+ Register

[Albalaha]

Hey JL,
             Please do not take my words otherwise but you should stop making claims relating to July. It is getting irritating and funny at times.
You may win a million with PB till July while other may still lose big with it. There is nothing eternal in PB that can shake the gambling world. It is just a trial and error and only as good as any other method. I believe in tests and not in words and I am afraid that tests say absolutely opposite of your claims.
               You are doing well with your methods and contributing very well in your own capacity. I deeply appreciate that. For haven's sake stop all these.

[spike]

Hey J It is getting irritating and funny at times. There is nothing eternal in PB that can shake the gambling world.  I believe in tests and not in words and I am afraid that tests say absolutely opposite of your claims.

Don't worry, he did the same thing in 2010 by setting a date
far in the future. That day came and went in 2010 and he never proved
any of the stuff he said he would.  There's nothing to prove.

[JohnLegend]

Hey JL,
             Please do not take my words otherwise but you should stop making claims relating to July. It is getting irritating and funny at times.
You may win a million with PB till July while other may still lose big with it. There is nothing eternal in PB that can shake the gambling world. It is just a trial and error and only as good as any other method. I believe in tests and not in words and I am afraid that tests say absolutely opposite of your claims.
               You are doing well with your methods and contributing very well in your own capacity. I deeply appreciate that. For haven's sake stop all these.

Albalaha, im not talking just about PB. Its the weakest of my methods. But for a mere 7 unit buy in. It was never meant to be supernova proof.

You don't put all your eggs in one basket. I use up to 6 methods to make my money. Im not doing this to prove anything other than I CAN WIN.

Spike will be talking me down forever. But what exactly has he proven?

The time for talking success is long over. You have to show it, prove it now. 10% of whatever I make from my chalenge will be donated to this forum. Victor will be very happy about that in the longrun.

[Albalaha]

I appreciate your spirit JL but you are mature enough to stop these debates. Better concentrate  your energy and experience to come up with better methods.

[JohnLegend]

I appreciate your spirit JL but you are mature enough to stop these debates. Better concentrate your energy and experience to come up with better methods.

My talkings done Al, I've already got the methods. The results will show what they can do from here.

[Bally6354]

Let's talk about advantage play for a moment. I think we all understand what that means. For my purposes here let's limit it to the ecart - that is, exclude anything that relates to the physical nature of the medium such as table or wheel bias.  What sort of advantages do you think we can find if we look into the  results as they unfold?  Gizmo has mentioned patterns and recurrence of past sequences, Bayes mentioned statistics, and JL some other method yet to be disclosed.  Remember - there has to be a REASON why the advantage exists.  Hence it should be measurable.

You would certainly know if your profits keep climbing. There are many tests that you can conduct to get an idea if your winning run is down to some kind of skill or just pure unintelligent luck!

There is a concept in statistics called 'chance of randomness' (COR). This is an exact statistical measurement of how 'lucky' a series of events are.

Doing something that has only a 1 in 30,000 chance of happening purely by luck is considered 'proof' by most experts that skill and not luck was involved.

Something like this is certainly a good starting point.

cheers.

[sqzbox]

Well, I was kinda hoping that measurable means that it can be simmed.


But in any case, maybe if I give you an example -


what do you think is most likely going to happen in the 20 spins following a run of 20 reds in a row?


OR


if, after 24 spins you have seen 18 unique numbers what is the likely scenario for the next 14 spins?

[Bally6354]

Well, I was kinda hoping that measurable means that it can be simmed.


But in any case, maybe if I give you an example -


what do you think is most likely going to happen in the 20 spins following a run of 20 reds in a row?


OR


if, after 24 spins you have seen 18 unique numbers what is the likely scenario for the next 14 spins?

I couldn't answer that sqzbox. I operate on a spin by spin basis.

I know we are not talking about AP play here.

However.......

Card Counters use the true count for the next hand.

Roulette Computers predict for the next spin.

I really don't think there is a way anybody could predict with any accuracy what the next 20 spins are going to bring unless you take bias into account!

I would love to be wrong on this however. Maybe this is where players who use some kind of regression to the mean strategy  could help out!

[Blood Angel]

OR


if, after 24 spins you have seen 18 unique numbers what is the likely scenario for the next 14 spins?

you will, on average, get 4 repeats in the next 14 spins.

[Ralph]

you will, on average, get 4 repeats in the next 14 spins.

Maybe! The average is not to much help.  You get an average red every second spin, you get a dozen every third spin, number 4 every 36 spin (on NZ). It is not easy to take advance of such knowing.


Over a range of much more spins it may be something into it.

[Bayes]

Well, I was kinda hoping that measurable means that it can be simmed.

But in any case, maybe if I give you an example -

what do you think is most likely going to happen in the 20 spins following a run of 20 reds in a row?

The same thing which is most likely regardless of what has gone before - somewhere between 8-12 blacks. There is no voodoo about regression to the mean, the next sequence doesn't balance BECAUSE the previous sequence was extreme, there is no law of cause and effect in operation here, and the effect is strongest in a purely random process. It works in both directions -

In statistics, regression toward the mean (also known as regression to the mean) is the phenomenon that if a variable is extreme on its first measurement, it will tend to be closer to the average on a second measurement, and—a fact that may superficially seem paradoxical—if it is extreme on a second measurement, will tend to have been closer to the average on the first measurement. - Wikipedia.

[Drazen]

But mr. Bayes what about exploring of this phenomena in roulette in practice? It isn't so simple as it might seem. It isn't just enough to start betting after 1st extreme is measured hoping that you won't have severe deviation in same direction. It can happen, and it will happen eventualy, and the true is that just after 2nd extreme, things might start go evened (but still not correcting in great ratio) and no MM can handle that...


But again if you take a short term indication to start the attack like if you see few wins in last few spins, you can also get in black serie of losses after that... How do you know deviation stops to grow and it is distributing normal or going back slight in back?


I think that is the main question and hardest thing to accomplish? It can happen that everytime you enter in the cycle thinking  few wins indicate that correction, black serie for starts again. You stop, waiting again such thing, but it happens again, just when you entered, which takes you with good number of losses already.


That happened to me while practicing this thing.. So what is your advice?


Cheers


Drazen

[spike]

There is no voodoo about regression to the mean, the next sequence doesn't balance BECAUSE the previous sequence was extreme, there is no law of cause and effect in operation here, and the effect is strongest in a purely random process. It works in both directions -

This truth will go right over they're heads, Bayes. You violate
the first law of pseudo science: Everything is true because
I say its true. This allows them to ignore proven facts, and
continue to march to the tune of the losing drummer.

Applying cause and effect to random outcomes is a huge
fallacy that trips up most system designers. A and B just
happened, C can't be far behind. So they take examples
of where that happened and ignore those where it didn't.
Logic won't work in roulette, a very hard lesson to learn.

[sqzbox]

Considering the 20 reds scenario, if the run is over, and there is no real way to determine if it is or not, but assuming it is, then the next 20 spins will be "normal" 95% of the time.  If the run is NOT over then I think I can safely say that we are closer to the end of it now than we were when it started. 


As so well-stated by Bayes, there is no cause and effect in operation here.  It is my belief that in this short time frame no regression to the mean will be visibly effecting. This is why I am of the belief that regression is not a tool that we should be attempting to use - it just doesn't happen in our time frame.  Attempting to capitalize on regression is, to me, a prime example of Gambler's Fallacy.


Considering the other scenario, the statistical expectation is 6 new numbers and 8 repeaters.  Naturally, those figures are the mean.   It would be more accurate to say the following.


When there have been 18 unique numbers at spin 24 then at spin 38 there will be -
a 92.5% chance of 4 - 8 new numbers (mean of 6), and
a 92.5% chance of 6 - 10 repeaters, mean of .


I am of the opinion that what we are all about in this game IS trying to predict - THE LIKELY FUTURE.  I capitalize it because I want to make sure you understand that I am NOT saying predict THE FUTURE.  But it seems to me that it is not unreasonable to try and predict the likely future.  We do it all the time.  I'll give you some predictions of the likely future right now - the sun will come up tomorrow; at my location we will see a new moon on the 12th Jan; if I run onto the motorway during peak hour traffic there is a greater than 99% probability of being hit by a vehicle (morbid!); and so on and on.  Our entire life is governed by our predictions of the likely future - these predictions keep us and our families safe and, hopefully, advance our life quality.  Why should it not be reasonable to do the same with roulette?

[Bayes]

As so well-stated by Bayes, there is no cause and effect in operation here.  It is my belief that in this short time frame no regression to the mean will be visibly effecting. This is why I am of the belief that regression is not a tool that we should be attempting to use - it just doesn't happen in our time frame.  Attempting to capitalize on regression is, to me, a prime example of Gambler's Fallacy.

Well, I think it's important to differentiate between Gambler's Fallacy and RTM. One is a fallacy, the other isn't. Here's a link you might find interesting. I would submit that attempting to capitalize on RTM may be an engineering problem, but it isn't a fallacy, although admittedly, in practice the distinction between them isn't so clear cut. I think many try to avoid anything that smacks of Gambler's fallacy - no-one wants to believe a fallacy, right? The trouble is, trying to capitalize on trends is also a fallacy (the inverse gambler's fallacy), as is trying to "beat random with random" and any other form of bet selection. In fact "bet selection" in roulette is an oxymoron; according to the mathematicians, scattering a handful of chips on the table with a blindfold on is just as good as any other "selection". I think that's nonsense, and outcomes do actually converge to their expected values more quickly than you might think.

In another thread you said something about how combinatorics might help to find good bets, giving an example of R/B streaking while more often that not, H/L isn't, or something like that. IMO this is a fruitful approach, and one I use myself. Take R/B for example, I don't just see R/B, but look at the data stream under multiple aspects, and these aspects aren't just "seeing patterns where none exist", but are objective patterns which must conform to probability. e.g., there are as many isolated singles as there are series, there are as many series of 2 as there are series > 2, etc. These are all implied by the fact that the bet has an even chance of winning. There is nothing special about R/B when viewed as such, and all the other relationships must conform to the "laws" of probability such that none goes "off the chart" as it were, which means that there are certain constraints because all the patterns are connected in so far as they all come from the same source. This is what is meant by saying that "random has limits".