Matrix TrendGrabber

[atlantis]

[As requested - moved to its own thread - A.]

This might be interesting to some.

Here is a 1000 spin continuous test (results from random.org)

Tracked 2 lines then look at last TWO vertical results in each column.
If  the results are different then bet for the CHOP to continue...
If the results are the same then bet for the STREAK to continue...
STOP AT A WINNER!

Use GLC EC progression?

H   L   L   L - first line results
L   L   L   H - second line results
L   H   H   L - L,L,L,W
H   H   L   L - L,L,W
H   L   L   H - L,L,L,L
L   H   L   L - L,W (6th step)
L   L   H   L - L,W
L   L   H   H - W
L   L   L   L - W
L   L   L   L - W
L   H   H   L - W
L   L   H   H - W
H   L   L   H - L,L,L,L
L   L   H   L - W (5th step)
H   H   L   H - W
H   L   L   H - L,W
H   L   H   H - W
L   H   H   H - L,L,L,W
L   H   H   L - L,L,W
L   L   L   L - W
H   L   L   L - L,L,L,W
L   H   L   H - W
L   L   L   H - L,W
L   H   L   L - W
H   H   L   H - L,L,W
L   H   L   L - L,W
L   L   H   H - L,L,L,W
L   H   L   H - W
H   L   H   L - L,W
H   L   H   H - L,L,L,W
L   L   L   L - L,W
L   H   H   L - L,L,W
L   L   L   H - W
H   L   L   L - L,L,L,W
H   L   L   L - L,W
H   H   L   L - W
L   H   H   L - L,L,L,W
L   L   H   L - L,L,L,W
H   L   L   L - L,L,L,W
L   H   H   H - W
L   H   L   H - L,L,W
H   L   H   L - L,L,W
L   H   H   L - W
L   L   L   H - L,W
H   H   L   H - L,W
L   L   L   H - W
L   H   H   H - L,W
L   H   H   H - W
H   H   L   H - L,W
H   H   L   L - L,W
H   H   H   L - W
L   H   H   L - L,W
H   L   L   L - W
H   L   H   L - L,L,W
H   H   L   L - W
L   L   H   L - L,W
L   L   H   L - L,L,L,W
L   L   H   H - W
H   H   H   H - L,L,W
L   H   L   H - W
L   H   H   H - L,W
L   L   L   L - W
H   H   H   L - L,W
L   L   H   H - W
L   H   L   L - L,W
H   H   H   L - L,L,W
L   H   H   L - W
H   L   H   H - W
H   L   H   H - L,L,W
H   H   L   L - W
H   L   L   L - W
H   L   H   H - W
H   L   L   L - W
L   H   H   H - L,L,W
H   H   L   L - W
H   H   H   H - L,W
H   L   L   H - W
L   L   L   L - L,L,L,L
H   L   H   L - W (5th step)
L   L   L   H - W
H   L   L   L - W
L   L   L   L - W
L   H   H   H - L,L,L,L
L   H   L   H - W (5th step)
H   H   H   L - L,W
H   H   L   L - L,W
H   L   H   L - W
L   H   L   H - L,W
L   H   H   L - L,L,W
L   H   L   H - W
L   L   L   L - W
L   L   L   L - W
H   L   L   H - L,W
L   L   L   H - W
L   H   L   H - L,L,W
L   H   L   L - W
L   L   H   H - W
H   L   L   L - L,L,W
L   H   L   L - W
L   L   H   L - L,W
L   H   L   L - W
H   L   H   L - L,W
L   L   L   L - W
L   H   H   H - L,L,W
H   H   H   H - L,L,L,L
L   L   L   L - W (5th step)
L   L   H   H - L,L,W
H   H   L   H - L,L,W
H   L   L   H - L,W
H   H   H   L - W
L   L   H   L - L,W
L   H   H   L - L,W
L   H   L   L - W
L   H   H   L - W
L   L   H   H - W
H   H   H   L - L,W
H   L   L   L - L,W
H   L   H   L - W
H   L   H   L - W
H   H   L   L - W
H   L   L   H - W
L   L   L   L - L,L,W
H   L   H   L - W
L   H   H   L - W
H   L   L   L - W
L   H   H   L - W
L   H   L   H - L,L,W   
H   L   H   H - L,L,W
H   L   L   H - L,L,W
L   L   L   L - L,W
L   L   L   L - L,W
H   H   L   H - L,L,W
L   L   H   H - W
L   H   H   L - L,W
L   L   L   H - W
H   L   H   L - L,L,W
H   H   H   L - L,L,L,L*
L   H   L   H - L,L,L,L*
L   L   H   L - L,L,W (11th step)
H   L   H   H - L,L,L,W
L   H   L   H - W
L   H   L   H - L,L,L,W
L   H   H   H - W
H   L   H   H - L,L,L,W
L   L   H   L - W
L   L   H   H - L,W
H   H   L   H - L,L,L,L
H   H   H   L - L,L,W (7th step)
L   H   H   H - L,W
L   L   L   L - L,L,L,W
L   H   H   H - W
L   L   L   L - W
L   H   L   H - W
H   L   H   H - L,W
H   L   H   H - L,L,L,W
L   L   L   L - L,W
L   L   H   L - L,W
L   H   L   L - W
H   H   H   H - L,L,W
L   L   H   L - W
H   L   L   L - W
H   L   H   H - L,W
H   H   L   H - W
L   L   L   H - L,W
L   L   L   L - L,L,W
L   H   H   L - W
L   L   H   L - W
L   L   L   H - W
H   L   H   H - L,W
H   L   H   L - L,W
H   H   L   L - W
L   L   L   L - L,W
L   L   H   L - L,L,L,W
H   L   H   L - L,W
H   H   H   H - L,L,W
H   L   H   L - W
L   L   L   H - L,L,L,W
H   L   H   H - W
L   L   L   L - W
L   H   L   H - L,L,L,W
L   H   H   H - W
H   L   H   H - L,L,L,W
H   H   H   H - L,W
H   L   L   H - W
H   H   L   H - W
L   H   L   H - L,L,W
L   H   L   L - L,W
L   L   L   L - W
H   L   H   L - L,L,L,W
H   L   H   H - L,W
H   H   H   L - W
H   L   L   H - W
H   H   H   L - W
L   H   L   L - L,L,W
L   H   L   H - L,W
L   H   H   H - W
L   L   H   L - W
L   L   L   H - W
H   H   L   H - L,L,L,L
H   L   L   H - L,W (6th step)
L   L   L   H - L,L,W
H   L   L   L - W
L   H   H   H - W
H   L   H   L - W
H   L   H   H - L,L,W
L   H   L   H - L,L,L,L*
L   H   L   L - L,L,L,L*
H   H   H   H - L,W (10th step)
H   H   L   H - L,W
H   H   H   L - W
L   L   H   H - L,L,L,W
L   L   H   H - L,L,W
H   L   L   L - L,W
H   H   L   L - L,L,L,L
L   L   H   L - L,W (6th step)
H   L   L   H - W
H   L   H   L - L,W
H   H   L   H - W
L   H   H   H - L,L,W
H   H   L   H - W
H   H   H   L - L,W
H   L   H   L - W
H   L   L   H - W
L   H   H   H - L,L,W
L   L   H   L - L,W
L   L   H   H - W
L   H   H   H - W
H   L   H   L - L,W
H   L   H   L - L,L,W
L   L   L   L - L,W
H   H   H   H - W
L   H   L   H - W
L   L   H   H - L,L,W
L   H   L   L - W
H   L   L   H - L,W
L   H   H   H - W
H   L   H   L - W
H   L   L   H - L,L,L,W
L   H   L   H - L,L,L,L
H   L   L   H - W (5th step)
L   L   H   L - W
H   L   H   L - W
H   L   L   H - L,W
L   L   H   L - L,W
L   H   L   H - L,L,W
H   L   L   H - L,W
L   L   L   L - W
H   H   H   L - W
H   L   L   H - L,W
L   H   L   L - L,W

The variable chains of same results in each column actually help this approach.

There are MANY wins (224) coming on the 1st to 4th step of the progression. In other words a win obtained on each of 224 lines! (896 of the 992 in-play numbers)

There were 5 wins at the 5th step; 3 wins at the 6th step; 1 win at the 7th step; 1 win at the 10th step and 1 win gained at the 11th step.
Maybe with some rules and a different progression (e.g. GLC's) and maybe a stop rule this could be good and pay dividends?

There may be also some merit in playing only after LL or LW for instance as Zer0Blue often suggests - although you can miss out on additional first column wins doing that...

Why not try it and play only after L,L,L,L or L,L,L,W?

Please share your thoughts and ideas with examples and any results obtained below.

Thank you.
Atlantis.

[atlantis]

Zer0Blue wrote:

;-)
That is it dear Atlantis. I know we have talked about this some weeks ago... It is very cool to apply it here also.
as for LW strategy it seems essential.

Colbster wrote:

Just ran with a grand marti (1,3,7,15,31,63).  Only lost at 4th level 1 time - think that might be a great stop point with 26 units at risk.  I went +89 before losing my 26 the first time.  Very cool idea.  Also, I played continuously without stopping when I had a win in the row of 4.

BuffaloWizard wrote:

That's a great shout colbster,
I like that progression...and if further testing proves that 4 levels are strong enough, then the odd 26 unit loss would be surmountable with this more aggressive progression.

[amk]

Hello atlantis,

Sorry, what do you mean with 7th/10th/11th step?

You are not using a six step progression?

How are you betting?

[ZeroBlue]

just tested this on 10000 games = 50008 spins


flat bet continuous- overall balance -438 units


it seems good.

[atlantis]

Sorry, what do you mean with 7th/10th/11th step?
You are not using a six step progression?
How are you betting?

Well remember this is still fairly new and not finalised (that is why in testing section) If betting EXACTLY as I showed in the 1000 spin example you need to have a stop; or a progression to cover the longer losing runs that may occur... (the highest was 10 L's)
Maybe George's (GLC) even money progression can be valuable here:
link:://rouletteforum.cc/money-management/even-chance-progression/
GLC turned my MV6 loser into a WINNER using that...!?
So maybe GLC can help us here with some input and thoughts?
There is my idea of betting only after LLLW or LLLL which might be valid in curtailing lengthy runs of L's too.
For better progression stability at this moment I am unsure of what is the most economical way forward...   
A bit of brainstorming required I suppose.

A.

[ZeroBlue]

maybe if we switch bet it will report different results


i will try this
bet the trend.


if last are different I bet different if they are same I bet the same so when I lose I switch. i guess you played the trend until a hit dont ya?



It is a random walk.



what do you think?

[GLC]

Well remember this is still fairly new and not finalised (that is why in testing section) If betting EXACTLY as I showed in the 1000 spin example you need to have a stop; or a progression to cover the longer losing runs that may occur... (the highest was 10 L's)
Maybe George's (GLC) even money progression can be valuable here:
link:://rouletteforum.cc/money-management/even-chance-progression/
GLC turned my MV6 loser into a WINNER using that...!?
So maybe GLC can help us here with some input and thoughts?
There is my idea of betting only after LLLW or LLLL which might be valid in curtailing lengthy runs of L's too.
For better progression stability at this moment I am unsure of what is the most economical way forward...   
A bit of brainstorming required I suppose.

A.

A,


I will take a closer look.  There are a couple of stretches where it looks like you have too many losses close together without a Win streak long enough to wrap a win up.  It may cause us to lose the progression.  I'll look close later and play those areas out and see how it holds up.
I think the progression based on a 5 step parlay can handle this well enough since it starts with betting 1 unit as a starter bet for the 1st 31 bets.


The one I listed that started with 1-2-4-8 may not hold up.  We'll see.


GLC

[GLC]

The problem with this system is that there aren't enough 4 or 5 wins in a row.  I don't think the let it ride progression will work so well here.


On the MV6 I think yes.  Much better at getting long streaks which is what these let-it-ride bets take.



GLC

[atlantis]

maybe if we switch bet it will report different results
i will try this
bet the trend.
if last are different I bet different if they are same I bet the same so when I lose I switch. I guess you played the trend until a hit don't ya?
It is a random walk.
what do you think?

Hi Zer0Blue,

Sounds OK - but not sure if switching, rather than staying with trend after a loss will make any appreciable difference or change matters...

I recently tried it FLATBETTING after LL and using O/E matrix of three... Betting max of twice per line. (no bet on first column) + stop at a winner - betting only for chop or streak to continue depending on last vertical pair results in col2 and col3.

OEE
EEO 
OOO LL*
EEO WW   +1     +1
EEE  LL
OOO LW    +0     +1
OOO LL*
OOO WW   +1     +2
OOE WL
EEO  LW
EEE LW
EEE WL
EOE LW
EOE LW
OEE LW
EOE WW
EEO WL
OEE LW
EEE WL
OOE LW
OEE WW
OOE WW
EEE WW
OEO LL*
OEO WL   +1       +3

+3 IN 75 SPINS.
A.
PS. Obviously the Zero is an enemy to this - so maybe try on nozero roulette?

A.

[vundarosa]

The problem with this system is that there aren't enough 4 or 5 wins in a row.  I don't think the let it ride progression will work so well here.


On the MV6 I think yes.  Much better at getting long streaks which is what these let-it-ride bets take.



GLC

-----------------------------
George,
Have you tried using the Tour de Force concept after each vertical double?!

vundarosa

[Colbster]

Atlantis,

After my message to you last evening about running with the Grand Marti, I did up a group of spins from actuals for demonstration purposes. On any 4-spin EC bet, there is a 1/16 chance that we will bust out.  In my tests, on all 3 EC groups, my results were shockingly consistent in demonstrating exactly these odds.  I busted out twice on each of the 3 ECs in the first 48 spins, with another bust on the B/R in the next set of 4 spins.  There is very possibly an argument for playing X number of spins after a virtual loss in a wwLL application, but that will need more testing.

[Colbster]

As an aside, this bet ends up being nothing other than penultimate bet.  If we have X then O, we bet X.  If we have X then X, we bet X.  If we have O then O, we bet O.  Just one-before-last.

[atlantis]

As an aside, this bet ends up being nothing other than penultimate bet.  If we have X then O, we bet X.  If we have X then X, we bet X.  If we have O then O, we bet O.  Just one-before-last.

Hi Colbster,
Well that is true - except to say of course that the results are not being recorded linear fashion; they are noted in a MATRIX grid - so it's not quite the same...
A.

[Colbster]

You are right - I didn't mean it at a criticism in the slightest.  I just made the observation because it is much easier to code or track when you ignore patterns and play the spin result 2 boxes up in a tracking matrix.  Less room for error as well.

[GLC]

-----------------------------
George,
Have you tried using the Tour de Force concept after each vertical double?!

vundarosa

Sorry V,

I haven't had time to look into this system much.  I have a lot of irons in the fire in my life at this time.  I do know that the Tour de Force progression is hard to defeat.  The key thing is that we can't go too long without a 5-in-a-row win series.  As long as we get one every 200 or so bets, we can stay ahead.  It can get to be somewhat of a grind.  But if we win in the end, who cares?

I know you're busy also, but maybe you could find time to see how it performs.  All we need is to determine how long before we get 5-in-a-row after reaching our 16th 1, because after that point we need 5-in-a-row to recover fully.  That's not taking into consideration the option of dropping back down the progression line after a 3 or 4 win.

George