Point 'n' Figure Strategy

[Bayes]

This is work in progress but thought I'd share the idea. The point and figure method of charting stocks is one of the oldest techniques of technical analysis used by traders, although other strategies seem to be more popular these days. The method attempts to put you on the current trend and keep you there as long as it lasts. Adapting it to an even money game is simple, for example, let x = red and o = black and suppose the first few spins are R, B, B, B, R, R, B, R, B, B, B, B. Your chart would look like this:
 
x   
  o x   x
  o x o  o     
  o        o
            o
            o

I think you get the idea.

You can plot the chart manually using squared paper, or use excel. I've written a program which makes it easier, here's a screenshot of a session I played this morning (note that I used different coloured o's instead of x's because the coding was easier) :

[attachimg=1]

So how do you stay on the trend? simple, if the CURRENT line of x's or o's moves to 1 higher than the PREVIOUS line, then you start betting for the "breakout" to continue, in other words assume that it's the start of a trend and bet on it. Refer to the screenshot for an example; the first few spins were wins for red, on the 4th spin red moved lower than its previous line, so I started betting at spin 4 on red. I continued betting on red until black achieved a breakout (this occurred at spin 23) then I switched to betting black. Success for black was short-lived, however, and red broke out again a few spins later, prompting another switch in betting.

That's really all there is to it, although there are many possible variations. I have gotten better results following the trend betting DBL (Decision before last). That means I don't take, for example, o = red and x = black, but o = a WIN for DBL, and x =  a LOSS for DBL (or vice-versa, it doesn't matter as long as you're consistent). You could use the method with ANY bet selection, in fact.

Another possibility (which I haven't investigated yet), is to take multiple spins representing each x or o. For example, instead of using 1 spin = 1 x or o, take 3 spins at a time and mark a x if there are 2 or more reds, and a o if there are 2 or more blacks, that way you only get a breakout if there is a fairly strong trend.

By the way, the reasons why this method supposedly works on stocks don't exist in roulette (or any fixed odds game); there is no psychology involved and no supply and demand, it's merely a way of trying to keep on the current trend.

[ego]

Nice work - very cool ...

I think is more easy to chart does horizontal with i know Johno at GG does (baccarat player) ...

x   
  o x   x
  o x o  o     
  o        o
            o
            o

Could be ...

X
000
XX
0
X
00000

Cheers

[Bayes]

If you want to know more about point n figure charting (different formations, optimisation etc) I've attached a nice book. If it doesn't work on roulette, you might have better results doing real trading. 

[attachmini=1]

[ego]

Very nice work - very nice PDF - it is with the same line of thinking when i experiment - great contribution Bayes.
I see this working using the method above and it also apply to different ways when it boils down to sport-betting.

Pretty cool.

PS i don't use it for real but i enjoy experimenting ...

[TwoCatSam]

Bayes

Very interesting.  Will you make the program available?

Thanks for the input.

Sam

[albertojonas]

 


thank you very much. That is great. i promise further explorations.

[Bayes]

Will you make the program available?

Hi Sam,

Yes, but there are still a few options I want to implement.  It probably won't run out of the box on anything except Win XP, so you might need to download "Dosbox", which is a DOS emulator for playing old games (for anyone under about 30 who doesn't know what DOS is, don't worry about it, I'll give instructions on how to set it up, it's easy).

Very nice session I played this morning, it's best when the outcomes are trending -

[attachimg=1]

Again, this was betting DBL.

The nemesis is "ranging" behaviour (more horizontal movement than up and down), but even then it has to be specific to give heavy losses. The worst possible outcome is when you get long streaks of exactly the same length on both sides of the bet selection, but this is rare, and with judicious MM you can ride out the worst of it when it happens.

[Bayes]

thank you very much. That is great. i promise further explorations.

Thanks, any feedback is welcome. Don't forget this is a kind of meta bet selection which can be used for any method of choosing your bets. If you have a what seems to be a good trending system, try it with this. One of the things I'm going to add to the program is an option to select wins and losses, so rather than have a selection "hard-wired" as it were, the chart will be built up from wins and losses, which can obviously apply to any bet selection. The chart will then show you when to switch from bet to anti-bet.

Another thing to make use of is the grid. It's rare that the chart spans more than 4 squares vertically over 100 or so spins, so you could use this as a factor in your decisions, a little like the Bollinger bands some traders use.

[ego]

 

Another thing to make use of is the grid. It's rare that the chart spans more than 4 squares vertically over 100 or so spins, so you could use this as a factor in your decisions, a little like the Bollinger bands some traders use.

Can you explain that further in more detail as i don't fully grasp it.

Cheers

[albertojonas]

Dear Bayes,


your efforts are priceless.


I am very excited with the option for wins and losses strings.


What i expect is to improve somehow the strike ratio of any bet selection. I will experiment today with imbalance and favorites (i am glad ego is participating on this thread also as he likes to have fun with this too), as i feel it very promising.


Cheers

[Gizmotron]

I'm so glad to see some of you working on trends and patterns.

For me it all comes down to visual dexterity. In my type of charts you can see the red & blacks but you can't  see the over all dominations because of the confines of the charts.

link:://rouletteforum.cc/index.php?topic=5931.0 If you look down the page you can see what I mean.

I like this kind of charting found in the graph of Bayes' software. It also identifies the over all domination.

link:://vlsroulette.com/index.php?topic=20003.0 BTW I did make things clear about my secrets.

[TwoCatSam]

Bayes


I knew there was some reason I started visiting the forums again. 

It is a pleasure to read someone who writes without benefit of obfuscation, bloviation, or smoke and mirrors.

I have thousands of old spins from Dublin.  I will learn to chart this on graph paper.

TwoCatSam

(And my wife read it and wants to go to Vegas!)

[Bayes]

Can you explain that further in more detail as i don't fully grasp it.

Just saying, I added the grid for a reason. I haven't worked out the standard deviations in terms of the squares, but if you did that then you could use the grid as a guide to the limits one side is likely to go (or not go beyond). For example, if you were to get 20 blacks in a row (or 20 wins in a row) then this would cover 4 squares (5 spins per square) vertically but the horizontal movement would be zero squares. This would be very rare, but what would the rate of change be in terms of squares for an SD of 3.0? (ie: no. of squares up/down divided by no. of square along).

[ego]

Cool ... This evning i will relax with some wine and read the PDF ...

Cheers

[albertojonas]

[reveal=Results]
w
w
l
w
l
w
w
w
l
w
w
w
w
w
w
w
w
l
l
w
w
l
w
w
l
l
l
w
w
w
w
l
l
l
w
w
l
w
l
l
w
l
l
l
w
w
w
w
l
l
w
w
l
l
w
l
w
w
[/reveal]

my first test results with a simple bet selection (bet against series of 3 EC).

Quite short as it takes long time to chart all the bets.

*Do You think that different bet selections provide more or less favorable charts to bet on?

Would love to hear your thoughts on this.

Cheers