Author Topic: EPISTEMOLOGY  (Read 1670 times)

0 Members and 1 Guest are viewing this topic.


« on: May 30, 2019, 08:15:08 AM »

                                                              of the science of roulette

As a general rule, on the forums, players come to philosophize about the use of different techniques to build a winning method. But this discussion is too often intuitive, so the end result will often remain very hypothetical. Is there a way out of this problem?

A few days ago, one of the most eminent members of our forum evoked a sort of "anti-law of the third" or also trying to read (or break) the "roulette code". These are original ideas which concern rather a theoretical search, which would begin with an epistemological study.

Epistemology refers to the analysis of scientific methods and knowledge, to define their validity criteria (their value). If we accept that the search for a winning roulette method is a quasi-scientific domain, then we should (all) begin by philosophizing about the epistemology of a science of roulette. For us, knowledge will be synonymous with a sufficient understanding of the laws, the phenomena of chance, and techniques, not only to beat chance but to beat the roulette wheel.

We will Distinguish the laws of chance and the rules of roulette for a simple reason. Some players will find some mechanisms that can beat the hazard (a selection with a chance imbalance in their favor). But they rarely manage to beat the roulette, because of the zero, or the advantage of the house. So, can we solve this problem?

An epistemology of roulette is therefore a reflection on the "value" of the concepts and techniques we use. It's a complicated word to finally talk about simple things. So, simplify. We naturally study this epistemology by doing tests to determine the "value" of our techniques, but we try too hard to achieve a "globally" positive result. Therefore, we do not know the individual value of each technique used.

According to the first question posed, to know if there is a solution to this problem of a search for a winning method, it would be to identify each technique to accept or reject it. To achieve this result, we should not think, first, to build a winning method, we should think to build our knowledge on roulette. It is this knowledge that will then help to build effectively.

For the anecdote, the mathematician Théo d'Alost had published in 1910 an excellent book on mechanisms on EC. Alyett, another author (circa 1930) then published a method, using the best Alost mechanisms. This method was a huge success. Almost 100 years later I still find, in some countries, discussions, references, or variants, on Alyett. The theorist had presented mechanisms, and the practitioner knew how to use them. To choose the best techniques, Alyett understood their value (by logical understanding and then by doing tests).
« Last Edit: May 30, 2019, 08:46:30 AM by Stratege »
The following users thanked this post: kav, Sputnik, scepticus, Third, Astutillo, fiben7


« Reply #1 on: May 30, 2019, 08:25:53 AM »

                                                                         SECOND PART

Let's go back to today! Intuitive search is not the best solution to our problem. We must prefer a quasi-scientific approach, and begin by determining the "value" of the techniques we will use. We will find examples on this forum of techniques that seem to reduce the deviance. These techniques constitute a first base but, systematically, they are insufficient. Presumably, they beat the odds but, absolutely not the roulette. This is usually the problem that the player cannot overcome.

My second question was, can we solve this problem? We can answer with another question. Why could we get a small advantage (by a small imbalance of chance in our favor) and not succeed in reproducing the same effect to beat the roulette wheel? But how to do it?

It is possible to present a common model between the concepts of "heat" and "compensation". If we choose a heat or a Sufficient delay, we obtain a "probability of repetition" which brings an imbalance in favor of the player but, very often without being able to beat the roulette wheel.

A general rule to define an attack will be to have a sufficient "probability of repetition", according to a favorable "configuration" in a sequence of spins, and playing according to a precise "duration". These 3 parameters are added to bring a greater advantage. The combination of several categories of effects is the key to the problem!

I give an example with an advantageous game that I know well and who can become a loser, in the hands of a bad player. For simplicity, on the EC, I have 2 very reliable spins per attack which brings 0.2 gross unit. On 100 spins, or 50 attacks, I get 0.2 x 50 = 10 units. I pay the zero, and there is +7 units in flat bet. The bad player does not want to play 2 spins but, 10 per attack. It will get approximately the "gross" positive effect of 0.2 units per attack but, playing 10 spins. On 100 spins played, or 10 attacks, 0.2 x 10 = 2 units. The bad player pays zero, and his balance sheet is negative, - 0.7 unit (with 2 - 2.7). In this example, not respecting the "duration" becomes a disaster. We also observe that a single bad parameter can ruin a player.

To obtain a favorable "configuration", we must eliminate attacks on sequences of spins that have no potential to let the phenomenon of "heat" or "compensation" that we seek.

The "duration" is precise, and the favorable "configuration" adds their effects to the sufficient "probability of repetition". These settings can be very sensitive (especially with a method that does not have a significant advantage).

It is therefore essential to find the right settings, otherwise the% of profit (or loss) can vary enormously. So, to study these 3 main parameters (probability of repetition, configuration, duration) becomes unavoidable. The idea of ​​a progression must come only after, otherwise it would completely destroy its initial advantage. There are also different strategies (after these 3 parameters) to further improve a% profit. But this kind of secrets can be discovered by analyzing the history of a winning method! I wish you to meet as soon as possible these secrets, which can become like a cherry bigger than the cake! "You have to have the spirit of combination," said a very old author. He was right !
« Last Edit: May 30, 2019, 08:42:26 AM by Stratege »
The following users thanked this post: kav, Sputnik, scepticus, Astutillo, fiben7


« Reply #2 on: May 30, 2019, 09:16:18 AM »
Stratege, instant classic posts here by you, priceless ones that will be referred to by members in the time to come.

Thank you for this, I especially find interesting your notion of "duration" which I think better complements my understanding of the roulette.

In an analogy and noting your comment of the "trend" pattern etc on my thread, I would suggest to seek whether a Bet Type selection is temporarily more Trending than Mean Reverting (or the other way around) and not whether a Bet Type selection is Trending or Mean Reverting in the longterm.

The spirit of combination is the way to success, totally agree across fields in science and life.

Thanks again for sharing your knowledge.
The following users thanked this post: Third, Stratege


« Reply #3 on: May 30, 2019, 09:30:26 AM »
Interesting posts Stratege.
 I think beating roulette  is clearly a maths problem . There are 37 numbers . Can we guess correctly better than random ?  If Yes - we  can profit. If no - we lose . The only way to profit is to increase the accuracy of your Bet Selection. Simple. 

 IMO The  only way this can be done is by using an hypothesis based on an Assumption.

Bettors can only choose a Low Risk - Low Reward strategy. Or a High- Risk - Higher Reward Strategy.  Negative Progression is a High Risk Strategy . I  prefer a Low Risk Strategy . Small profits add up. Their Accumulation allows the Bettor to increase the size of his / her bets.  Flat Bets allow proper calculation of a Betting Bank .A  Progressin does not.
« Last Edit: May 30, 2019, 09:32:24 AM by scepticus »
The following users thanked this post: Stratege, Astutillo, fiben7

« Reply #4 on: May 30, 2019, 09:52:05 AM »
Interesting posts Stratege.
I  prefer a Low Risk Strategy . Small profits add up. Their Accumulation allows the Bettor to increase the size of his / her bets.  Flat Bets allow proper calculation of a Betting Bank .A  Progressin does not.

This could be a phrase of mine except for flat betting. I prefer very mild progressions with levels.

Scepticus, I say the following without sarcasm or irony but... how can you win flat betting? Progressions were invented just because its mathematically impossible to win flat betting. So... how the hell you win?  :o
The following users thanked this post: Third, Stratege


« Reply #5 on: May 30, 2019, 10:12:19 AM »
Awesome thread that must not die nor be derailed!

My first contribution will be:

We cannot win without positive variance and thus our method must find a way to seek it.  Eventually positiver variance will seek us but "eventually" is a terrible unknown and we must still be on the table playing when it happens.  The only way to stay on the table is to keep our expenses as low as possible.
Sorry that's all I have for now because I must go out.
The following users thanked this post: Stratege, fiben7


« Reply #6 on: May 30, 2019, 11:17:32 AM »
To avoid the accusation of derailing this thread -( or Trolling !  ;D    ) I'll answer your question in the Thread Sceps roulette strategies.
The following users thanked this post: Astutillo


« Reply #7 on: May 30, 2019, 02:09:49 PM »

Thank you fiben7, for your positive feedback! The notion of "duration" must be the most accurate measure of the phenomenon we are exploiting. It's exactly like the concept of "motivation". There is an "energy in one direction" (for us, a probability of repetition), an "intensity" (our configuration that accentuates our probability) and a "duration" (our most exact temporal measure). Without the combination of the 3 parameters, it will be the uncertainty (in general).

I remember better this precision: trend or return to the mean? An early response is to take spin slices (100 or 200 spins? - or combinations of several evens) and observe the conditions of a trend or a return to the mean. It's a long study. The pioneers of roulette started like this. And I admit that I had to follow this path to define "my" length of a gap / "my" ideal spin length. But that would be about the same idea with "heat": a number of repetitions and numbers / a number of spins. 
« Last Edit: May 30, 2019, 03:01:46 PM by Stratege »
The following users thanked this post: Third, fiben7


« Reply #8 on: May 30, 2019, 03:19:25 PM »
Thank you Scepticus, to interest you in this discussion. You say, can we guess the hazard?

We can find specific sequences in a spin history that shows a constant imbalance in favor of the player. The simplest explanation is to follow the pattern of a test I did 20 years ago. We know from Marigny's work that the maximum deviance is less than 6 STD. If we look for deviations worth 3.5 STD, to play on a few tens of spins only (not counting zero), we will have more chance to win than lose! Because the negative deviance can only gain another 2.5 points (6 - 3.5 = 2.5) but, the positive deviance does not have this constraint, on the contrary, it has a possible tension pressure of 5.5 STD in front of her! It will be the same with the concept of "heat" but in the opposite direction. Numbers will come abnormally to bring also a temporary imbalance  ;)
« Last Edit: May 30, 2019, 03:23:57 PM by Stratege »
The following users thanked this post: fiben7


« Reply #9 on: May 30, 2019, 03:22:42 PM »
Past results do not have affect on future performance. They should wright it on roulette table.
The following users thanked this post: scepticus


« Reply #10 on: May 30, 2019, 03:23:34 PM »
@Strat: My simulations show that a single number can hit within expectation up to 31 times in a row but only miss expectation 13 times in a row.

Just for clarification's sake, how does this apply to your above post?


« Reply #11 on: May 31, 2019, 11:04:25 AM »
Thank you, Third, for thinking that this thread deserves a deepening and a future.

If I understand your search on a single digit, you can consider that 37 spins is a cycle, just as 2 spins on EC is a cycle. You can use the calculation of the STD (Universal Calculation for Deviance) to a single number. The "binomial law" will give the same variances. Even if, with a single number, we can have 5 or 6 repetitions in a cycle, it does not seem (according to my observations) that the greatest differences are, proportionally, greater than those of the EC. The same formula is therefore possible on 24 numbers.

You also evoked with your game, the obligation of a positive variance to win before leaving the table. I will, in my next messages, talk about different things that will bring multiple insights. My next message will be on "sensitivity of the criteria".  ;)
« Last Edit: May 31, 2019, 11:09:16 AM by Stratege »
The following users thanked this post: fiben7


« Reply #12 on: May 31, 2019, 11:15:53 AM »
Ok, great!

I have been so tired I haven't read Part 2 yet! >.<
« Last Edit: May 31, 2019, 11:18:59 AM by Third »


« Reply #13 on: June 01, 2019, 07:58:26 PM »

                                                              The sensitivity of the criteria.

The notion of validity (value) of our criteria is fundamental to intelligently construct the structure of a method. But often, we might think that most roulette techniques (or criteria in general) don't have a real positive effect.

It must be said that a criterion will have a greater or lesser "sensitivity" according to a type of context. On a completely random spin sequence, a criterion will never have a very interesting effect on our spin selection.

Our game concept must first provide a favorable "probability", so that a criterion can use the strength of the tendency to imbalance brought by this probability. It's the same with a good "configuration". A criterion (or a technique) will also be helped by favorable configurations, so we must absolutely reject those that will not facilitate our techniques.

We can say that the more favorable the context, the more predictable the reactions will be. On the other hand, with a totally random context, the techniques will behave almost completely randomly.

Only one game criterion (or one technique) cannot win. Because there are different forces in the random currents that form a history of spins. These forces are a set of laws. We need to identify the main forces that will guide the currents that will produce a favorable spin distribution for us. This is why I have already said that "a law taken alone does not bring any advantage to the player, it is the combination of the laws between them" which form the phenomena that we seek. And to understand this architecture of the hazard is, in a way, "read the code of roulette"!

To illustrate this simply, the concept of "heat" indicates that the" law of the third, "or more accurately," binomial law " should act on certain numbers. But other laws also act at the same time and differently on all the numbers, or on some more particularly (the sleepers, those come once, or twice ...). We must not forget the effects on certain groups of numbers, on certain sectors ... A game concept thus isolates a particular sequence where some laws (more or less known) have a particular strength to act in a fairly predictable way for the player.

Above all, there are a multitude of effects produced by different laws (or forces), and these effects are captured by different criteria. But for criteria to begin to have a predictable reaction (a good sensitivity), we must start with a general framework, that of a favorable probability.

On this subject, if we find in a sequence of spins several favorable probabilities that fit together, we will observe that the laws which must produce numbers or opposite series have great difficulties to react. The more we add probabilities in our favor, the less the opposing forces will be unpredictable.  :o
« Last Edit: June 01, 2019, 08:10:16 PM by Stratege »
The following users thanked this post: Third, fiben7


« Reply #14 on: June 02, 2019, 03:03:34 AM »
NICE!  Maybe this is a law, I am not sure, but I observed it and definitely works:

If a group of numbers consists of hot numbers and cold numbers, the colder the cold numbers are, the greater the chances of that group to hit in the next spin.

What is interesting about this observation is that the same group of numbers could have hit last spin but the chances of hitting next spin are even greater!  This seems to be a streak predicter?  Or a run predicter?

I have found this phenomenon still falls under the curve fitting curse.
The following users thanked this post: Stratege