+ Register

[Drazen]

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.

True, and actualy this is the biggest problem here. If one can solve this engineering problem efficiently then he has beaten the game.
Without this efficient solution, RTM means well, pretty much nothing. 


Regards


Drazen

[sqzbox]

Thanks Bayes - well put and a nice article you referred us to.  I admit that I struggle a little when trying to differentiate in my mind between RTM and GF - it is the on-going struggle that helps the mind grasp and integrate the concept.  As I said earlier though, I do believe that RTM exists (since it is a mathematical certitude) - just generally not manifesting in our convenient time frame.  Help me with the following scenario.  I will try to be very careful in my wording so as to minimise any possible ambiguity.

We will have 200 coin tosses.  We will examine them in 2 groups - the first 100 and the second 100.  In the first 100 we see 90 heads. Would you expect the next 100 to comply with the normal probability distribution or to weigh in heavily on the tails side?

[Albalaha]

It is very much possible to get 600 heads vs 400 tails in 1000 trials/tosses. A coin has no memory to balance things out. 500 heads and 500 tails are merely rule of probability. Randomness knows no rules.

[Bayes]

sqzbox,

90 heads in 100 coin tosses is 8 standard deviations below the mean, which is pretty much impossible GIVEN a fair coin. If I actually saw that happen I would assume the coin was biased and bet heads in the next 100 tosses, but I assume you merely want to make the point that the first measure was an extreme one.

Since the mathematical expectation NEVER changes, the most likely scenario is +/- 1 std over the next 100 tosses, meaning between 45 and 55 tails (around a 68% chance of this). I agree, it can be tricky to understand the difference between GF and RTM, maybe because we're so used to thinking in terms of cause and effect.

It might help if you think about a roulette bet which has a very high chance of winning, such as covering 36 numbers on a single zero wheel. If you were to make a series of such bets and recorded a "W" each time one of the 36 numbers hit and a "L" each time the uncovered number hit, the sequence would have very few consecutive "L"s. It's more likely that a "W" would be followed by another "W" than a "L", but you don't have to invoke any prior losses in order to see that it would be true, because you can see intuitively that the higher the probability of a SINGLE bet winning, the more likely it is that you will get long winning runs and few consecutive losses. Extreme events are by definition rare, so it follows that a win (a non-extreme event) is much more likely to be followed by another win, and a loss is more likely to be followed by a win.

[Drazen]

sqzbox,
Extreme events are by definition rare, so it follows that a win (a non-extreme event) is much more likely to be followed by another win, and a loss is more likely to be followed by a win.

[Bayes]

It is very much possible to get 600 heads vs 400 tails in 1000 trials/tosses. A coin has no memory to balance things out. 500 heads and 500 tails are merely rule of probability. Randomness knows no rules.

albalaha, I wouldn't say it's "very much" possible. That scenario represents 6.3 STD below the mean, which would be incredibly rare, if not impossible. I would certainly have doubts about the fairness of a coin which generated those results. If randomness knows no rules then why not 200 head in 1000 tosses?, or 50?

[Bayes]

[Drazen]

Can this be concluded it is better to attack  just after W? If lose wait for another win and attack again? But again, where is bettors judgment here, which superman and you are mentioning, this can automated, right?


Cheers


Drazen

[Ralph]

Many methods win most of the time, and it is just "rare" events which kills it. The "rare" events use to come a bit faster than we expect. 


High bankroll and progressions stand until a "rare" events happen.

[Superman]

Can this be concluded it is better to attack  just after W? If lose wait for another win and attack again? But again, where is bettors judgment here, as superman and you are mentioning, this can automated.

Speaking for myself only, the W would have already been my bet, IF my prediction is correct, I'll use streaks as an example again, if I thought my indicators were saying the next red hit will be a streak not a single event, I would bet once for that to happen, if it was a single event again, I would wait for the next red and bet red again, lose? then I would repeat the wait, this would be my 3rd attack on that specific trigger, if the 4th bet lost I would still keep the trigger alive, waiting for the next red but just flat bet ignoring the last 3 losses, this is where the hole starts getting deep as THAT specific trigger was faulty but it eventually came through.

Then, with the 3 losses in mind, when my next trigger has appeared I may bet 2 units to try and recover a little of the last triggers losses, BUT if I had done this for 3 or 4 triggers and it still isn't working I would just start over, forget the losses and start from 1 unit. This is the problem for many players, the thought of being in the negative is hard to accept for most people so they try to recover as fast as possible thus risking more units.

A big problem on most gambling forums is that 95% of the members want to earn easy money from gambling in general, they've just found roulette, found a forum, read a few pages and think it's easy, it aint, there's also those that just keep going round in circles from one method to the next, there's no simple way of winning long term. I know I speak for a few people in stating that I have been testing, checking and result crunching for the best part of 6 years, only now for the last year have I been able to actually make consistent winning sessions WITHOUT stupidly huge risks, anyone can play a martingale but we know where that ends up.

In short, there is no SINGLE method and no SINGLE MoneyManagement plan to win consistently, and I don't think what I have is a method in reality, it's a way of playing the game in general.

But again, where is bettors judgment here, which superman and you are mentioning, this can automated, right?

Did you mean can't automated? or can BE automated?

I still don't think it can be automated sadly.

[Bayes]

Can this be concluded it is better to attack  just after W? If lose wait for another win and attack again?

No. As per the Marigny strategy, better to attack after a L. The loss could come at any time, but two consecutive losses are less likely than a single loss. At this point you might protest that it's GF because you're dealing with independent events and the loss which has just occurred makes no difference to future trials, and while it's true that the odds of the losing sequence haven't changed (hence you need to be cautious), it's a better entry point than merely picking one at random.

[Drazen]

Thanks mr. Bayes it helps. I have something in mind trying with this.


@ superman


True, many think that gambling could give them independent lifestyle just like that.. Oh that is very very hard to accomplish no matter which way.


Oh numbers of hours I putted in this game is measured in thousands in few years for sure, and still not 100% sure I can beat the game.


But something is interesting here my friend. The way you play can't be tested on millions of spins fast to be sure you can beat it, (as with any other systems you are ding) becasue it can't be coded as you say and pretty much all depends on your personal evaluation of the situation.  You are actualy beating the game in walk.. you never know how session will end and how long will it be. So it seems to me that you are in a way still testing and getting confidence with this by real playing with low limits, even after so many won sessions

[sqzbox]

Thanks Bayes - I was hoping that that would be the answer.  I only used 90 heads in 100 tosses because that was what the link you referred us to used as an example.  The point I was actually hoping to indicate was that after an extreme event (and I know we can't specify exactly when it is over) then normality reasserting is the most likely scenario.  Some might say that RTM would cause an overload of tails in the short term future (i.e. the next 100 spins in my example) in order to achieve statistical balance over the 200.  But here is where I disagree.  It is my assertion that the likely future is ALWAYS normality.

[Superman]

You are actualy beating the game in walk.. you never know how session will end and how long will it be. So it seems to me that you are in a way still testing and getting confidence with this by real playing with low limits, even after so many won sessions

Sounds about right BUT mentally I intend for the session to end in the plus, I have thought a few times of ending a session in the negative, temper mainly, but that would be defeat so mentaly regroup then plod on again on bad sessions this can be tough. I'm still learning though and have been playing with the idea of using 0.20 as a base bet, when I played this way before JL had my account I was playing in 0.01 units LOL

[Albalaha]

albalaha, I wouldn't say it's "very much" possible. That scenario represents 6.3 STD below the mean, which would be incredibly rare, if not impossible. I would certainly have doubts about the fairness of a coin which generated those results. If randomness knows no rules then why not 200 head in 1000 tosses?, or 50?

Bayes,
           What is the probability of getting 5 or 7 reds in a row? Don't u feel that they can appear more often in a session than the mathematical probabiility says?


See 75 Blacks, 105 Reds and 5 Zeroes.