December sunrise from my house

Quote from: alrelax on February 06, 2024, 03:15:52 AMLast winter in Kansas City, just a totally painted sky in the early morning.

Terrific!!!!

as.

BTW: try to spot two different mechanical random walks getting a slight different pace capable to exploit the most asymmetrical bac features and you'll get one of the solution to get a sure indeniable long term edge.

We'll see how a multilayered progressive plan will get the best of it.

as.

Betting plan

Baccarat world is made by infinite situations that we could restrict by considering streaks.

More hands we'd try to guess lesser will be our profits, or better sayed, greater will be our losses.
There's no fkng exception about that, believe us.

Mathematicians and 'so called' gambling experts will say to us that no matter how's diluted our betting, the EV remains negative yesterday, now and in the future.

This is a complete false statement as it considers any shoe dealt springing from the same shuffling source.
We've seen that it makes quite a difference when a shoe is shuffled by a machine or manually or by "only gambling gods know" preordered shuffles.
The 'cut' made by players (when allowed) isn't a proper tool to break the shoe sequences, it just postpones them.

Hence we must be prepared to deal with sd values (at both positive and negative ways) and machine sd values are way better controllable than other shuffling procedures.
Next are manually shuffled shoes and at the bottom of the list come preordered shuffled shoes where we do not get any hint about how they were shuffled.

1) How many 5/5+ streaks are we expecting from any shoe dealt?

We strongly think that percentages presented above are accurate, we do not care less about millions of shoes dealt especially when a homogeneous source wasn't considered (so you could put in the trash every data you've collected unless coming out from the SAME source). In the same way we shouldn't give a fk about B/P long term raw percentages as sure as hell they will be approaching more and more the 0.5068/0.4932 useless ratio.

We need to assess the average streaks distribution after having devised one or more proper random walks.

After having classified every "long streak" happening as belonging to the 5/5+ category, the positional distribution of them is paramount as way more than half of the times we'll expect zero or one or two 5/5+ streaks happening along the entire shoe; obviously more such long streaks happen at the start or intermediate parts of the shoe greater will be the probability to get a shoe surpassing the inferior than three/superior than three streaks number ratio.

Remember that per every seven shoes dealt on average you'll get ZERO shoes presenting a 5/5+ streak, and at the worst situation you're almost 3:1 favorite NOT to get three or more 5/5+ streaks.

More simply, well more than half of the shoes dealt will include just one or two 5/5+ streaks whatever distributed.

Best scenarios and worst scenarios

Well, best scenarios are when zero 5/5+ streaks will happen along the entire shoe (nearly 13.8% of the times), worst scenarios are when the "four" cutoff streaks number  will be reached or surpassed (say it's  about a 5% value or so).

So we're nearly 2.77:1 favorite to get zero 5/5+ streaks than having four or more 5/5+ streaks at any shoe dealt.

Problems arise when we have to decide the more valuable spots to deny such possible 5/5+ streaks

We'll see this issue in a couple of days.

as.

Here another set of thousands of live shoes dealt by preordered shuffling (same HS room); notice that the previous shoes data were extracted by a Shuffle Master Machine: 

Zero 5/5+ streak account for 14.51%

One 5/5+ streak account for 28.49%

Two 5/5+ streaks account for 29.56%

Three 5/5+ streaks account for 19.89%

Four or more 5/5+ streaks account for 7.52%

Comparing those two sets of data we see that at the second sample one and two 5/5+ streak shoes concede more room to three and four(4+) streak shoes.
In addition at both data zero 5/5+ streak shoes percentage is almost corresponding and one and two streaks categories keep being distributed "even money" between themselves.

IMO streaks lenght and average distribution are the best indicators of the binomial model movements affected by a kind of asymmetry (bac rules, key cards impact acting at finite sequences, etc).

So for example at the first sample we'll get an average 82.63% cumulative probability to cross zero/one/two 5/5+ streaks per shoe; at the second sample the same probability decreases to 72.56%.

At any rate no matter the shuffling procedure employed, shoes presenting one and two 5/5+ streaks account for respectively 69.46% and 58.05% of total shoes.
At both data the zero 5/5+ streaks percentage (13.17% and 14.51%) acts as a kind of constant "bonus".

It's quite obvious to deduce that we shouldn't really interested to guess what precise patterns will show up along any shoe dealt (so there's no point to bet against streaks formation around any corner), just to assess the more likely classes of precise streaks distribution happening per every shoe dealt. Thus being able to understand the acute damage (or favor) 5/5+ streaks will get to us on average and in important relationship with the actual shuffling procedure.

Those are just the basics, we need a plan to try to get the best of it.

See you later

as.

Suppose we want to adopt a more risky progressive plan, one which can get us innumerable wins with very very low or insignificant levels of probability of losing the entire bankroll.

IMO first we should consider random walks producing many "low" or "moderate" streaks, so getting rid of the singles production thus considered as neutral events.
Then and even though long term statistical findings teach us that doubles are slight more likely to happen than any other pattern, only the proper assessment of the actual shoe dealt will help us to find what and when to bet.

Casinos prosper about the unlikelihood that players will guess right for long by ratios surpassing 51.2% at B bets and 50% at P bets.
A progressive plan doesn't shift such ratios in our favor, unless the average streaks distribution provides low sd values, up to a 4 or 4.5 sigma (e.g. a 16 or 20 B or P streak or any other unlikely proportional deviation).
And providing the use of proper random walks, it fkng does.

If the time works for casinos, let the time be working for us.

Basics

How many 5/5+ streaks per shoe are going to be produced on average when the main random walk is acting?

Following data come from thousands of real live shoes dealt at MonteCarlo casino where 8 decks are utilized and almost two decks are cut off from the play.   

Zero 5/5+ streak account for 13.17%

One 5/5+ streak account for 34.73%

Two 5/5+ streaks account for 34.73%

Three 5/5+ streaks account for 14.97%

Four or more 5/5+ streaks account for 2.39%

Since per every shoe sample made of nearly 60 hands dealt we'll expect on average a slight lesser amount than four 5(5+) streaks percentage (after 64 hands it's 3.125%, so now more than that), we know we're getting an edge as the average statistical findings seem to strongly deny such expected value.

Obviously in practical terms it makes a lot of difference if we'd start up to bet against the 5/5+ streaks happening after a new column is filled than wagering after a given streak shows up.
We'll see this issue next.

as.

Hi KFB!

Say that in terms of different patterns extracted by infinite random walks the less likely situation happening for each shoe is a balanced or close to balanced ratio, obviously influenced by the various portions of it.
After all, gambling is a game of clusters at either way, meaning that what we have won at positive (clustered) situations will be easily lost at negative clustered sequences.

The trick is to adopt random walks capable to shorten the streaks lenght as, generally speaking, baccarat is a game where streaks are better defined in their average appearance than a 50/50 independent model along any shoe dealt (thus considering an infinite succession of 72-78 resolved hands).

Q1/A1: I wouldn't bet ever toward 5/5+ streaks as our two random walks eliciting the algos action   find very few occasions to expect them. As an interesting part of shoes do not present any such (long) streak.
So in such instances any streak is valuable to be attacked up to the point that even 4s sometimes do not happen for that shoe.

It's true that moderate streaks (as 3s or 4s) coming out consecutively (so without any inferior pattern intertwined) make a relative more room to expect a longer streak (a 5/5+ streak).
Sooner or later some shoes must come out by filling a way shorter than average number of columns and this can only happen by the shoe producing long streaks.

Q2/A2: Good question.
By far the best probability to get 5 or more wins in a row is by "hoping" that A or B or both will take a uniform clustered direction (singles in a row, singles/streaks or streaks/singles, etc).

More later

as.

Independently of the random walk utilized, longer streaks (say 5-5+) tend to dispose themselves by three different shapes:

a) Singularly interspersed between lower degree patterns (singles/doubles/triples and 4s)

b) Diluted or not showing up at all

c) Clustered (back to back) at various levels of consecutiveness and density.

Each shape of presentation has its merits with important practical reflexes.

Shape a) is the more likely course of presentation and could be applied at lower streak levels (4s or 3s).
Actually a (risky and unnecessary) multilayered progressive plan adopted to get 5s, 4s or 3s showing up as 'isolated' cannot be wrong for long as the isolated/clustered streak ratio tends to produce low sd values.
Problem is that by adopting this strategy we rely upon a "general" probability that could be voluntarily (virtually) altered once casinos know what we're doing.

Shape b) is not so rare to happen as we'll face more shoes presenting zero or one/two 5-5+ streaks and even 4 streaks than shoes producing a way larger number of expected 5-5+s.

Shape c) is very interesting as generally speaking, clustered 5-5+ streaks will make more room to inferior streak classes to happen in the next portions of the shoe.
Notice that whenever a 5-5+ streak cluster shows up, we're somewhat discontinuing a constant losing betting strategy as what we should interested in is to approximate the streak "ranges" per any shoe dealt getting rid now of the clustering effect (so we need a different event or a couple of different events to start or restart the r.w. action.

To get an idea about how our two algorithms work, instead of considering the filled BP spots (or every other either/or succession whatever intended) think about the empty areas shapes limited by the different columns lenght and rows interruptions.
Obviously deeper we go down in assessing such empty slots, greater will be the probability to get consecutive rows not filled by actual hands. Always considered by empty 'ranges'.

Even though algorithms work by numbers (probability after effects), with some practice such propensity could be geometrically ascertained up to the point that even a random betting at the proper times will pick up more wins than losses.

Someway algorithms base their action upon a ultra selected "negative" multishaped probability oriented to get something heterogeneous at one side and homogeneous at the other one.
At the end suggesting bets in order to get empty 'ranges' and not filled slots.

as.

Hi Al!

Actually there are many mechanical approaches getting a very slight advantage over the house but to really succeed they need several parameters to converge into univocal betting spots.
And such parameters are more in direct relationship of the actual shoe production than about long term statistical findings.

Long statistical data help us to understand the more likely random walks movements roaming or distancing from a 0 starting point.
Obviously most part of them consider small or moderate steps at either left or right direction, an issue best studied by "streaks" lenght (widely intended as a 6 or 7 chopping line is a streak).

Then even such streaks may be classified into "isolated" streaks, two back to back streaks (cluster of two), three streaks (cluster of three) and so on.
Naturally any specific streak will fight against superior patterns (that is for doubles singles are ininfluent, for triples doubles are ininfluent and so on)

That's now that the actual shoe production becomes the main succession to be interested upon as each card distribution is asymmetrically shaped by definition and itlr everything will be equaled (or close to it).

If a distribution is asymmetrical even streak clusters will be asymmetrically grouped even though we do not know which streak classes will be clustered unless they came out at least one time.
The trick to consider two specific streak classes simply facilitates the problem as now we need each class to show up at least one time before making them to fight vs opposite streaks.

The beauty of baccarat is that streaks groups cannot show up isolated for long or, better sayed, that every shoe dealt in the universe is virtually destined to form at least one streaks cluster.
Sometimes such streak clusters are so long to prolong for the entire shoe and of course longer the streak classes considered, greater will be the probability to not cross an unfavourable streak not belonging to the classes wagered.

More later

as.   

Al, I believe you, wagering small sums between serious bets is a perfect reasonable way not to get any heat from casinos.

"Successful players" I was referring to are people who like to get the best of it on every cent they put the money at.
Mostly I'm talking about Vegas players who are well aware that casinos consider baccarat as a unbeatable game, so they do not need to camouflage their action being labeled as worthless.

Technically I'm deadly sure that unless a huge betting spread is adopted, wagering 12-15 bets per shoe strongly reduces/erase/invert the EV+ and actually such people is interested to know that some "rare" spots could be EV+.

I mean that it's almost impossible to convince a part of HS no-bac players that baccarat is beatable by being more right than wrong at 12-15 hands bet per shoe.
For the remaining part (best baccarat scholars), the answer is a sure "no", an answer we take as a papal bull, confirmed by our findings.

Several times I've publicly invited youtube geniuses, system sellers and forums "I can't lose" claimers to mentor some HS people gladly accepting to concede a percentage cut of the winnings (worth thousands of $$), but with the disturbing downside that to avoid consequences such "winning players" must be right beyond any shadow of doubt (say capable to overcome a possible 4 or 4.5 negative sd value).

So far, nobody accepted the challenge but us.
Let me know if some "Banker is the best bet no matter what" or "I got a 53% winning rate" or other fkng bullsh.it claimers want to get a formidable freerolling on their money (but only at their money).

as. 

Quote from: KungFuBac on January 18, 2024, 03:01:02 PMAsymBacGuy at the bottom of post #963: "...Now you might wonder why some "uninterested" players will patiently wait then betting 10k-20k after a couple of consecutive 1s came out in order to get any number different than 1..
.."
Like

Hi KFB! 

Actually not every successful player will bet as huge as 10-20k per hand, surely all successful players will bet very few hands per shoe, the best being just one-two (or zero) bets per shoe.

In fact, more bets we'll place, greater will be the probability to fall directly into the EV- world as the profitable spots are quite rare to happen.

Technically those spots are 'convergence of probability' situations where different parameters belonging to a given random walk do converge to a more probable statistical line.

It's like using a kind of "score" by assigning numbers to specific r.w. patterns. Whenever for each pattern the score is negative or neutral or too light positively deviated, we'll simply do not bet (even though we would have been winners). In the only other scenario a bet is suggested by the algorithm.

It's obvious that everything revolves around the clustering effect of various patterns elicited by the same BP succession.
More clusters of something are happening (that is more numbers different than 1 are showing up) dictating a same hand wager, better will be the probability to catch "prolonging" clusters.
In poor words we'are just "approximating" that clusters won't stop right at the spot considered worthwhile, always setting up the "minimum" profitable value, that is one step going forward.

After having reached that minimum cutoff point, we should not be interested to bet anymore as we will simply find ourselves in the position of gambling.
Such "gambling" attitude, providing to assess carefully the starting points of the clustering effect, could be a (minor) viable tool for those who are capable to manage a positive betting plan without risking to hope for too long positive sequences erasing previous losses.

Anyway if you'd measure large shoes data (enduring almost every possible variance deviation) you'll see that it's way better to win few spots than chasing (challenging) the inevitable short positive successions typical of baccarat.

More later

as.

Taking the same shoe sample let's see how third row gaps went (practically those are single/double streaks sequences):

2-2-1-1-9-5-(2)
1-4-7-5
11-1-2-6
1-4-6-10
4-14-6-5
4-1-1-5-4
1-11-3-2
2-5-3-1-1
6-3-9
1-2-7-13-1
4-1-2-1-6-1-1
1-11-3-4
4-1-4-3-1-4-3
3-2-3-8
2-2-9-5-4-(2)
2-13-1-1-(2)
14-5
2-1-4-2-1-1-4
3-7-6-3-4-6
1-2-2-9-1-5
6-1-1-3-6-5
3-5-1-15
1-6-1-2-8-3-3-4
3-8-7-5-3-3
1-10-5-1-2-5-3
4-4-1-4-4-2-(3)
3-4-6-2-4-1-6
1-4-4-5-10-1
1-1-2-2-4-1-3-3
4-2-1-4-5-1-4-1
5-2-15-2-1-4
5-4-2-2-13-4
9-3-10-1-8
2-2-4-7-3-1-1-2
2-1-4-6-2-5
2-1-1-4-2-2-2-1
1-2-2-13-1-1-4-2
4-8-5-11
7-2-2-13-1-5
12-1-3-2-8-1
1-11-3-4-5
10-1-1-5
9-5-1-8-2-4
3-1-16
3-4-2-3-1-5-8
1-7-1-2-1-1-6
3-2-6-1-17
4-2-1-6-3-3-(3)
1-4-5-4-1-3-5
3-2-5-5-2-1
1-4-2-5-16
2-4-1-3-1-13
10-6-9
4-1-9-1-9
2-13-8-7
2-1-1-5-2-5-5
1-1-3-1-2-1-4-5
8-6-6-2-5
4-2-1-1-1-1-1-6
16-2-2-5-6
-2-3-2-5-8-4
2-1-2-2-1-7-1
1-2-2-7-16
4-11-4-5-8
6-4-3-1
7-10-7-9
9-5-3-5
1-2-2-3-6-1-5
2-2-4-3-1-4-1
1-1-2-8-1-4-2-3
3-3-3-19
7-2-2-3-2-1-3
3-5-1-4-9-7-(2)
7-9-3-6-1
16-4-4-1
1-1-5-3-3-4
3-12-4
2-5-3-7-1-3-1-2
4-1-1-5-2-1-4
2-7-3-8-2-2
9-13-5-1
1-1-2-2-3-3-3-3
3-2-10-3-5
2-6-1-4-2
4-6-6-2-2-1-(2)
8-1-4-2-4-4-4
1-1-3-1-2-4-4
4-1-1-2-3-1-1-3
4-4-7-4
1-1-4-5-1-1-3-2-1-3
1-1-5-2-5-2-5
2-5-1-2-4-4
7-3-3-4-8-3
2-4-3-2-3-2-1-2
3-1-2-3-1-1-4
7-1-1-4-4
4-4-3-1-1-1-3-4
2-9-3-4-4-(2)
1-22-2-1-6
4-5-3-1-2-6
5-2-5-6-6-4
6-3-1-3-7-(2)
2-1-4-1-9-5-4
7-3-2-1-2-4
2-1-7-2-2-1-(2)
3-1-2-1-1-4-5-2
1-3-1-4-2-1-2-10
6-1-3-5-5-10
2-7-7-3-2
4-8-5-8
3-2-3-2-1-6-6
6-4-1-5-1-4-2-(3)
2-1-3-4-2-8-1
6-4-2-3-2-1-3
10-1-4-18
2-3-6-5-1-1-1

Now you might wonder why some "uninterested" players will patiently wait then betting 10k-20k after a couple of consecutive 1s came out in order to get any number different than 1.
Are they gambling?

as.

Streaks gaps and the permutations issue rule algos action

Consider an A/B finite sequence (e.g. 12 steps) where (p)A = 0.75 and (p)B=0.25.

An 'ideal' distribution would be something like AAABAAABAAAB..etc
Actually this is the least likely scenario to get through as a perfect balanced A/B ratio came out, furthermore by a "perfect" pace.
In this case we are not worried about the HE as we know that the probability of success will be so defined that even if our winning bets we'll be payed 0.50:1, we'll crush the game.

Notice that here we're not trying to get a kind of positive variance as W=L.

Problems arise when:

a) The FINAL A/B ratio deviates too much from the expected values at either side (for example 2:1 or 2:2, etc) and obviously we do not know whether A or B are priviliged to show up;

b) Despite of a decent (low sd) final A/B ratio, temporary A (or B) situations are so clustered or so whimsically distributed that we won't know what and when to bet.

Whereas a) factor can't be resolved other than from a long term statistical point of view (huge variance), b) factor could be easily evaluated by "more probable" lines where streaks lenght make a decisive role in the betting options.

One of the main tools our algos rely upon is the probability any shoe dealt will fill (or not) certain rows (that is the streaks lenght) but always by "gaps", meaning that consecutive streaks filling second or third rows are not included in the process as no gaps could be classified here.

If you'd consider geometrically second, third and so on rows BY GAPS, you'll see that empty slots will be more and more asymmetrically shaped up to the point that it's a child's joke to "guess" when consecutive empty slots will take place, almost always by a different shape happened so far.

Providing a proper random walk, of course.

In fact every row is in direct relationship of the long streaks probability: further we consider rows, greater will be the probability to get higher than 1 gaps, at the same time getting rid of the consecutive long streaks making no room to form gaps.

More later.

as.

Thanks KFB!

Back to the 3-4 streak clusters (We've already classified 2-3 and 2-4 streak clusters).

L-L-L
L-W
W
W-W
W
L-W
W
W-L
W
W
W-W
W-W
L-W
W-W
W-L
W
W-W
L-W-W
L-W-W
W
L-W-W
W-L
L-W
W
W
W-W
W-W
W-W
W
W
W-W-W
W
L-W
W
W-L
W-L-W
W-W-W
W-W
L-W
W-W
W
L-L-W
W
W
W-W
W
W-W
W
W-W
W
W-L
W-W
L-W
W-L
W-W
W-L-L
W-W
W
W
L-W
L-W
W-W
W-W
W-L-W
L-W-W
W-W
L-W-W
W-W
L-L
L-W
W-W
L-W
L-W
W
W
W-W
W-L-L
W-L-L
W-W
W
W
W-W
W-W
L-W
W
W-W
W-W
W-L
W-L
W
W
W-W-W
W
L-W
L-L
W
W-W
W
L-W
L-W
L-W
W-W
W
W
W
W-W
W

W=147 L=47 (x3=141)

Nothing extraordinary, even though single Ws include several back-to-back wins and no loss (an obvious consideration happening at other streak clusters).
Out of 107 shoes, single Ws = 35 (nearly 1/3 of the total outcomes).

Comparing this small sample with 2/3 and 2/4 streaks we got:

2/3 streaks: W=249, L=92 (x3=276)  -27

2/4 streaks: W=188, L=41 (x3=123)  +65

Of course those data could easily change with other samples, the common denominator is that 5(5+) streaks remain less likely than what a binomial model dictates.
More importantly is the fact that doubles (2s) are more likely to produce a way higher amount of winning spots even if they could endure the variance.

Then, if a slight (but important) propensity is going to happen, the only obstacle we have to overcome is the permutation issue , best assessed by the old clustering effect.

After all, what happened could repeat once or more times but if it didn't happen so far it just remains in the 'potential world' we shouldn't give a fk about.
And this thing keep showing up at the various levels of clustering probability as 'isolated' patterns are the less likely to happen.

More later

as.

Now let's consider the same CSM sample from a specific lenght clusters factor.

Algo #1 random walk will suggest a bet towards 2-3 streaks clusters vs 2 or 3 streaks isolated situations.

W-W-W
W-W-W
W-L-W
W-L-W-W
W-W-L
L-L-W-W
L-W-W
W-W-W
W-W-L-W
L-W-L-W
W-W-W
W-L-W
W-L-W
W-W-W
W-L-W
L-W-L
W-W
W-W-W
W-W-L-W-L
L-L-W
L-L-W-W-W
W-W-L
W-W-W-L
W-L-W
W-W-W
W-L-W
W-W-W
W-W-W
W-W
W-W
L-W-W
W-W
L-L-W-W-L
W-L
W-W-W
W-L-W-W
W-W-W-W
W-W-W
W-L-W-W
W-L
W-W-W
W-W-W
L-L-W
W-W
L-W-W
L-W-L-W
W-L-L-W
W-W-W
W-W-W
L-W-W
W-W-W
L-W-W-W
L-W-W-W
W-W
W-W-W
W-L-W-L-L-W
L-L-L-W-W
W-W
W-W
W-L-L-W
L-L-L-W
W-W-W
L-W-W
W-W-W-W
W-W-L-W
W-W
W-L-W-W
W-W-W-W
W
W-W-W
W-W-L-W
W-W-L-W
W-L-L-W
W-W-L
W-W
W-L-W-L
W-L-L-W
W-L-W
W-W-W-W-W
W-W-W
W-W-W
W-L-L-W
W-W-W
W-W-W-W
L-W-W-W
W-L
W-W
L-L-W-W
W-W-W
L-L-W-W
W
W-W-W-L-W
W-W
L-W-L-L
L-L-L-L-W
L-W-W
W-W
L-L-L-W
W-W
W-W-L
L-W
W-L-L-L
W-W
W-W
W
W-W
W-L-W

Now the same strategy using 2 s and 4s (3s considered irrelevant)

W
W
W-W
W-W
W
W-W
W
L-W-L
W-W
L-W
W-W
L-W
L-W
L-W-W
W-W
W-L-L-W
W
W
W-W
W-L-W
W-W
W-W
W
W-W
L-W-W
W-W
W-W-W
W
W
W-W
W-L-L
W-W
W
W
W-W
L-W
W-W
W-W
L-L-W
W-W-W
L-W-W-W
W-W
L-W-W
W-W
W-W
W-W-W
W-L-W
W-L-W
W
W-W
W-W-W-L-W-W
L-W
W-W
W-W-W
W-W
W-W
W
W-W
W-W
W-W
W
W-W
W
W-W
W-L-W
W
W-W
W-W
L-W-W
W-W-L-W
W-W-W
W-W
W
W
W
L-W-W
W
W
W
W-W
W-W
W-W
W
W-W
W-L-W
W-W-W
W-W
L-W-W
L-L-W
W-L-W
W-W-W
W-W
W
W-W
W-W
L-L-W
W-W
L-L-W-W
W-W
W-W
L-W-L-W
W-W
W-L-W
W-L
L-W
W-W
W-W-L
W-W-W
W
W-W
L-W-W

Now 3s and 4s (next)

as.