That line was obviosuly a sarcastic note about how is easy to distrust other's ideas and works putting atop of it the old fkng math issue saying that bac cannot be beaten.

Maybe it could.

as. 

 

Nice posts Al.

Now expect some people telling you that such bets have an astoundingly EV negative itlr, or that those long streaks happen very rarely. :-)

If the game isn't beatable mathematically or statistically (and of course it is) this thread should demonstrate that some betting lines may cause the house to get a bit of discomfort, albeit temporary.

as.

Quote from: alrelax on May 24, 2019, 01:46:04 PM
I am hurting today, about 1 and a half hour sleep.  What a session, what a great session.  And I did this last night because of the great email I got from AsymB, thank you buddy!  Those few sentences sent me out with the song I love in my mind, Burning Down the House.

Greeeaaat job buddy!! 

I like that song too...

Now do not try to give the money back. :-)

Soon it'll come a day when casinos will be forced to treat baccarat as black jack as every single bet they are offering is 100% beatable.


as. 










 

Thanks Al.
We're speaking the same language even though is taken by different angles.

There's no fkng way that playing into a random taxed model we'll be more right than wrong by instinct or by following trends or by mechanical placements unless we assess carefully what's happening, what's happened so far and what is slightly more likely to happen.
Moreover, we should know what's our real goal per every session played. We can't hope to win every session and we can't break even after an harsh losing session.
Are casinos going to win every fkng day or week? No way.   

The certainty of a given outcome can only be extracted by the proper use of time or by utilizing other tools (defect of randomness for example).

Example.

We know that the probability to get a big road shoe without any B or P single or streak or double or 3+ streak is close to zero.
But if we consider the common three additional derived roads is absolutely zero. ZERO. Mathematically.

Going down to some of less likely outcomes and testing a lot of real shoes, we'll see that what happened so far tends to represent it in the same shoe and, at a lesser degree, what not happened so far gets a slightly increased probability to appear.
The process is endorsed by a supposedly flaw of randomness.

The problem is when to start to bet and when to stop it.

It should be an idiocy to stop the betting when crossing a winning streak, but the exact counterpart (stopping to bet when losing) will provide huger benefits.

Casinos want us to gamble, playing every hand or betting side bets.
Therefore we should disappoint their hopes, so betting very few hands (or betting small every hand and wagering 10x or more on key hands).

It's very likely that a team formed by me, Al, Sputnik, roversi, Bally and some others will crush every casino in the world by the simple concept of convergence of probability taken by different angles.
To get a decent profit we must join $1000 or $2000 min tables, let casinos think we're stupi.d gamblers.

Our goal will be to be banned in the casino we're playing at.

Our motto is

put your fkng math edge in your behind

as.

Thanks for your inputs and replies.

Look how awesome is to know that one specific situation is going to happen (or not) per a given class of shoes dealt.
But the specific situation must be firmly set up in mind BEFORE playing and adjusted accordingly to what the shoe produces but always in terms of "playable" or "not playable" shoe. 

In order to do that we need to take advantage or, even better, to build several random walks endorsing the probability to look for the searched situation.

Playing by instinct or by experience may be valuable random walks but too much affected by emotional and actual factors. More importantly playing instinctively leads to bet too many hands.

Five random walks are directly diplayed on the screen (big road, bead plate road, etc) but we could build infinite random walks even not based upon the B/P results (for example about the first and second card dealt).

For example, one of the artificial road one can easily add is the third to last hand registration.
We wait three resolved hands then the fourth hand will be classified going back three hands, then registering it into two separated columns (S=same, O=opposite).

for example:

BBPPPBPPBPPBBBB is

OOOSSSSSOOS

Of course there's no a direct value in registering the outcomes in such a way, it's just a ploy to raise the probability to cross the searched situation that could be delayed on other roads.

But the real value of registering multiple random walks simultaneously is whenever our plan dictates to get B or P on more than half of all random walks considered.

The reason is all about the difficulty to get a sudden inversion of probability's plan on many roads, at least on more than half of them.

In case our plan suggests all roads to get the same outcome in the same point (a relatively rare finding even adopting only 3 or 4 roads), our wager will get an astounding EV+.

Despite the wonderful profits such scheme will produce, I know there's a methodological issue to be solved: the presence of ties. At least theoretically.

Since we have to discount ties in our registrations (besides big road and bead plate) we know that results' distribution may be affected in some way.

In a word, shoes particularly rich of ties at the start should be avoided (along with the shoes not fitting other conditions we're looking for).

as. 

What should make "unbeatable" a given plan?

My answer: the certainty that a class of events must show up per any shoe at different degrees of presentation.
I'm not talking about a very very high probability that something is going to happen but the certainty that something happens.

Of course just knowing that something is 100% going to show up doesn't help us too much as we need to estimate when and how many times those events come out per any shoe.
In a word, we must build a proportion between searched events and number of attempts.
If we know that some shoes will provide just one searched event, we must restrict at most our attempts to spot this event as we're risking many to win one.
Conversely, knowing that some shoes will present many searched events, money utilized to spot those situations will be spent with a way higher probability to be right AT LEAST IN ONE SPOT. That is the minimum requirement of certainty we should look for. 

That is I do not want to win several bets within limited intervals of time with high degree of uncertainty but to win very little in safe conditions of certainty within relatively large amount of samples.

If such certainty would be ALWAYS limited in the space of 5-6 attempts per shoe, a simple martingale would solve the problem.
Unfortunately not every shoe will provide the room to make 5-6 attempts, in other words certainty becomes certainty only in selected circumstances.
And not by magical forces, just for a matter of space as any shoe is a finite separated dependent world.

See tomorrow

as.   









 

Quote from: Jimske on April 05, 2019, 05:16:32 PM
You are the scientist, mathematician and statistician analyst.  You been working on this for how many years?  Were even going to write  book about it.  What is it that you have to report after all this time?

J

I've been studying this game for 8 years long, I think the game is beatable only as decks are not properly shuffled.
My best accomplishment was and still is when I was contacted by a couple of high stakes players willing to be mentored by me and getting me a cut on their winnings. And, you know, it's a lot of money when people are wagering $5000 or more per hand.

My final conclusion is that the game can't be beaten mathematically or "humanly", the solution remains in the middle (virtus in medio stat)
Key words are time, space and asymmetry.

Time: you need time to get some searched probabilities happening. And of course time may work in casino's favor mathematically or in player's favor statistically. It's up to us to decide when and how. 

Space: baccarat is made by finite portions of probability. Many factors work in that. 

Asymmetry: a constant slight asymmetry works either by rules and/or by actual conditions (card distributions, outcomes' nature, etc).
For example, it's impossible (not only high unlikely) to get certain dispositions in the same position shoe per shoe.

as.

I'm looking at this post just now.

I like to work trying to adhere at most by a scientific approach, that is every observation/theory must produce measureable results. And, more importantly, every theory must be proved by falsification.
Did you measure your points?
Human thoughts, instinct or presentments even based on objective outcomes cannot be a valuable guide to get the best of it at baccarat unless they are carefully measured.

And of course the more heterogeneous are the parameters involved in a theory, larger should be the sample confirming such theory.

as.
 

At least when I'm sober and focused :-)

as.

Quote from: Jimske on March 27, 2019, 08:02:20 PM
I guess I missed anything Glen wrote about unplayable shoes.  Maybe you could opine on the subject and present an example "unplayable" shoe?  Thanks.

J

In order to win two units one needs to win one unit first.
It's a mathematically undisputable fact that it's a lot more likely to be ahead of one unit than to be ahead of two units and so on by a logarithimic scale.

Therefore any strategic plan should be oriented to get that one unit profit per a given series of trials and this is an awesome result as it will deny the math negative edge.
When an initial loss comes around, we must hope that subsequent outcomes will first balance the previous loss, then inverting the losing line dictated by the first loss.
In a word, to get a profit we need two positive results to balance that first loss.
Without a verified edge, attempts directed to get a balancement even by the use of progressions represent a totally worthless effort. That is it will be more likely to be +1 after a 0 cutoff scenario than to be +1 after a -1 spot.

People who like to state they can get multiple winning units per any given shoe are bighornshitting themselves and anyone reading them.

It's not fkng possible to be ahead of multiple units not only per any single shoe but per a decent sample of shoes unless whether a lucky (unlikely) positive variance is acting.

We can't be right on every shoe dealt, maybe placing the "turning point" to -1 is the best option to get an edge itlr, that is to get rid of that shoe (best if the process is made fictionally).

To get a long term edge at this game one needs to bet huge and very rarely and it would be an outrageous statement to say otherwise. Especially if someone tries to demonstrate that every single shoe is controllable, a total fkng bighornshit.

No human can be more right than what probability and math dictate, otherwise this game wouldn't exist.

I'll be more convinced of the contrary if any "foolproof system" claimer would bet $1000 or more per hand, and this thing isn't going to happen.

Watch me when I'm playing and you'll get a better idea of what I'm talking about.

as. 























   

Nice reply Al. Thanks.

It's quite known that I'm hardly working to set up a no brainer mechanical method capable to get the best of it no matter how whimsically will get the hands distribution.

Many shoes are not playable at all just for the reasons Al outlined in his several posts.

This as sh.it tends to come out in clusters than being balanced in subsequent portions of the shoe.
As well as good outcomes can be detected right at the start.

Into a finite and dependent model, probability works by various degrees and whereas some levels are "unlikely" reached, our best move is to not play at all.

Notice that imo we do not want to follow "unlikely" lines, we just want to get rid of them.

Baccarat is a game of balancements and deviations, whenever a given deviation is "due" we follow it, whenever is harshly going against the "expected" (it's sufficent to lose 2 bets in a row) we better wait the next shoe.

Imo never ever try to adhere at an "unlikely" distribution as it's more inclined to form difficult detectable results.

It's like poker where to get an edge you have to fold some possible best hands.

as.


















 

Summarizing, the main feature why bac can be beaten is because certain spots can't be missed, meaning that probability plays a huge and decisive role in that.

Of course such feature cannot be theorized whether a perfect random world is working, otherwise I have to put into the trash bin all the math involved.
And I can't do this.

as. 

Baccarat provides an important feature many times we forget about.
The casino's winning probability cannot be less than 50% unless bets are placed on Banker side.
Yet the economical return favors casino every bet we'll make.

Thus even if we're the world champion geniuses of bet selection, we are still playing a 48.94% or 48.76% proposition on our winning probability unit wager.

The only way one could lose only 1.06% or 1.24% or an average  mix of two of the total money wagered is by flat betting.
It doesn't matter which kind of selection one utilizes, by flat betting one is math expected to lose from 1.06% to 1.24% of his/her total bets.

Actually most bac players want to recover losses by increasing the bets, but they forget that the more they'll increase the wagers better are the opportunities to lose everything.
The same is about increasing the wagers when positive streaks seem to come out.

There's no one single possibility in the world that after a decent trial of shoes one can get the best of it by increasing the wagers unless certain "battles" provide a very low variance impact.

Say we have found an astonishing -4 +4 random walk working onto two opposite situations.
We know that when the +4 level is reached an unlikely still possible 8 losing streak is going to happen (that is a shifting force going toward -4 point). Are we going to bet?
No fkng way.
The same about a +3 or +2 random walk position.

We do not want to raise our bets to win just one fkng lousy unit, we want to get the best of it by increasing the probability to get one unit profit immediately or, at least, after two bets.

Say we are going to join an HS table after forming a bet selection site team, where each member put $5000 at risk . Our standard unit will be $10.000 and the maximum wager on that table is $20.000.
Our bankroll is $200.000, that is 20 units. That is 40 members.
Our goal will be to win just one unit after 3-4 shoes dealt, so every member will get $250 and the probability to lose the entire 20 bets for each player is very close to zero. Say it's zero.
In fact the probability to win one unit after 3-4 shoes dealt is 99.999%.

Are you going to join such team or do you think you'll get a better edge?

I know many players tell you they'll get a better edge but they are deadly wrong.

We do want to lower the variance at most and, by the way, nobody has shown to you that a given method can get the best of it by such "low" win rate.

Still you can't be wrong about this "silly" method, it's just a matter of waiting the right opportunities to come along.

And actually it's what we do, putting at risk 20 units to win just one unit after 3-4 shoes dealt by 99.999% accuracy.

Imo one needs to risk a relatively huge bankroll to win something after a given period of time, think that casinos are going to put at risk virtually infinite bankrolls to win our miserable buy-ins.

Guess what casinos think when we're constantly betting $100 or $300 (EV- situations) and suddendly we're raising the bets to $10.000 where our EV will be positive.
They'll think we are i.diots, but they'll fear our bets as they need 100 or 33.3 wrong bets to balance our previous no edge wagers.
Situations that cannot come along.

as.   













     






 


 

Yep Al! Still many players like to wager via strong progressions.

Back to the subject.

When we consider two opposing events A and B having the same (or almost the same) probability to appear, we'll expect deviations according to the binomial model.
Such events could be as simple as a Banker or Player hand or highly complicated specific situations (for example what's the next winning hand after a side had won with a natural 7 vs a drawing hand, etc)

No matter how sophisticated is our approach to select two opposing A and B situations, itlr everything will equalize with the well known unbeatable deviations (burdened with the vig).

Wait.
This is true whether the game is perfectly randomized and it's very difficult to negate that shoes do not present such feature.
Thus in order to try to demonstrate that shoes are not that random, instead of assessing the randomness by statistical tests (chi-square, etc), we should work more empirically, say thinking in more practical terms as it's what really counts.

If I'm able to find out the spots when two opposing situations do not adhere to the common deviations (that is they are more "restricted") I'm on cloud nine.

In fact, there's no way I could spot favourable situations per se, the only hope is to get what I name "limited random walk", a sort of pendulum which moves from the left to the right and vice versa within a restricted range and crossing several times the 0 point.
Since I do not think I'm a genius capable to dispute math laws, the only explanation is that cards distribution of every single shoe couldn't be that random as we think.

Therefore and thanks to my long analysis I dare to state that not every A/B opposing situation will produce the same expected deviations and, more importantly, that not every shoe is playable as some shoes are so polarized at the start that we better get rid off them without betting a dime.
I mean that we can't try to be right on every shoe dealt as many times the possible unrandom effect can't be properly grasped by human eyes.

And this is proven by the fact that no matter how many random walks we wish to set up, a given card distribution will present similar lines on each of them.

Now the question is how to classify a "not playable" shoe.
Next time.

as.

It's important to take track of derived roads (four are displayed right on the screen but you can construct infinite roads).

BEB, SR and CPR are three displayed "derived roads" that are representing three different random walks being a direct reflex of hands distribution.
To simplify the issue a bit, BEB represents a 1-step random walk, SR a 2-step random walk and CPR a 3-step random walk.

It's interesting to notice that the main road (the main BP distribution) will almost always form omogenous distributions on derived roads according to the main road.

The most imprtant parameter to look for is the constant asymmetrical distribution on such four distinct situations.

There are several factors that endorse such assumption.

- one is the general asymmetry of the game

- two, cards are depleted once they are used, so the future deck is always asymmetrical even if we do not know which side will be favored by such asymmetry.

- three, the shoe we're playing at is not a perfect random model by any means.

It's up to us to define and restrict the values and assess the limits of such different random walks.

Say we want to restrict the variance effect thus trying to find the situations when A can't be higher or lower than a -4 or +4 B deviation respectively.
We know that whenever such "cutoff" limits are reached we're playing a 100% edge game.
I mean that whenever A reached the -4 cutoff value (or B the +4 value), our bets could only have a positive expectancy.

Thus the main issue is to find opposite situations where A or B can't produce higher deviations than 4.

Secondly we must approach our strategy in order to get the least deviations, even if we know that an 8-step martingale will get the best of it in any case.

Are we going to bet 256 units to win just 1 unit? Maybe, but it's a worthless and risky effort as the certainty to get an 8-step random walk cannot be achieved by 100% accuracy.

Moreover, we shouldn't forget that a continuos +1  -1 random walk provides us a light loss (vig impact).

So what's the best strategy to adopt?

See u tomorrow.

as.