Author Topic: Triggers in Systems  (Read 1776 times)

0 Members and 1 Guest are viewing this topic.

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3968
  • Thanked: 1278 times
Re: Triggers in Systems
« Reply #90 on: December 07, 2017, 12:44:27 AM »
In case of black/ red it doesn't really make sense to wait anything and martingale after. Tested - do not work.

Hey Mr. P, I was just thinking about this yesterday and since the LotT must provide a felted cluster that will hit greater than any other such cluster, wouldn't Low/High be a better bet selection than Red/Black & Odd/Even?  I mean RBOE doesn't take into account geographical locations on the felt, right?

Regardless, I must agree that 18 numbers is most likely too many to gain the proper statistical focus for hot numbers. 

What do you think about Dozens?   
« Last Edit: December 07, 2017, 12:48:24 AM by Reyth »
 

Fyodor

  • New
  • **
  • Posts: 49
  • Thanked: 56 times
Re: Triggers in Systems
« Reply #91 on: December 07, 2017, 01:00:53 AM »
Please pardon my "Jump-in"
If you are looking for felt based "clusters" that involve dozens, please consider the "trans-dozen on the EU wheel, that is covered by the 19-30 sequence.
Could be covered with two chips, (@2 to 1) so very economical, and nine of that dozen are (close to) perfectly distributed around the wheel.
The pattern emerges as also being useful to partition the wheel sequence into manageable neighbourhood groups.
 
The following users thanked this post: MickyP

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1358
  • Thanked: 787 times
Re: Triggers in Systems
« Reply #92 on: December 07, 2017, 01:59:58 AM »
Reyth, generally l do not think in such a terms... l used to when play bias exclusively. .. but since then my game adanced a bit..
   1st 12 numbers got nice splits/ coners/ 3numbr to cover numbers around 10 ( tiers...).
 Example is 5/8.. 10/11, 5/6 if you decide to center a bit aside... if you want insurance bet on other side, can include 4&12.. 3/0...
   But generally, less numbers not targeted you bet -better. Covering entire 12 numbers delute too much posible edge.  You should be thinking like..  take edge for best number add extra ( not included in the bet) number edge and devide by 2. If you follow this logic, you will see that it's not good idea.
 
The following users thanked this post: Reyth

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1358
  • Thanked: 787 times
Re: Triggers in Systems
« Reply #93 on: December 07, 2017, 10:58:15 AM »
BTW. .  Would be nice to have " modify button" back. I often type on the go...on the phone as well... where spellchecker do not work. It's only practical..   
 
The following users thanked this post: Reyth

Mike

  • Veteran Member
  • ****
  • Posts: 624
  • Thanked: 50 times
Re: Triggers in Systems
« Reply #94 on: December 07, 2017, 12:34:48 PM »
The problem is that you don't play roulette.  When you have skin the game it makes a difference in the way you approach it. 

Why is the fact that I don't currently play much roulette (I still play online occasionally, and have played a lot in the past) a problem? I would guess that most stats professors have NEVER played it; does that mean that their analyses are invalid? Of course not. And you yourself play RNG exclusively; some people would say that's not even roulette.

If you're talking about experience, and whether that affects the way you play and how successful you are, then I would say yes - experience can make you a better player (some people, like compulsive gamblers, never learn from their "experience").

Quote
I have saved countless thousands of BU's by letting losses go by and profited again and again because I delayed my bet.  Snap running a simulation just simply doesn't provide that perspective.

So why doesn't it show up in the simulations? You're a competent programmer, surely you could write a program which simulates the way you play on a session by session basis.

I think you're applying double standards here Reyth. You make use of simulations to test the limits of systems and generate stats which you then make use of in designing systems. You take the results (quite rightly) as representing objective truth and not mere opinion, and yet a simulation which shows that virtual bets have no merit are somehow discounted or not taken seriously.

Quote
Which is more important, a long-term theory that takes place over millions of spins or what is more likely to happen RIGHT NOW while I am spinning with money on the felt?

Please, not this again. How can you possibly defend a thesis which asserts that a short-term test is more informative and representative than a long-term one?
It's totally contradictory to any principle of statistical inference and intuitively everyone knows this. All serious system creators will test their fledgling systems manually for as many spins as they feel are necessary (which often isn't enough) before committing to real money play. They don't think "I'd better not test TOO much, because the system might fail, I'll just test while I get positive results".

Quote
I dislike theoretical discussions that have the effect of discouraging people from developing  themselves because it IS only your opinion and there is room either way for people to form their own opinion.  If you would simply approach it that way, I would have no problem.

Sure, people can form their own opinions and can also choose to ignore any evidence which contradicts their pet theories. If someone likes using virtual bets because it gives them some psychological security blanket or the added structure it gives to a system, then I guess that's a good enough reason for using them, even if it's been shown that the virtual bets don't give any actual advantage or safety. And virtual bets don't make your results any WORSE - you're not more likely to lose if you use them. The only concern I have is that they give players who believe in them a false sense of security, so they make riskier bets than they would have if they didn't use the trigger.

Casinos are well aware of this. Why do they provide leader boards and other stats like hot and cold numbers? If they gave the player any true advantage they surely wouldn't provide them. I would venture to suggest that their presence is a substantial boost to the casino's coffers.
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3968
  • Thanked: 1278 times
Re: Triggers in Systems
« Reply #95 on: December 07, 2017, 01:00:37 PM »
We are repeating ourselves here.  This is where we clearly disagree:

Just because a simulation over millions of spins shows that a trigger has no effect doesn't mean that there is no effect at all from triggers.

Roulette doesn't do things in "straight lines", it tends to "zig zag" to accomplish its statistical "work".  With a sequence length of millions of spins, there is a long-term zig-zag that occurs slowly to produce the end result.

You leap from the results of this long-term result to "all triggers are useless no matter the sequence length" which is clearly not true.

This is why we say that the fact you do not actually play roulette is a factor in the way you think.  Surely you do not play roulette systems because if you did, you could not avoid agreeing with us.

Once again, my hands are tied.  I don't have a choice, I MUST employ the method that gets me the best results on the felt, RIGHT NOW, while I am playing, PERIOD.  Its a player's perspective.

Sorry, I would comment more but I am pressed for time this morning at work. :)
 
The following users thanked this post: Mike

Mike

  • Veteran Member
  • ****
  • Posts: 624
  • Thanked: 50 times
Re: Triggers in Systems
« Reply #96 on: December 07, 2017, 01:49:54 PM »

Just because a simulation over millions of spins shows that a trigger has no effect doesn't mean that there is no effect at all from triggers.

Roulette doesn't do things in "straight lines", it tends to "zig zag" to accomplish its statistical "work".  With a sequence length of millions of spins, there is a long-term zig-zag that occurs slowly to produce the end result.

You leap from the results of this long-term result to "all triggers are useless no matter the sequence length" which is clearly not true.

Sorry Reyth, you've lost me here. If a simulation shows that triggers have no effect why does that allow you to conclude nothing? If there is such a big difference between simulation and practice, why do you run simulations in the first place? The fact that roulette doesn't do things "in straight lines" is correct, but that just highlights the need for a bigger sample, since a small sample can easily "trap" only a positive result which would be misleading. You need a larger sample to get the big picture.

I don't think you can say that it's CLEARLY not true that triggers have no effect when the only evidence we have that they DO work is anecdotal. There is no reputable source which says that they can work in theory, and no systematic empirical evidence published anywhere, so at best the issue is controversial.
 

Sputnik

  • Veteran Member
  • ****
  • Posts: 534
  • Thanked: 453 times
Re: Triggers in Systems
« Reply #97 on: December 07, 2017, 03:20:50 PM »


 @ Mike i tell you why i collect data and make simulations of different betting strategy.

For example the horse market where the odds change based upon how the public bet.
That way you get favorit horse and others with higher odds.

When i collect data and make my observations i can see that the favorit does not has the strike ratio that it should have if the public was right. So then i want to make a benchmark to see how correct the public market is when they put money on a horse and how the odds reflect the true chance for a horse to win.

So around 68.3% the market should pick the right horse within 1 SD
So around 95% the market should pick the right horse within 2 SD
So around 99.7% the market should pick the right horse within 3 SD
So around 3% the market should not pick the right horse within 3 SD or lower.

This is not a true playing model using the horse market with the odds that has 50/50 situation to win.
So i want to know the values look like and what kind of benchmark i can expect playing a favorit horse.
How many times has the market right and how many times has the market wrong.

So i get a swings measuring the overrepresented hits and the underrepresented hits.
For example a horse with 2.0 odds will most likely not get a strike ratio or sequence around 14 wins and  2 loses

When i collect this strike ratio or sequences of hits and no hits i get waves to explore and trigger when to bet and when not to bet.

This also work with higher odds as i can use binomial probability calculation.
For example 3.0 odds has a 50/50 winning situation within two attempts.

This is the reason i collect and simulate data.

Cheers
 
The following users thanked this post: kav, Reyth

MickyP

  • Mature Member
  • ***
  • Posts: 199
  • Thanked: 132 times
  • Gender: Male
Re: Triggers in Systems
« Reply #98 on: December 07, 2017, 10:47:52 PM »
I'd like to float back down to earth for a moment.
I do understand the need to test a trigger, a system, a strategy or an approach but I can't see the logic in having to over extend the length of a test. These million plus spin tests are like tapping softly on a piece of glass, it will eventually break. The result will be a warning to everyone not to tap on glass as it will break.

How many spins can occur in a 24 hour period at a B/M table in an average casino. Multiply this number by 10 or 20 and that should be a sufficient number of spins for a test.
What is the actual aim of a million plus spin test? Is it to prove that roulette can be beaten or is it for you to determine how much bankroll you need to survive these millions of spins?
There are many bankroll defences that players include in their game.
I agree that triggers do not always work and I agree that we at times read the game wrong and yes at times we loose. These are the facts of the game. Everything that can be tested is tested to the extreme yet one all important ingredient in the game is overlooked. The state of the players mind at the time of play. This can alter the outcomes of the game drastically.
The human factor will always be more unpredictable than the game itself because we are not all designed identically. Our thoughts do not function in harmony to the greater good of defeating the game.
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3968
  • Thanked: 1278 times
Re: Triggers in Systems
« Reply #99 on: December 07, 2017, 11:01:02 PM »
Millions of spins tests allow me to see what is the worst that I am likely to experience.

Usually the rarest downward swings will not even show up in a simulation before 1M spins have occurred.  However, this doesn't mean that as long as we play less than 1M spins we will be fine; one of these events can occur right from the first coup attempt in our first session.

I say that it is "clearly" not true because of the empirical results that Pales, Kav, myself and I am sure very many others achieve.

I would definitely agree that the topic is controversial however and that is because over millions of spins, the trigger advantage is nullified; I don't go as far as to say that we ignore this, I just defer this because I don't have a choice.

So, what does it mean that I defer it?  I don't know.  All I know is that I have reaped alot of profit and that it is going to take much more correction than I have yet seen to reverse that.

Even Bayes uses triggers (even though he too despises them); he called his a "fuzzy trigger".  Most of my triggers are not simple and represent convergence of multiple statistics (a favorite tactic of Bayes as well).

Once again, I agree that the discrepancy should be investigated.  If we look at the statistical output, with some hard work, we should be able to identify what causes the neutralization and maybe understand what we should "expect next".

Because I have discovered that roulette changes its focus with every new session, I also think testing can be done with properly resetting the random seed to simulate a session reset.  My session resets are not simple either and occur for several different reasons.
« Last Edit: December 07, 2017, 11:07:48 PM by Reyth »
 
The following users thanked this post: kav, MickyP

MickyP

  • Mature Member
  • ***
  • Posts: 199
  • Thanked: 132 times
  • Gender: Male
Re: Triggers in Systems
« Reply #100 on: December 07, 2017, 11:02:47 PM »
I understand that in these million plus spin tests you are looking for a bad series of spins long enough to prove failure. If you have to use this many spins to prove failure then how many spins do you require to prove success?

 

MickyP

  • Mature Member
  • ***
  • Posts: 199
  • Thanked: 132 times
  • Gender: Male
Re: Triggers in Systems
« Reply #101 on: December 07, 2017, 11:14:36 PM »
Reyth , Mike posed the question to you regarding triggers being neutralised after millions of spins of testing yet they work in shorter test batches. Is the trigger neutralised due to the ratio of trigger hits to misses over these extended tests?
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3968
  • Thanked: 1278 times
Re: Triggers in Systems
« Reply #102 on: December 07, 2017, 11:57:03 PM »
The way I discovered that triggers have no effect on the long-term results is by running a system over 16 million spins and noting the win rate.  Then I add in the trigger and the win rate doesn't change. 

Maybe one of the reasons this happens is because eventually the rarest statistical results begin to cluster enough nullify the advantage gained over the entire sequence length.  I think it takes a very rare set of results for this to happen which usually wouldn't be seen in smaller samples.

I guess eventually at some point, we can get hit with a very rare downswing of successive losses before we will be able to recover.

Does this prove that triggers don't work?  Not in my mind.  For one, I don't believe that a single string of 16M spins is an adequate test of a system; unless I force the sequence to reset when I say it should, I am not satisfied with the results.  Secondly, I don't believe that just because some testing of certain types of triggers show no change in the results, means that we are helpless and just throw up our hands and give up.

What our critics continuously fail to admit is that it is ALWAYS POSSIBLE to bet correctly and not lose.  ITS POSSIBLE.  That is enough to remove the false air of "objective fact without question" from their arguments even though they refuse to allow that fact to get in their way.

They appeal to authority and appeal to ignorance but none of that changes the FACT that it is always possible to win in roulette using a system. 
« Last Edit: December 08, 2017, 12:21:56 AM by Reyth »
 
The following users thanked this post: kav

MickyP

  • Mature Member
  • ***
  • Posts: 199
  • Thanked: 132 times
  • Gender: Male
Re: Triggers in Systems
« Reply #103 on: December 08, 2017, 06:40:36 AM »
Thanks Reyth . Did you discover the neutralisation of the trigger in only one test or did you run several tests that produced the same results?
 

Mike

  • Veteran Member
  • ****
  • Posts: 624
  • Thanked: 50 times
Re: Triggers in Systems
« Reply #104 on: December 08, 2017, 09:43:43 AM »
Reyth,

I don't deny that is POSSIBLE to win using a system, only that virtual betting is ineffective. It's not really fair to say that "we" make the fallacy of appealing to authority. Everyone appeals to authority; when I visit my doctor I trust that she knows what's she's talking about. We all have to rely on experts. Appealing to authority can be fallacious when 1) The authority is not an expert on the matter, 2) the authority is an expert, but is biased, 3) the authority is an expert, but his opinion is unrepresentative of expert opinon on the subject (it's always possible to find an expert who supports your view). But these cases don't apply in any of the experts we cite. Mathematicians and statisticians are overwhelmingly of the opinion that roulette is a game of independent trials. To claim otherwise seems perverse. Besides, it's not as though we ONLY appeal to authority.

As for appeal to ignorance, strictly speaking, you're correct. I can't really say "there is no evidence that triggers work, therefore triggers don't work". It's also true to say that there ARE situations where it's a least conceivable that there is some real (rather than imagined) dependency going on. However, the arguments that are made in favor of a universal kind of dependency are based on a misunderstaning of what statistical independence means (like saying that "no one ever sees 100 reds in row, therefore there is no such thing as independence"). And the gambler's fallacy involves a straightforward contradiction, so it's not a matter of finding some empirical evidence that it's not false.

Quote
The way I discovered that triggers have no effect on the long-term results is by running a system over 16 million spins and noting the win rate.  Then I add in the trigger and the win rate doesn't change. 

I think it would be more informative to record not only the number of busts, but where they occur in the sequence. If you and others are right in believing that triggers do make a difference, there might be some pattern in where they're successful and where they're not, but you wouldn't notice this just by looking at the long term win rate.
 
The following users thanked this post: Reyth