Developping a valuable indicator of tendency

[Arteinvivo]

In the new section VLS opened (Arte's studio) I will develop this new thread "Developping a valuable indicator of tendency" one bit at a time as an exploration of uncharted territory. I hope to add some more possibilities to my Alembert system some of you have already begun to test.

I am presently discussing with a french professionnal trader who graciously accepted to share with me different ways to use an indicator of tendency which *might help to improve any systems. I'll discuss these possibilities and more but in relation to the Alembert system I have developped recently to decelerate my level of activity in bad moments but increase it when there is concentration of hits. I hope you'll join the discussion till then have a good time and take care.

  -Arte

*might : I want to make this clear though... I don't think we can use the past in a game of independant events as a way to predict with 100% degree of certainty the next outcome. But we certainly can use the past to structure our bets in order to exploit a stream of decisions in a very systematic way. My numerous attempts to find a good system have showed me that such structure is more efficient than a pure blind luck approach but you have the right to disagree. It is my contention we can more easily measure the level of productivity a system can deliver and react accordingly if we approach the game in a very systematic way by opposition to a system which follows no clear directions but works on a hunch and guess basis.

[Arteinvivo]

What is wanted is a formal frame of reference rich enough, without subjectivity in order to measure precisely what's going on.

We want to know when the level of productivity of our system is ...

1) abnormally in good shape (STRONG RISING tendency)
2) in good shape (RISING tendency)
3) perfect equilibrium (ABSENCE of tendency)
4) in bad shape (FALLING tendency)
5) abnormally in bad shape (STRONG FALLING tendency)

[Arteinvivo]

What kills any method of play is when an ecart large enough occurs it cannot support. This fluctuation from the norm literally explodes the capital of play, especially when a progression is being used. It is thus advisable to correctly interpret the value of such a variation.

To correctly estimate the maximum amplitude of a fluctuation or ecart for a given event, we need to compare it to the ratio of average profits. It is generally known as the financial variation, whose very simple formula is this :

financial variation = maximum variation / ratio of average profits

For example, a variation of 20 is an important ecart for any even money chance whose ratio of profits is 2 to 1. However, this same variation of 20 is a very positive ecart or variation for a straight-up number whose ratio of profits is 36 to 1.

Let's plug these values into our formula :

Thus, in our example, one will have

- even money bet, financial variation = 20 / 2 = 10

- straight-up number, financial variation = 20 / 36 = 0.56

In general, it is considered, that a method of play whose financial variation is lower or equal to 4 (<= 4), even associated with a light progression, is excellent.

Knowing this, it becomes easier to understand that the more one has a low financial variation, better it gets and the more the game is getting interesting. The financial variation 10 we got in the example above is very bad. Unfortunately, a method of play which does not have an effective management of these variations, shows a financial variation around 10.

To decrease the financial variation, there are the choice between decreasing the maximum ecart or increasing the ratio of profit. If we use a method of play which always exploits the same type of chances, we can't change the ratio of profits. In this case, we can only act on the maximum ecart and try to reduce it.

Note: while decreasing the variation, the ideal would be a system of play in which one could, under particular circumstances, pass from a type of chances to another offering a higher ratio of profits.

What factor contributes the most directly to the maximum ecart ? Obviously, the rate of success of the method of play. The more the percentage of success increases and the more the maximum ecart will decrease.

Consequently, how can we increase the rate of success? By selecting which spins should be played, while choosing, if possible, optimal periods to play.

For example, using any even chance, there is a fixed rate of success equal to 48.6%, but this rate will undertake or pass through multiple variations over time. Thus, on series of 100 spins for example, one will observe a higher rate of success than 48.6% and sometimes this rate will be inferior. It is this variation we want to exploit or take advantage of.

The first difficulty which comes to mind is how to know if a rate of 30% out of 100 spins observed is really an abnormal variation or not in the course of a game. One needs an approach which makes it possible to answer it and, this, for all type of chances or all combinations one might wish to contemplate. We thus need a general formula which would tell us if the tendency is in good shape or if we are getting accross a turbulent zone where the ecart is getting too large, immediately and without ambiguity.

The general formula, which will become our indicator of the current tendency is this :

DE = (NSW - (NSP * p)) / SquareRoot(NSP * p * q)

In this formula the various variables means :

NSW : Numbers of Spins Won during a game
NSP : Numbers of Spins observed and Played
p : probability of winning (for example, for any even chances, p = 18/37 = 0.027)
q : probability of losing = 1 - p
DE : degree of ecart for the given chance.

How to interpret this degree of ecart value (DE) ?

[size=12]The formula is built to provide different levels of interpretation[/size]

If DE is greater than 1, the game is abnormally in good shape (STRONG RISING tendency)
If DE is comprised between 0 and 1, the game is in good shape (RISING tendency)
If DE = 0, the game is in perfect equilibrium (ABSENCE of tendency)
If DE is comprised between 0 and -1, the game is in bad shape (FALLING tendency)
If DE is lower than -1, the game is abnormally in bad shape (STRONG FALLING tendency)

Example :

During a game where we are playing the even money chance "Even", we are at spin 85, so NSP = 85, we have won 30x, so our NSW = 30, p = 18/37, q = 19/37 or 1-p then DE would equal :

DE = (30 - (85 * 18/37)) / SquareRoot (85 * 18/37 * 19/37) = -2.463

As DE = -2.463 we come to the conclusion the game is abnormally in bad shape (STRONG FALLING tendency)

In order to avoid making calculations by hand, one can build a small table by type of chances or combinations played :

[Arteinvivo]

Using the above table of references, relative to a chance or particular combination, we can define entry and exit points of attack, i.e. moments for which we consider the timing is good to start or stop playing a particular chance.

For example, one way to proceed could be to play the chance only when DE is between -1 and +1, but stop when DE passes below -1 or above +1.

Another way could be to wait until DE passes initially below -2, then to wait until DE passes once again above -2. At this point, we start to play but stop when DE gets below -2 or above -1 or 0 or +1.

The many ways to proceed are varied and only some tests will make it possible for everyone to choose his own personal tactic, to determine our own levels of reversal of tendency. But we must admit, with a game of mere chance, nothing never guarantees that a reversal of tendency will occur or continue while it seems to show. But, a thing is sure, partial returns to balance will appear with this type of indicators.

[Arteinvivo]

OK, the presentation is terminated. This is the theory, now we need to apply it to our systems. If you have ideas or questions you would like to add or ask then go ahead. I have deleted 2 commentaries to avoid breaking the theory between multiple comments.

[J.Daniels]

Very good job arte,

the first thought I had, is to use this indicator tendency with systems that covers many numbers, like 2 dozens.. like that the tendency of good and bad shape will be much more regular. Although I have a question  

NSP : Numbers of Spins observed and Played

For example playing your DAlembert system, NSP would be the spins that you actually placed a bet on, or "played and observed" (all of them) ?

Thanks

JD

[Arteinvivo]

I would measure the complete stream of decisions observed played or not. If my target is Red even though i make some virtual bets i will measure all Red versus all Black. I want to know where Red is going, i.e., is it rising or falling.

[Arteinvivo]

Comrade Bliss from VIP posted this interesting complement of information :

Your formula doesn't take into account "inertia", "momentum", "the tendency to keep going", which is Newton's first law:

Every body perseveres in its state of being at rest or of moving uniformly straight forward, except insofar as it is compelled to change its state by force impressed.

Let's assume, as you are, that Newton's law does apply to roulette.   [smiley=shocked.gif]

What you need is a way to measure the rate of change of the current conditions. Your formula does a good job of showing you what the trend is, just as a moving average will do, but what may be more useful is to know when a change of phase is approaching, a "turning point". A simple way to measure the rate of change is to divide the current DE by its value X spins ago. If the resulting number is relatively large, it indicates an acceleration, if small a deceleration (or levelling out). See also nolinks://en.wikipedia.org/wiki/Rate_of_change_%28technical_analysis%29

It might be worth giving it a try, what's to lose?

[Arteinvivo]

-