Alembert-Martingale Hybrid

[Bigbroben]

Hi all,

I've been working last weeks on a type alembert-martingale hybrid that's been not so bad, so far.

Allright, let's see:

This applies to even chances.

Let's compare Alembert and Mart for the following outcomes, betting Black:
Outcome:    R    R    B    R    R    B    B    R    B    R    B    B    B 
Alembert:   -1   -2  +3    -2  -3   +4  +3   -2   +3  -2   +3   +2  +1
Cumul:       -1    -3   0    -2  -5   -1    2     0    3    1    4     6     7

Marting:      -1  -2   +4   -1   -2  +4  +1   -1   +2   -1   +2   +1  +1
Cumul:        -1  -3    1     0    -2   2   3     2     4     3    5    6     7

So, even for Alembert and Martingale on this one.
   
Let'S hybrid these to to create a mix, where doubling when losing, but going down a level when winning, instead of going back to 1:
Outcome:    R    R    B    R    R    B    B    R    B    R    B    B    B
Hybrid:        -1  -2   +4   -2  -4   +8  +4   -2   +4   -2  +4   +2  +1
Cumul:        -1   -3    1   -1   -5    3   7     5    9     7    11   13  14

Or again, still betting on black:
Outcome:    R    R    R    R    R    B    B    R    B    R    B    B    B
Hybrid:        -1  -2   -4   -8  -16 +32 +16 -8  +16 -8 +16 +8  +4
Cumul:        -1   -3  -7  -15 -31    1    17  9    25  17  33  41  45

A pure Martingale would have given a score of 6.


So, we see using up and downs as a d'Alembert, combined with a Martingale progression brings higher rewards.

True, there is more risk: a losing streak, say 8 in a row, could bring you from level 3 to 11, instead of 1 to 8, risking your bankroll or table limit.

Outcome:  R    R    R   B    B    R    R   R   B      R      B      R      R    R    R   R   R   R  R
Bet:            1    2    4   8     4     2    4   8   16     8     16    8      16    32  64......

Advantage of using a factor of 2: losses are cut as soon as the first win, allowing decision of player to go on in the progression or coming straight back to 1, or any other lower level.

This system, though, allows for various progressions.  For who said it had to be exponent 2, as in Martingale?!?

I like using a factor of 1.4 or 1.5, even Fibonacci, which is a 1,6xxxx factor.  I could also use a 1.3 factor to make sure I can deal with progressions of up to the 20s, 30s, which happens, on a 200+ spin game.

I'll update as developments happen, of course.

Comments?

Attached: example of a game.

[DoctorSudoku]

I am not a big fan of long negative progressions, but this is still interesting stuff. Thanks for posting it.

[Kav]

Interesting approach, combining progressions.

The problem with combining D'Alembert with Martingale is that both are doing fine in choppy spin sequences and are doing poorly with long losing streaks.

Try combining a progression that performs nicely with short streaks with one that performs good with long streaks like Oscars Grind.

[Bigbroben]

Well, this being said, I never play with a factor of 2, unless for a quick game and a stop at losing the 512 mark ( which isn't too bad when playing with pennies).

I'd go 1 1 2  2 4 4 8 8 16 16 32 32... It "feels" safer.  Equivalent to a progression of sq.root 2 (1.414) or almost.  If swinging from level 3 to 4 all the time, betting 2, then 2 again, it's going nowhere.  Adjustments still being worked on.

Or, for some positive progression: when winning a 3rd spin in a row, depending on the level I'd be at, I'd go:

  W     W   W   W    W    W      W
  64  32   16   32    48    64    80  ...
My comfort zone is below 128.  So I would not start betting higher if not back to lower levels.
I've been working on different pos.prog a bit, not too long ago, still in progress.

Normally playing factor 1.4 or 1.5, and deciding to stop on the moment's feel.

I also tend to track red and black qty coming out and betting the trend: deviations do exist, and can be a real factor when playing 200 spins: sometimes one is 20 or 25 above the other at spin 200.  Although the example I put attached was betting same as outcome. 

What are you guys relying to when deciding what to bet on?  I think dominant color is good, reaching on average a lower prog level than betting same, opposite or full red or black.

[cht]

Well, this being said, I never play with a factor of 2, unless for a quick game and a stop at losing the 512 mark ( which isn't too bad when playing with pennies).

I'd go 1 1 2  2 4 4 8 8 16 16 32 32... It "feels" safer.  Equivalent to a progression of sq.root 2 (1.414) or almost.  If swinging from level 3 to 4 all the time, betting 2, then 2 again, it's going nowhere.  Adjustments still being worked on.

Or, for some positive progression: when winning a 3rd spin in a row, depending on the level I'd be at, I'd go:

  W     W   W   W    W    W      W
  64  32   16   32    48    64    80  ...
My comfort zone is below 128.  So I would not start betting higher if not back to lower levels.
I've been working on different pos.prog a bit, not too long ago, still in progress.

Normally playing factor 1.4 or 1.5, and deciding to stop on the moment's feel.

I also tend to track red and black qty coming out and betting the trend: deviations do exist, and can be a real factor when playing 200 spins: sometimes one is 20 or 25 above the other at spin 200.  Although the example I put attached was betting same as outcome. 

What are you guys relying to when deciding what to bet on? I think dominant color is good, reaching on average a lower prog level than betting same, opposite or full red or black.

Only a very few things work - but any system or method that does is based on numbers appearing greater than expected (hot) and the math being in the player's favor instead of the house. - TurboGenius

[Bigbroben]

This is what I meant:

running up and down a progression like d'Alembert. 

Here:
bet x2 if loss, bet/2 if win.  Reset progression at 100 units above last br high.  Stop-loss at a lost 512u bet.
This is what I play until I find the HG with straight-ups...

So far so good, more wins that busts...

[Thanatos]

...
So, even for Alembert and Martingale on this one.
   
Let'S hybrid these to to create a mix, where doubling when losing, but going down a level when winning, instead of going back to 1:
Outcome:    R    R    B    R    R    B    B    R    B    R    B    B    B
Hybrid:        -1  -2   +4   -2  -4   +8  +4   -2   +4   -2  +4   +2  +1
Cumul:        -1   -3    1   -1   -5    3   7     5    9     7    11   13  14

Or again, still betting on black:
Outcome:    R    R    R    R    R    B    B    R    B    R    B    B    B
Hybrid:        -1  -2   -4   -8  -16 +32 +16 -8  +16 -8 +16 +8  +4
Cumul:        -1   -3  -7  -15 -31    1    17  9    25  17  33  41  45

A pure Martingale would have given a score of 6.
...

Uh shouldnt that blod R have been a (B)lack?

Outcome:    R    R    B    R    R    B    B    R    B    R    B    B    B
Hybrid:        -1  -2   +4   -2   -4  +8  +4   -2  +4   -2   +4  +2  +1
Cumul:        -1   -3    0   -2   -6    2    6     4    8    6   10   12  13

Still a very good result on hybrid marty 

[Bigbroben]

No, I meant it.  It was to show a different scenario.

Of course, to stop at 100u above is just to make sure not to get a streak of losses that would bust the 512u stop ( which could be at any level, really, say: 128, or 2048).  So, 512 is 2^9, counting the nr1, this represents a progression of 10 levels.  If you get 10 losses in a row, loss of 1023 (1+2+4+8+16+32+64+128+256+512).  9 losses in a row and 2 wins is enough to get 257u profit (0-1-2-4-8-16-32-64-128-256+512+256)= -511+512+256=257.  Kinda hard to lose.

Let's say one plays until the 512u level loses, instead of stopping at +100.  At one point one can easily see the following:

Result: L     L     L     L     L       L     L        W     L      L        W       L      W      L      W      W       L       L      L      W       L       W     W      W      W   
Level:   1    2     3      4     5      6     7        8      7      8        9        8      9       8       9       8       7       8       9      10      9     10     9        8        7   
Net:     -1   -2   -4     -8   -16   -32 -64  +128  -64  -128 +256 -128 +256 -128 +256 +128 -64  -128  -256 +512 -256 +512+256+128 +64

Cumul final: 1217 profit, 25 spins,15 losses, 10 wins. Now anything can happen, but for sure at this level, even if busting 512, that is having lost 1023 u, one is still in profit in the game.

This is pretty powerful.  I like to imagine it as a ratchet inching up slowly the floor of the maximal loss, every win raises the floor by the difference between 2 levels.  At some point, the floor is higher than zero and a loss is impossible ( unless of course missing discipline, betting 1024 and losing).  Emphasis on DISCIPLINE!!!

I make quicker money and more money playing this way, as opposed to LOTT.  So far...

Bet selection (black or red, high or low, odd or even) is another topic, but I believe it is not so relevant.  Just bet on the dominant color, or last color out, whatever.

Less aggressive progression are ok, such as 1,2,4,7,11,16,22,29,37,46,56, etc (+1,+2,+3,+4,+5,+6,+7,+8,+9,+10, etc), but the loss/win ratio has to come closer to 1 as the difference in %age between 2 levels is lower.


So, people, if anyone tries it and sees something worth mentioning, please share!

[Bigbroben]

A quick question here:
  what is more likely: to lose 10x in 10 spins, to lose 10x and win 1x in 11 spins, or lose 10x and win 2x in 12 spins?

Are the odds that different?

[Bigbroben]

For anyone not in a hurry,  used the 30 progression shown in attachment.

1,2,4,7,11,16,22,29,... (+1,+2,+3,+4,+5,+6,+7,...).

Not as aggressive as a factor of 2, but safer.  Going up and down the progression, reaping the difference between steps.

P.S.: the custom progression on R-sim is quite useful for this kind of automatic play.  Configurate the prog, start, and let it run for the night...

[jekhb76]

Nice work!
So basicly we have a 30 step progression and we play the Color that hits hoping for a second hit on that Color. When not, we go up 1 or 2 Steps on the ladder and when win we drop 1 step Right?

[Bigbroben]

Right,

there is always this one time when you'd go through the floor from losing many times in a row with r-sim.
It is snot perfect, either: the simulation once lost at around spin 600, with 50% red and 50% black overall.  The reason is many wins would stay at level 1 if already there. 
Possibility: Guetting or Reverse Labby when winning at level 1, or starting at level 10 or anywhere in the middle...

[jekhb76]

Maybe Always reset to 1 when any profit is reached, No matter how Small. And maybe An earlier stop loss for each session.
Did you ran More tests with this progression? And did you reached step 30 on the ladder More then once? When playing with pennies, you could Make the ladder much longer. Just thinking out loud.

[reiand91]

<img src="link:://:.pichost.org/images/2018/03/20/proba128eab9.jpg" alt="proba128eab9.jpg" border="0" />

[reiand91]

<img src="link:://:.pichost.org/images/2018/03/20/proba128eab9.jpg" alt="proba128eab9.jpg" border="0" />

Excuse me for the last post. This is the first time I upload an image.

Last night I did the first test. Looks good.
I will continue testing.