Standard Deviation and z-score in plain English

[Bayes]

Just noticed this, I'm fine iggiv, how's yourself.

I see that the unicode math symbols have been messed up in this thread for some reason. Don't know why that is, maybe Vic can shed some light on it when he returns.

[albertojonas]

hy bayes!

Maybe this is a silly question...
Is there any instrument to "tell" or give a clue on when or if the trend is inverting?

like completing a cycle?

i am more or less familiar with bollinger trends analises...

Great posts by the way
Congrats.
AL

[Bayes]

Hi AJ,

If you know about bollinger bands you probably also know about other technical indicators like moving averages, elliot waves, fibonacci retracements etc used by technical traders. I suppose you could try any of these on roulette but you'd have to define the number of spins over which you choose to identify a trend, and it's all quite subjective. Personally I like point and figure charts; they're quite simple to use and are in some ways more objective than other methods because there are clear rules for when to enter markets. Just keep in mind that they're based on supply & demand, a concept which doesn't apply to roulette outcomes. 

[albertojonas]

That is very true: supply and demand. And roulette is of a different nature. maybe one can also use

link:://en.wikipedia.org/wiki/Stochastic_oscillator

Precisely what I struggle with is to focus any observation in a window.
we all have made the naive question once: where do we enter the randomness?

-I try to develop some kind of detection to center myself on random before I start play.
it could be either z-score=x, wait till all 37 numbers appear, or other elaborate events depending on what I am betting.
do you use any or know of others? This is kind of a fun exploration to me.
Very interesting stuff,
thanks bayes.

[Gizmotron]

Here is a sample of some "maximum sleeps":460 - 540 spins with a straight up not showing. (maths expectancy is a hit once every 37 spins)
309 splits with a split not showing. (maths expectancy is a hit once every 18.5 spins or so)
178 spins with a street not showing. (maths expectancy is a hit once every 12.3 spins or so)
155 spins with a corner(square) not showing. (maths expectancy is a hit once every 4.6 spins or so)
93 spins with a line (double-street) not showing... (maths expectancy is a hit every 4.6 spins or so)
36-40 spins with a dozen/column not showing.(maths expectancy is a hit once every 3 spins or so)
20- 36 spins with any even money chance not showing. (expectancy is a hit once every two spins)
10 numbers may sleep for about 45-55 spins

That's great information. It's mathematical data on the existence of the elegant pattern  concept. That's very interesting about the sleeping dozen. By experience I see runs up to 30 every once in a while. A sleeping dozen occurs for 16 spins all the time, meaning during almost every four hour session. But that's from playing experience.

Also you mentioned later about this:

Personally I like point and figure charts; they're quite simple to use and are in some ways more objective than other methods because there are clear rules for when to enter markets. Just keep in mind that they're based on supply & demand, a concept which doesn't apply to roulette outcomes.  

Here is the easy to read charting I use in real casinos:


| L M H | 1 2 3 |  | B _R | L  H | O  E | -- ## -- Line
| X     |   X   |  | X    | X    |    X | --  8 --  1
|     X |     X |  |    X |    X |    X | -- 36 --  2
|   X   |     X |  | X    |    X |    X | -- 24 --  3
| X     |     X |  |    X | X    | X    | --  9 --  4
| X     |     X |  |    X | X    | X    | --  3 --  5
|   X   |     X |  | X    | X    | X    | -- 15 --  6
| X     |     X |  |    X | X    |    X | -- 12 --  7
|---------------------------------------|  -  0 --  8
|   X   |   X   |  |    X |    X | X    | -- 23 --  9
| X     |     X |  |    X | X    | X    | --  3 -- 10
|   X   | X     |  | X    | X    | X    | -- 13 -- 11
|   X   |     X |  | X    | X    | X    | -- 15 -- 12
| X     | X     |  | X    | X    |    X | -- 10 -- 13
|   X   | X     |  |    X |    X | X    | -- 19 -- 14
|   X   |     X |  |    X | X    |    X | -- 18 -- 15
|     X | X     |  | X    |    X | X    | -- 31 -- 16
| X     |   X   |  | X    | X    |    X | --  2 -- 17
| X     |   X   |  | X    | X    |    X | --  8 -- 18
|     X |   X   |  | X    |    X |    X | -- 26 -- 19
|     X |     X |  |    X |    X | X    | -- 27 -- 20
|     X | X     |  |    X |    X |    X | -- 34 -- 21
|   X   |     X |  |    X | X    |    X | -- 18 -- 22
|   X   |   X   |  | X    | X    | X    | -- 17 -- 23
|   X   |   X   |  | X    |    X |    X | -- 20 -- 24
|     X | X     |  | X    |    X | X    | -- 31 -- 25
|   X   |   X   |  | X    | X    | X    | -- 17 -- 26
| X     |   X   |  | X    | X    |    X | --  2 -- 27
|     X | X     |  |    X |    X | X    | -- 25 -- 28

[albertojonas]

gizmotron, nice to have you here.  

i play even chances in pairs looking for inbalance as a constatant
from the window chart above it would have the following results:

+7 on BR
+6 on lH
+3 on OE

how do you apply it gizmotron?

[Gizmotron]

Gizmotron, ... how do you apply it gizmotron?

I use a juxtaposed form of condition identification. I picked up a good tip from John Patrick, mainly his use of the Regression System combined with "Up & Pull." I had to invent my own MM for betting on two dozens or two columns at once. But I also juxtapose that two doz/col with betting flat or progression with a single dozen or column bet. It all comes down to my ability to read the current conditions. I also just focus on the doz/col and leave all the rest to go off and do whatever it does. I don't watch the EC's anymore. I have my reasons.

[albertojonas]

may you elaborate on that?

[Gizmotron]

May you elaborate on that?

I look for singles or sleeping dozens occurring in the doz/col sections of the table. If they are dominating then I bet with them. I do this by betting the other two dozens that are not sleeping. Or I bet that another single will hit. If these conditions are absent and repeating doz/col are running strong then I bet single doz or col bets. Sometimes I bet flat, sometimes I bet my progression or my Up & Pull Money Management technique. I figured these things out on my own.  I don't see a compelling reason to share good MM techniques on an open forum. People are smart enough around here to figure out good balanced methods on their own. It's fun to watch them try too. (Look up "John Patrick, Up & Pull, and regression.")

[hanshuckebein]

hi Bayes,

thanks a lot for this tremendous piece of work you've done for and presented to us all. 

I must admit that reading and trying to understand it really keeps the blood in my brain circulating. 

my schooldays were over about 30 years ago and math was never one of my most beloved subjects. I never listen too well to what the teacher tried to tell me. it seems now it's all coming back to haunt me. 

may I ask one question: can I use the z-score also to measure the runs and the changes of an even chance?

say I have a result of R R R B R R B B R B.

so I have a run of  R first, then change to B, change to R, run of R, change to B, run of B, change to R, change to B.

and now I'm not interested in R or B as such but interested in how often the ball produces a run or a change no matter what colour

well, I hope I've made myself halfway clear and hope for an answer that suits a math-dummy like me. 

thanks and cheers

hans

[Bayes]

Hi Hans,

You can indeed use the z-score on runs and chops. The 'law of series' says that a double  is half as likely as a chop, a run of 3 is half as likely as a double, a run of 4 is half as likely as a run of 3 etc. This means that you can use the standard formula with p = 0.5 (ignoring the zero effect). So to get the z-score of the singles vs other series (2,3,4...) just count the number of series as though they were red and the number of chops as though they were black (or vice-versa) and plug them into the formula. You can also use it for series vs longer series, so for example, ignoring all the chops, you could just look at the runs and calculate the z-score of the runs of 2 vs the remaining series (3,4,5,6...).

Hope this is clear, if not let me know and I'll post an example.

[ZeroBlue]

Hi Hans,

You can indeed use the z-score on runs and chops. The 'law of series' says that a double  is half as likely as a chop, a run of 3 is half as likely as a double, a run of 4 is half as likely as a run of 3 etc. This means that you can use the standard formula with p = 0.5 (ignoring the zero effect). So to get the z-score of the singles vs other series (2,3,4...) just count the number of series as though they were red and the number of chops as though they were black (or vice-versa) and plug them into the formula. You can also use it for series vs longer series, so for example, ignoring all the chops, you could just look at the runs and calculate the z-score of the runs of 2 vs the remaining series (3,4,5,6...).

Hope this is clear, if not let me know and I'll post an example.

Bayes do you presently apply Marigny de Grilleau »Le Gain scientifique d'une seule unité« ?
[/size]That is a very cool subject i would love a thread on that as it is part of the not very well told roulette myths.
[/size]
[/size]nice and concise explanation by the way.
[/size]Congratulations, Bayes

[Bayes]

Hi Zeroblue,

Member 'ego' has discussed Marigny's methods in detail on the forum. Search for 'cut point methodology'. He's the expert on it so you should contact him if you have any questions.

[hanshuckebein]

hi bayes,

thanks for your explanation. an example would be great. 

this is why I asked my question:


link:://rouletteforum.cc/roulette-and-gambling-framework/one-more-time-the-layout/

cheers

hans

[Bayes]

Ok, the 'master' formula for calculating the z-score is <drum roll>

z = (w - np) / √(np(1 - p))

n = number of spins or events.
w = number of wins.
p = probability of the 'event' you're interested in.

Using your example:

say I have a result of R R R B R R B B R B.

Whenever you're using the formula, you have to be careful to define the number of events (n) and wins (w) in terms of what you want to know. For example, if you wanted to know the z-score for red vs black, you would count the total number of spins to find n, but in this case, the so called 'sample space' is divided into not red and black, but streaks and chops, so the total number of streaks and chops (n) is 5. Note that I've ignored the final B in the sequence because we don't know whether it's a chop or the first element of a streak.

It doesn't matter whether you choose wins to be chops or streaks, as long as you're consistent when interpreting the results. Let's take w to be the number of chops, in that case w = 2, and p = 0.5.

so z = (2 - 5 x 0.5) / √(5 x 0.5 x 0.5) = -0.447

The essence of Marigny's method is to track the chops and streaks until you get a z-score of  ± 3.0. You then bet for the 'correction'.