Game Rules
Learn: Blackjack, Baccarat, Craps, Roulette, Slots, Video Poker, Pai Gow Poker, Texas Hold'em, Caribbean Stud, Let It Ride, Casino War, Big Six, Sic Bo (NEW!)

Systems Gallery
Win more, more often with proven betting methods from the systems gallery

Order Form
A secure on-line order form to purchase products advertised by Let's Talk Winning.

Playing Online
All about gambling at online casinos. Find out where the best deals are

Best Casino Bonuses - NEW!
The best collection of on-line casino bonuses, where you can cash $802 on your sign-up.

Links and Things
Sign our guestbook, refer a friend, link to other gambling sites.

Gambling News - NEW!
Latest Gambling News - updated daily including gambling articles, joke of the day and a gambling dictionary.

Newsletter
Click above to read the current edition of our weekly newsletter. Then get your own free subscription right here. Enter your e-mail address for immediate delivery of an introductory issue, including a FREE ROULETTE SYSTEM by clicking here.

Please tell all your friends about it.

On-Line Forum
Post a gambling question or opinion - get an answer . . .

Below is a selection of the most recommended
on-line casinos
and sportsbooks

Check out:
On-Line Casino of the Month

================================================    Sunday, January 9, 2005
 
Baccarat Bet Selection Design

Hello everyone,

An interesting post was made by Oen at the LTW discussion forum on how to design a bet selection for Baccarat and I'd like to share it with you today.

"There are two types of bet selections: a pattern based and template based bet selection. For the discussion today, I will discuss only the pattern based bet selection.  A pattern based bet selection is a way of bet tracking based on previous or past decisions, while a template based bet selection is  tracking, which is not based on previous decisions and is predetermined and independent on previous results.

There are systems, which are independent of bet selections. They treat bet selection as insignificant. These systems usually follow blindly the progressions.

As you may know, we cannot predict a future outcome. This does not mean, however, that past decisions cannot be useful. Past decisions will form patterns or formations, which we can use as a reference for our bet selections.  We see many times a shoe running choppy, streaky and getting into other formations.

Let's take a look at  the following mappings:

Last 1 decision

There are 4 ways to map for the last 1 decision.  Read: b->b, if the last 1 decision is b, then bet b.

b->b
p->p

b->p
p->b

Last 2 decisions

There are 8 ways to map for the last 2 decisions.

bp->b
bp->p
pb->b
pb->p

pp->b
pp->p
bb->b
bb->p

Last 3 decisions

There are 16 ways to map for last 3 decisions.

bbb->b
bbp->b
bpb->b
bpp->b
bbb->p
bbp->p
bpb->p
bpp->p

pbb->b
pbp->b
ppb->b
ppp->b
pbb->p
pbp->p
ppb->p
ppp->p

Last 4 decisions

There are 32 ways to map for last 4 decisions.

…
bbbb->p
bbpp->b
ppbb->p
…

Last 5 decisions

There are 64 ways to map for last 5 decisions.

…..
bbppb->b
ppbbp->p
…..

If you will, there are 2*power (2,n) ways to map for the last n decisions.

Let me give some examples for classic and commercial bet selections using the above mappings (or rules)

A. Bet p only
1. p->p
2. b->p

B. FTL (Follow The Last)
1. b->b
2. p->p

C. oFTL (opposite to Follow The Last)
1. b->p
2. p->b

D. DBL (Decision Before Last)
1. bp->b
2. pb->p
3. pp->p
4. bb->b

E. oDBL (opposite to Decision Before Last)
1. bp->p
2. pb->b
3. pp->b
4. bb->p

F. Wrangler march
1. Wait for one decision, then
........ 1. b->b
........ 2. p->p
2. Bet the side until two losses in a row, then switch to the other side.
3. ppbb->p (pure ppbb)
4. bbpp->b (pure bbpp)
5. ppbbp->p (pure ppbb)
6. bbppb->b (pure bbpp)

The next step is to analyze the losing and winning streaks of the designed bet selection, take FTL as example.

…111111..., you will have losing string …LLLLLL…..
…121212…, you will have losing and winning sequence …LLWLLWLLW…..
…131313…, you will have losing and winning sequence …LLWWLLWWLLWW…..
…222222…, you will have losing and winning sequence …LWLWLWLW…..
etc.

So, the weakness of FTL is the single chop pattern for example: …bpbpbp…, you will lose every hand.

Beside the rules, which refer to the past decisions, you could certainly add other rules (complex or simple and no one hinders you from doing that) to your designed bet selection (fictive or no bet). These additional rules are called restrictions. You restrict the bet selection because of the losing streak you want to handle.  This is one example why you would want to add additional rules.

For example:

First variation of FTL:
FTL (Follow The Last) with restriction
1. b->b
2. p->p
3. If it runs two losses in a row, then stop betting and resume betting after a win.

Second variation of FTL:
FTL (Follow The Last) with restriction
1. b->b
2. p->p
3. If it runs two losses in a row, then switch to b->p or p->b. After a loss, switch it back.

You can find many ways to implement a restriction.

The next important step is to design the progressions you want to use to accompany the designed bet selection. Progressions can be positive or negative.   Principally try to avoid steep progressions. Flattened progressions or flat bets with a partial parlay are favorable.

The last step is to perform tests. Find out which betting strings form losing and winning bets.

When you use the above for real in a casino, there are a few other important factors you need to consider, such as MM (Money Management). This is not really a progression.   It has much to do with the way you want to handle the money you play with.  Other factors are self control, patience, emotions, attitude, etc.

Another way to design a bet selection refers to how you wish to handle losing streaks.  Please remember that every bet selection has its losing and winning strings, there are no exceptions. The question is how often they occur.  You can find out about this by testing or actual play.

For example that …22222… statistically may happen more often than …44444…..

For example, how can one handle the string: 1111…., ….22222…., and 1212121?

The rule which handles …11111…. is
1. bpb->p
2. pbp->b

The rule which handles …22222…. is
1. ppbb->p (pure ppbb)
2. bbpp->b (pure bbpp)
3. ppbbp->p (pure ppbb)
4. bbppb->b (pure bbpp)

The rule which handles …12121212…. is
1. pbb->p
2. bpp->b
3. pbbp->b
4. bppb->p

Additional rules to handle a streak of 3 are:
1. bbb->b
2. ppp->p

This is only one solution to the above question and you may find others. You only need to use common sense by observing the shoes.

No matter how simple or complicated your designed bet selection is, there will be times you will encounter winning streaks and losing streaks. Stay on simple rather complicated bet selections."

Some comments made on this post:

"One other way that I have been experimenting with is to bet Banker only, since statistically Banker will show more times than Player.

The rules:

When you begin a shoe, wait for a Banker to show, then place your first bet on Bank.
Continue betting on Bank until 3 losses in a row.
After three losses, do not bet until Banker appears again - then resume.
If Banker loses immediately after resumption, stop and begin the process again.

Since you are guaranteed to hit those three losses in a row, I prefer a more flat type bet progression - called Alexambert. You can find the rules for playing Alexambert on the Winn-Maxx.com site (you have to be a member, but that doesn't cost anything so sign up to learn about the progression).

The Alexambert progression I have been using is 1-1-1-4-4-4-7-7-7, etc. (it's more than just an up-as-you-lose, down-as-you-win, so check out the site for the rules).

I don't go beyond 1-1-1-4-4-4-7. Winn-Maxx lists several other variations for you to check out."

"There are a few other bet selection techniques, such as 'anti-streak": waiting for X number of decisions and betting the opposite. I know somebody who is doing extremely well playing a the same "template based" 4 bet combination over and over with different levels of a martingale.

For example BPBB, any win say at $25 units, then the next two bets are watched, should they lose all 4 bets, another 4 step Martingale is used say at $50. Another template based betting technique is a simple cross-over: always bet the opposite of your last bet. This way one can't simply say you will lose against PBPP, you would only lose against PBPP if it occurs when you commence your 4 step betting sequence (independent betting for a game of independent trials).

Cross-over will provide a win "somewhere" ?

1-1-1-1-1 (50% of the time)
1-2-1-2-1
2-2-2-2-2
3+ 3+ 3+

However your MM plays an important role here.  It's not a bad bet selection, considering it only loses against 50% of all chops, so you have now narrowed down the field.  More suitable for Martingale than anything else, or twice your normal playing unit level.

In regards to the 2-1's and 1-2's, MM techniques can turn these from potential losses to pushes playing FTL.

You can also base your game on probability, similar to anti-streak. For the same reason that an entire shoe is not going to chop or produce 2-2 results. You may see the following combination 3-1-3-1-3-1 (a pattern of 12).  This is not going to continue.  It will eventually have to break (we just really don't know when).  You need to chose your own anti-pattern figure, and bet against it.  You continue to bet higher on the anti-pattern figure and the more abstract the pattern the better it is.

The frequency of occurrence is an interesting concept, unless you have an extremely streaky shoe.  The first line is more populated than the second, the second more populated than the third line etc.

Given the above,  if you played FTL only, you lose all first line decisions, and there certainly are more of them.

My advice is that you should never play with one tactic in mind, be DBL, or FTL. You rather need to use a selection according to what the shoe is doing, although this can change in an instance.

An important aspect is what I refer to as "Trending the Shoe".  We can work our strategy around it.  If the shoe is streaky play FTL, if it's choppy play DBL, but a choppy shoe can easily produce a string of 2-2 results, refer to as the 2 chop.

So we need trigger points to change from one tactic to another. Those can be identified by observing the shoe (last few results), or by a run of losing bets. Yes I'm totally in I agreement with the rule restrictions, not forgetting the very useful tactic bet selection known as WATCH, if the trend is hard to gauge.

In regards to the partial parlay and my experience with it, it relies too much on luck , which could be fine for a possible recovery, and it can surely protect your bankroll, but the odds are very much against you considering the frequency of obtaining back to back wins.

I'm all for deep progressions, while they can but you in a mess, they can work wonders, but do they have other requirements such as more than 1 win? Short progressions in my opinion can have you going around in circles and getting you nowhere. The casinos don't mind this, of course.  However with deep progressions, they do place counter measures such as table limits.

My local casino has tables that have a min $10 and $10K table differential. How can you lose with that kind of spread?  That's a 16 step progression using a normal Fibonacci, playing with a $25k bank and playing $10 unit value.  You would die of boredom, but you could do it for a living.

You mention "you use common sense by observing the shoes", while I couldn't agree more. However I wouldn't say the trend of one shoe will be the same as the next after a shuffle. But this statement and action is crucial, because if we are not using a template based bet selection, it should determine our bet selection method."

"The d'Alembert progression can be considered as an arithmetic series:

1-2-3-4-5-6-7-etc. with difference=1

The difference between two immediate steps is 1.

The progression rules say:

1. Start with 1 unit.
2. L -> last bet unit + 1 unit
3. W-> last bet unit ¨C 1 unit (stay 1 unit when reach step 1)

Alexembert progressions are variations of d'Alembert (see www.win-maxx.com for further information). The progression steps vary by the difference between the two immediate steps, which can be 2,3,4,.

A. 1-3-5-7-9-11-13-etc. with difference=2

B. 1-4-7-10-13-16-19-etc. with difference=3

C. 1-5-9-13-17-21-25-etc. with difference=4

The progress rules state:

1. Start with 1 unit
2. Number of losses = Difference -> last bet unit + Difference
3. Number of wins = Difference -> last bet unit - Difference (stay 1 unit when reach step 1)

We bet each step up to "Difference± losses or Difference ± wins". Its intention is to bet lower bet units during losing streaks and higher bet units during winning streaks.

Ernie's bet selection:

1. Wait for a BANKER to show:
1. b->b
2. Keep betting until 3 losses in a row, then stop betting and resume betting after a BANKER appears.
3. If BANKER loses immediately after resumption, stop and begin the process again.

The variation B. 1-4-7-10-13-16-19-etc. is used for the above bet selection.

Let me analyze the weakness of the bet selection for the worst cases:

111111 ­ you have the L and W strings  ­.. LWLWLWLW­
121212 ­ you have the L and W strings  ­.. LLWLLWLLW­
131313 ­ you have the L and W strings  ­.. LLLWLLLWLLLW­
141414 ­ you have the L and W strings  ­.. LLLLWLLLLWLLLLW­
222222 ­ you have the L and W strings  ­.. WLLWWLLWWLL­
232323 ­ you have the L and W strings  ­.. WLLLWLLLWLLL­
242424 ­ you have the L and W strings  ­.. WLLLLWLLLLWLLLL­

You will encounter losing streaks when you shoes running like this:

B
PP
B
PP
BB
PP
B
PPPPP
BB
PPP
BB
PPPP
B
PPP
B
PP


It means that you see more PLAYER than BANKER within section of a shoe. I checked the Zumma and found out there are many shoes running similar like this.

Certainly what you can do is that stop betting, change table or bet the other dominant side with the same bet selection. If you are confident that by statistics say that BANKER will show more times than PLAYER, you should bet the way you bet. I personally also saw that shoes ran more BANKER than PLAYER. Once I saw that after half of the shoe have been dealt there are only few PLAYER and many and many of BANKER. Further what you can do is to qualify the table first for a few hands whether the table will show more BANKER or PLAYER before betting."

"Actually I plan to post separately the discussion of how to design a template based bet selection. You are right that there other bet selection techniques such as template based and others (such as systems depending on MM only) not classified as pattern and template based bet selection.  I will outline a few, which you have mentioned.

For example:
1. Banjo's template
2. Predetermined bets (like Izak's "How to make 10 units in even money bets" system)
3. bpbpbpbp­. (zig-zag bet no matter what will happen)
4. bbbbbbbb­. (one-side bet no matter what will happen)
5. Coin tossing (H=b, T=p)
6. n-figures betting (for example: bb-bp-pb-pp, n=2 or ppb for n=3 ¡­)
7. Anti-pattern (for example waiting for a streak of n, than bet the opposite)
etc.

As long as that the bet tracking does not refer to past decisions, then this tracking will be categorized as template based bet selection.

Let me outline Cross-over bet selection. Cross-over can be considered as zig-zag template based or pattern based depending on the first bet selection and thereafter.

Template based Cross-over selection; LB=Last Bet:
1. ->p or ->b
2. p(LB)->b
3. b(LB)->p

[p(LB)->b will be read as if Last Bet is p, then bet b]

Pattern based Cross-over selection; LB=Last Bet:
1. Wait for the first decision:
.... 1. b->p
.... 2. p->b
2. p(LB)->b
3. b(LB)->p

As to the Anti-Pattern, which can be treated as pattern based or template based bet selection. It depends on how you see the decisions and the template.

You are right in regards to progression that you could make living when you use deep progression. But on the other hand, I think that a few gamblers will use a deep progression. 

Basically I would say if you could design a good and proper bet selection then you can be sure that you can enlighten the MM/progression you are using."

"So far I have discussed pattern based bet selections with the reference of past decisions. Now I am going to touch on statistical recordings based on counting and the calculation of the p and b appearances or occurrences, which can definitely be used to chose a proper and  bet selection.

---Power(2,n) choices---
Let me calculate the number of shoes if we play n numbers of hands.

If there is only one hand to play, then we have two choices of decisions:

B
P

If there are two hands to play, then we have four choices:

BB
BP
PB
PP

If there are four hands to play, then we have eight choices:

BBB
BBP
BPB
BPP
PBB
PBP
PPB
PPP

If there are four hands to play, then we have sixteen choices:

BBBB
BBBP
BBPB
BBPP
BPBB
BPBP
BPPB
BPPP
PBBB
PBBP
PBPB
PBPP
PPBB
PPBP
PPPB
PPPP

If there are five hands to play, then we have thirty two choices.

So, if there are n (>=1) hands to play, then we have power(2,n) choices.

Let's assume that a Baccarat session will end by approximately more or less within 60 hands. So, if there are 60 hands to play, then we have power(2,60) = 1.152.921.504.606.850.000 combinations of decisions.

---Law of Series---
The distribution of decisions can be grouped into series. There are series of 1 or also called as singles, series of 2, series of 3 etc. The law of series appearance (Binomial distribution) can be expressed as the following:

Single appearances will hit twice as many as series of 2.
Series of 2 appearances will hit twice as many as series of 3.
Series of 3 appearances will hit twice as many as series of 4.
Series of 4 appearances will hit twice as many as series of 5.
Series of 5 appearances will hit twice as many as series of 6.
Series of 6 appearances will hit twice as many as series of 7.
.
.
.
Series of n appearances will hit twice as many as series of n+1.

For example:

SERIES.............. PLAY......... %

1 series of 7......7 plays ..... 0.78
2 series of 6.....12 plays .....1.57
4 series of 5......20 plays.....3.15
8 series of 4......32 plays.... 6.30
16 series of 3....48 plays.. 12.59
32 series of 2... 64 plays...25.19
64 series of 1... 64 plays.. 50.39

TOTAL 127 series... 247 plays...100.00

---Law of even distribution--- equilibrium/equi-partition---
The law of even distribution states that p and b's are distributed evenly on the long term when a large number of hands is dealt.

---Deviation of Equipartition---
Deviation of Equipartition expresses that in the short term b's will be dominant over p's or p's will be dominant over b's.

Let me show you an example for the application of the Law of series to bet selection:

1. Bet singles
2. If singles bet lost, bet series of 2
3. If series of 2 lost, then stop betting
4. Resume betting after the series is broken.

Or
1. b->p
2. p->b
3. bb->p
4. pp->b
5. (bbb¡­)->no bet
6. (ppp¡­)->no bet
7. (bbb¡­p)->b
8. (ppp¡­b)->p

With the above background, we can raise the following questions.

How many p's and b's (1-figure) are dealt in one shoe?
How many pp, pb, bp, and bb's (2-figures) are dealt in one shoe?
How many ppp, ppb, pbp, pbb, bpp, bpb, bbp and bbb (3-figures) are dealt in one shoe?

How many singles, series of 2, series of 3, series of 4 etc. are dealt in one shoe?

How many wins and losses are dealt in one shoe?

What is the difference or deviation between p's and b's (1-figure) dealt in one shoe?
You can ask the same question for 2-figures, 3-figures etc.

What is the distance between p and b (1-figure) dealt in one shoe?
You can ask the same question for 2-figure, 3-figure etc.

The list of questions can get longer and longer.

By observing statistical records,  you will be able to find out if the shoe ran choppy, streaky, mixed, p-dominant or b-dominant, etc.  With this information on hand,  you will be able hopefully to chose a proper bet selection and design a bet selection based on the statistics. For higher figure play you will be able to see how each of the groups is developed.

Let me show you a small example for 1-figure (p and b) by recording the number of p's and b's being dealt and calculate the difference between p and b.

#-....decision...p.......b.....(p-b)
1........p............1........0.......1
2........p.............2.......0.......2
3........b.............2.......1.......1
4........b.............2.......2.......0
5........b.............2.......3.......1
6........b.............2.......4.......2
7........b.............2.......5.......3
8........p.............3.......5.......2
9........b.............3.......6.......3
10.......p............4.......6.......2
11.......b............4.......7.......3
12.......p............5.......7.......2
13.......p............6.......7.......1
14.......b............6.......8.......2
15.......b............6.......9.......3

# is the number of hands being dealt
p is the count of the p occurrences
b is the count of the b occurrences
(p-b) is the difference between the occurrences of b and p

P
4 :
3 :
2 : *
1 :* *
0 :-------*------------------------------
-1 : * *
-2 : * * * * *
-3 : * * * *
-4 :
B

You will see how b's and p's develop whether it will be choppy, streaky, p-dominant or b-dominant. You can even chart the (p-b) and see the wave formation. After recoding a few hands you will be able to choose a proper bet selection. So, the statistical recording serves as a tool for you to choose a bet selection.

For example:

1. (b-p) is absolute increased ->b
2. (b-p) is increased ->p
3. (b-p) is equal ->no bet

Let me show you another example for 1-figure (p and b) by recording the p's and b's being dealt and calculate the distance between p and/or b.

#....decision....pd/bd
1.........p..............1
2.........p..............1
3.........b..............1
4........ b..............1
5.........b..............1
6.........b..............1
7.........b..............1
8.........p..............5
9.........b..............2
10.......p..............2
11...... b..............2
12.......p..............2
13.......p..............1
14.......b..............3
15.......b..............1

# is the numbering of hands being dealt
decision is either p or b
pd/bd is the relative distance between two immediate occurrences of b or p

The statistical recording can prepared as well for other questions, for example 2-figure (pp, pb, bp and bb) or single (p,b), 2-series (pp, bb), 3-series (ppp,bbb) etc.

So what I have outlined is the use of statistical recording of past decisions as a tool for choosing a proper bet selection or even can be used for designing a new bet selection. You can do it and it is only with simple math."

Thanking Oen for his useful post.

Wishing you all the best,
Until next week,
Izak

If you missed any newsletter, click here for an archive.