 50/50 rule dictating that the probability to win (or lose) four hands in a row is 1/16 (6.25%) and so on.
But such probability is real only when the shoe is producing random outcomes in the sense that no matter which spot of the shoe we start to bet those outcomes will be unaffected by previous events (place selection).
Naturally and in absence of informations, we have no means to detect whether the first four hands dealt of a new shoe are really randomly placed or not.
In any other scenario, if we think the future four hands will give us a sensible better 6.25% WL ratio in either direction we'll get an edge. Same about lower or higher probability circumstances.

Imo, the more we wait for favourable dependent and allegedly unrandom situations, the better will be our results, providing we proper classify the playable shoes.
It's easier than what you think.

Tomorrow I'll talk about what I improperly name as "limited" random walks.

as.

Good post Al.

Anyway I think that gambling forums defend readers by definition as it's quite easy to falsify directly the theories, ideas or methods presented (without insulting of course).
Providing they are offered for free.

Since many games are still considered unbeatable and no matter how prestigious is the source, there's no point to purchase anything from anyone and this is just a very good starting.

as.