### Author Topic: Probability Theory is not useless  (Read 1596 times)

0 Members and 1 Guest are viewing this topic.

#### Bayes

• Moderator
• Veteran Member
• Posts: 688
• Thanked: 563 times
• roulettician.com
##### Probability Theory is not useless
« on: March 09, 2017, 09:29:18 AM »
Some players believe that probability theory is "useless" for the gambler. Reasons put forward for this opinion may be that your playing sessions are too short for the rules of chance to be applicable. E.g. in the Talos_Dump thread:

Quote
CANON 1

I studied a lot of math. Roulette is not a probabilties game
(you play too few shots for the rules of chances to be involved).
It is just math, plain and simple.

But what math is needed to create systems? Even if you ignore bet selection (as DrTalos advocates), you still need to design a money management plan appropriate for the location(s) you play, and that plan must necessarily take into account, even if by trial and error, the probabilities of the events.

This is quite easy to show by a simple example. Suppose you've never played roulette but find yourself in a casino. You learn that single numbers pay 35-1 and that appeals to you. You have \$100 to play with and decide to put \$10 on number #17 until your bank is lost, or until a win. Is that a sensible use of your resources, given that you're looking forward to an evenings entertainment? Obviously not, because the chance of a win is only about 24%. Any MM scheme must take account of the probabilities of winning/losing runs if it's to be "successful" (however you define that).

Also, to say that probability theory doesn't apply in the short run is not true. It applies just as much to 1 bet as a million, the difference being that in the latter case the "error" is much less, but in any case variance or dispersion can also be quantified using probability theory. After all, probability theory is successfully used by advantage players, who play just as many or few spins as the average system player in a session, so why should it be useful to the former but not the latter?

Then there's the objection that probability theory, being merely a "theory", is a kind of junk science offering only tentative conclusions. People fixate on the word "theory" not realizing that in a scientific context the word doesn't mean what it means in everyday speech, where it usually means a "hypothesis" or guess. In my dictionary a (scientific) theory is defined as a well-substantiated explanation of some aspect of the natural world; an organized system of accepted knowledge that applies in a variety of circumstances to explain a specific set of phenomena. So, not a "guess", then.

Another objection made by many system players is that because the "theory" says no system can win, it might as well be abandoned as useless, or just wrong. But this is based on a very narrow view of probability theory, which is a huge field of study with new applications and results being discovered all the time. To reject the entire field on the basis that you don't like one isolated result is to throw out the baby with the bathwater.

It's true that the theory says, given certain assumptions, that no winning system is possible, but you can use the principles and results of the theory to test those assumptions, or develop methods which are appropriate for your style of play and risk preferences.

Please don't trash probability theory; it's bigger and better than you think.
« Last Edit: March 09, 2017, 09:54:20 AM by Bayes »

The following users thanked this post: kav, Reyth

#### Jesper

• Hero Member
• Posts: 1425
• Thanked: 738 times
• Gender:
##### Re: Probability Theory is not useless
« Reply #1 on: March 09, 2017, 10:32:40 AM »
The probability of losing the bankroll is essential to know.   Like chip up and then reduce the chance of winning.

The following users thanked this post: Bayes, Reyth

#### kav

• www.Roulette30.com
• Hero Member
• Posts: 2086
• Thanked: 1077 times
• Gender:
##### Re: Probability Theory is not useless
« Reply #2 on: March 09, 2017, 12:59:23 PM »
Like I have said elsewhere Bayes, the problem is that very often, when I use the expectations of probability theory and try to compare it with actual data, people have "accused" me of GF.

Also IMO it is not true that
Quote
the theory says, given certain assumptions, that no winning system is possible
but that would start another long discussion like those we had here:
Virtual Losses and the Limits of Randomness
and here: Probabilities before and after the event
« Last Edit: March 09, 2017, 03:12:35 PM by kav »

The following users thanked this post: Reyth

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4181
• Thanked: 1442 times
##### Re: Probability Theory is not useless
« Reply #3 on: March 09, 2017, 01:50:35 PM »
I have no problem with the validity of probability theory except when people try and twist it into an absolute to fit their own view/agenda; I mean when people try to use probability theory to tell us that human interaction cannot change the results of the game (i.e. winning or losing) even though we have free choice, I consider that an outrage and am personally offended.

I agree with Talos about our sessions being too short to consistently get a reliable sampling of probability in a specific bet selection.  The short run <> the long run as Dobble says.

Talos is speaking partially in code, his full meaning is hidden from view behind his intentionally clever mode of expression.

Its almost as if he is being humble when he says that bet selections do not matter because what he means by extension is that he has discovered a method that goes beyond any individual bet selection (i.e. that applies to the entire felt at once) -- something that is really genius.

Does this mean he ignores the felt entirely?  No, but he is saying that we CAN do so and still win.  It is clear that he is paying attention to the felt geographically because he plays High/Low as his EC of choice and I find that fits very well with his subsequent heavy use of double Dozens.  Its also clear that he pays attention to the felt to remove the necessity of decision making and to avoid betting mistakes:

Quote from: Dr. Talos
CANON 30

I have a way to select what to bet, in order to avoid confusion or indecision. If system tell me to play a couple of dozens (for example, is not the way how it works...) I have a pattern to chose which dozens are to use. What you choose doesn't affect the final result, as long as you bet two dozens!
This "pattern" helped me testing the system without tweaking results. I tried different patterns with the same spin outcomes, and worked exactly the same.
A winning system must work in every situation, or is not a winning system. If you need to check late, hot, or lucky numbers, is not a winning system, it is a strategy

I am trying to wrap my mind around this concept to the point to where I can write a simulation that demonstrates how his ratios ubiquitously function across the felt and that is the perspective that is required, to look at the whole board at once, continuously:

Quote from: Dr. Talos
when I say selection is not important, I exactly mean that no matter what is your pattern, your way to choose the right number/area/color/section has no influence in the final outcome.

When I started, I took 2000 spins I witnessed in a casino. I selected a pattern and I used my system (for a better understanding, let's say a used the number 1 as a pivotal point of my system). It worked. then I used a different pattern, and then another (let's say number 2, and then number 3... finally 36). In every occasion worked, even when a number, a split or a street skipped for two hundreds spins! Sometimes I profited more some other less, naturally, but at the end I was winning no matter what.

He has already surpassed 100K personal spins some time ago and there is no way to know how many spins he manually tested but he has the character type of an accountant; obsessively meticulous.

EDIT: It took some doing but I have come up with the idea for the simulation required; I just have to code it.  It should be quite enlightening!
« Last Edit: March 09, 2017, 05:17:38 PM by Reyth »

#### BlueAngel

• I always express my opinion
• Hero Member
• Posts: 1571
• Thanked: 243 times
• Gender:
• Do you want truth? You cannot handle the truth!
##### Re: Probability Theory is not useless
« Reply #4 on: March 09, 2017, 02:58:22 PM »
In my opinion probability theory is good basis to set the borders in which a person could create a pack of parameters.
Conditions and parameters may or may not change from time to time, probability theory is like the logic, useful but not sufficient enough to create a long term winning method.
This is where all comes down to individual in order to use creatively his/her mind ''canvas'', and  paint his/her own masterpiece within the probability's frames.

I'm amazed how ignorant remain some persons who pretend to be experts but make such elementary mistakes such as assuming that the long term is something completely irrelevant from the short term!
Short term is a part of the long term, as 1 is a part of 100, they are the very same persons who believe in hit and run as viable solutions for long term profits.

I have decided not to stick my head in the sand and naively believe that I can fool probability but in reality fooling myself.
It's really hilarious how some poor gamblers are trying to re-invent the wheel by wrongfully assuming that their observations are more accurate than the probability theory itself...

For example after a few thousands observations they arrive to the conclusion that the average hit of a dozen is 1 out of 9 instead of the 1 out of 3, so in their own mind are more accurate than probability, perhaps they deserve the nobel prize with such revolutionary ideas!

Then knowledgable, persons like user real, are becoming the bad guys when they are trying to say the obvious.
I'd prefer 1000 times more to learn something from users like Bayes, Real and Kav and not Reyth's fallacious idols! (dr Talos is not one of them)

Unfortunately this forum is infected by inaccurate assumptions, commonly known as gambler's fallacy and the most ludicrous of them all is the number of their followers, that's why the majority of gamblers are long term losers.

#### Bayes

• Moderator
• Veteran Member
• Posts: 688
• Thanked: 563 times
• roulettician.com
##### Re: Probability Theory is not useless
« Reply #5 on: March 09, 2017, 05:42:53 PM »
Like I have said elsewhere Bayes, the problem is that very often, when I use the expectations of probability theory and try to compare it with actual data, people have "accused" me of GF.

Without a specific example it's hard to comment, but you have to agree that GF is a common mistake (not saying you're guilty of it, BTW).

Quote
Also IMO it is not true that
Quote
the theory says, given certain assumptions, that no winning system is possible

There are no absolute probabilities; they are always conditional. If the conditions are present then the result must follow, but of course that doesn't mean the conditions are necessarily present in reality.

Quote
I agree with Talos about our sessions being too short to consistently get a reliable sampling of probability in a specific bet selection.  The short run <> the long run as Dobble says.

So the chance of winning a bet on 30 numbers is the same as the chance of win when betting on one number, if you're only betting for a few spins? That's what you're saying. DrTalos didn't say anything about a "reliable sampling", but that "roulette is not a probability game", which is nonsense. He himself admitted that the search for his system would have been much shorter if he didn't have to use trial and error.

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4181
• Thanked: 1442 times
##### Re: Probability Theory is not useless
« Reply #6 on: March 09, 2017, 06:03:17 PM »
So the chance of winning a bet on 30 numbers is the same as the chance of win when betting on one number, if you're only betting for a few spins? That's what you're saying. DrTalos didn't say anything about a "reliable sampling", but that "roulette is not a probability game", which is nonsense. He himself admitted that the search for his system would have been much shorter if he didn't have to use trial and error.

He is obviously referring to the effects of variance as you mentioned earlier but you seem to be taking the actual meaning behind his comments a slight bit too far to infer that he means that probability doesn't apply in any way.

His meaning (and he admits his english is less than precise for which he apologizes) is that the variance is too large in our short sessions to obtain useful results soley based on probability; in other words, he has failed enough to know that relying solely upon probability to win surrounding a specific bet selection is a losing proposition.

He doesn't throw probability completely out of the picture, just reliance upon it in conjunction with a specific bet:

Quote from: Dr. Talos
CANON 52

As I already said, the selection is not important. If with a pattern of bets I would have win at the fifth spin, with another would have been at 7th.
The average hit/number played is what I rely upon. 1/60 can be seen not enough, in a short term, but I play a lot of numbers for a lot of spins, so sooner or later I will get it, because "mathematically" the average will tend to be 1/38, am I wrong?

There is a subtle genius here that is easy to gloss over and reject and many have done this.

It however is very interesting that you mention series of bets regarding his system & philiosophy because I believe this is a great unspoken wisdom about how he plays.

If bets on a series of smaller bet selections can be made in less expensive chunks, it can be more advantageous to make those bets instead of bets on larger bet selections but with more expensive bets.

The overarching point is that our attachment to a single bet selection entraps us, limiting our ability to maxmimize our cost:risk & payout:cost ratios, hence the need to decry bet selections with over-reliance on probability because we won't defeat variance that way.

Its like we are addicted to thinking: "Wow this neat betting pattern, if I could just get it to work better, I can use a trigger and a raising schedule, I can do it!"; he is trying to break that spell that roulette has over us and get us to stop thinking that way.

I realized this moments ago when I noticed that all the quads go down the center (except the zero quad) and I started dreaming about waiting for a series of hits in CL 1 & CL 3 and then betting etc. and my mind just snapped when I realized what I was doing!
« Last Edit: March 09, 2017, 11:49:57 PM by Reyth »

#### scepticus

• Hero Member
• Posts: 2401
• Thanked: 521 times
##### Re: Probability Theory is not useless
« Reply #7 on: March 10, 2017, 06:19:30 PM »
Reyth
As all quads contain the middle column might it be possible to leave out a particular column ? Say the First or Third ?

I use Probability Theory but am not getting involved in this controversy but merely restate my belief that it does not produce an " absolute " and is therefore a "Guess". An educated guess is still a guess whether you phrase it as a Prediction  or Forecast.
As mathematicians tell us that we can win in the short term but will lose in the long term  then it seems that they agree that short term probabilities  differ from long term probabilities  - just as Dobelsteen .Palestis , harry j. and Reyth  claim .

The following users thanked this post: Reyth

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4181
• Thanked: 1442 times
##### Re: Probability Theory is not useless
« Reply #8 on: March 10, 2017, 06:33:55 PM »
Well I don't believe this method will succeed and won't be pursuing it but my "old self" or "instinctual self" would want for the middle column to be sleeping for a maximum amount of time and then bet 2-3 "john wayne flowers"?  I mean something like that so that we aren't just throwing the flowers out there, we have a reason to bet them and do not bet every spin?

We can see that the Dozens can also get involved in the placement of the flowers.

Even though I think this is doomed to failure, I think its better than thowing flowers out there to bet every spin.

Its fun thinking of these too even though I think its roulette trapping us like the Sirens:

« Last Edit: March 10, 2017, 06:42:49 PM by Reyth »

The following users thanked this post: kav

#### BlueAngel

• I always express my opinion
• Hero Member
• Posts: 1571
• Thanked: 243 times
• Gender:
• Do you want truth? You cannot handle the truth!
##### Re: Probability Theory is not useless
« Reply #9 on: March 10, 2017, 06:52:51 PM »
Nothing is certain in short and long term, you may place a single bet, win and get lost like kid who snatches a chocolate and as easy as it seems to win one or few units, equally possible is to lose the same amount.
You might try to make it harder to lose by risking more money and indeed you would lose less times but more money.
Hit and run was never the solution and it'll never be.

Finally, a friendly tip for you Scepticus;
Do you know why no-one understands your 9 columns? Because it's NONSENSE!
No matter which combination of outcomes you observe, you can never discard any possibility simply because it still occupies its position inside the wheel.

There are not ''normal'' and ''paranormal'' results, only frequent, rare and everything between them.
« Last Edit: March 10, 2017, 07:19:34 PM by kav »

#### scepticus

• Hero Member
• Posts: 2401
• Thanked: 521 times
##### Re: Probability Theory is not useless
« Reply #10 on: March 10, 2017, 07:49:36 PM »
Actually, BA, I am not at all bothered if you, or anyone else . think it is nonsense. This was devised by mathematicians . If the winning number is not zero then it must be in a column and a dozen . As you progress in spins then the probabilities lessen. I think that helps - you  don't . Our opinions differ.

#### BlueAngel

• I always express my opinion
• Hero Member
• Posts: 1571
• Thanked: 243 times
• Gender:
• Do you want truth? You cannot handle the truth!
##### Re: Probability Theory is not useless
« Reply #11 on: March 10, 2017, 07:53:10 PM »
Quote
the probabilities lessen

Probabilities are constantly stable.

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4181
• Thanked: 1442 times
##### Re: Probability Theory is not useless
« Reply #12 on: March 10, 2017, 08:00:17 PM »
To cut down on the length of a repeat argument, I think its clear that both of you are right but that Scepticus is referring to the successive probabilities within a series of bets; i.e. the indisputable fact that successive streaks diminish.  Nobody ever disputes the odds of a single spin but just as valid is the odds of a series of spins.

The following users thanked this post: scepticus

#### BlueAngel

• I always express my opinion
• Hero Member
• Posts: 1571
• Thanked: 243 times
• Gender:
• Do you want truth? You cannot handle the truth!
##### Re: Probability Theory is not useless
« Reply #13 on: March 10, 2017, 08:30:21 PM »
I'm going to give an example for you to understand better what I'm trying to say;

Let's say Reyth plays his favorite system which uses the trigger after 4 reds/blacks to bet the opposite, Reyth sees the opportunity and thinks he has much more chance to win now but me who has just sit on the table I'm not aware of what happened before so I just bet black which has happened the last 4 spins.
Me and Reyth have placed 18 numbers bet but only one of us can win, not both.
In case either of us wins, the stake will be return plus one unit, the problem is that Reyth thinks he has more than 95% probability while the foolish casino would pay 100% of his unit value.
Indeed is such a bargain, what are you waiting for?!
From the other hand I'm betting (not observing) for a parlay in power of 4, in order to win 1+2+4+8=15 units thus risking 1 to gain 15, this is the true probability which reflects money wagered ant not outcomes observed.

By observing you don't only save loses but you also miss wins, therefore no matter how many observations you had in order to use them as triggers, the ratio between risk and reward remains the same!
You cannot sidestep probability, just the facts!
« Last Edit: March 10, 2017, 09:15:17 PM by Reyth »

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4181
• Thanked: 1442 times
##### Re: Probability Theory is not useless
« Reply #14 on: March 10, 2017, 09:14:10 PM »
By observing you don't only save loses but you also miss wins, therefore no matter how many observations you had in order to use them as triggers, the ratio between risk and reward remains the same!

The trigger issue has been first brought up by you here.  We have only mentioned the odds of a series of spins which is undeniable.  We don't need GF arguments which is the rest of your post.