Author Topic: Testing systems, by hand, on small samples of spins is a complete waste of time.  (Read 5376 times)

Real

  • Fighting the war on absurdity one foolish idea at a time.
  • Hero Member
  • ******
  • Posts: 1201
  • Thanked: 120 times
  • Gender: Female

Testing systems, by hand, on small samples of spins is fun, but in reality it's a bit absurd.  If you're doing it for entertainment, then have at it.  However, real research requires careful controls.

There is no way by which the player can derive any meaningful entry/exit points, in order to optimize play... when hand testing on just 30 or 40 spins here and there.
Attempting to do so is called "curve fitting".   You're not going to find "exquisite symmetry" or "analytical combinatorics"anymore than you will find real bears and unicorns in the clouds.  You also can not determine what the best "triggers" or "entry and exit points" should be.
 
Curve fitting is a fools folly.  Meaningful testing takes place over tens of thousands of spins and hundreds of thousands of spins.  Such testing requires careful controls.
 
If you want to test...
 
1. Make your observations
2. Form a hypothesis
3. Make a detailed prediction and carefully define the  test.  This is to limit curve fitting and the degrees of freedom.  Periodically changing the rules along the way will lead you back to the "fools folly".
4. Test a statistically relevant number of spins using a computer.
5. Conduct additional out of sample testing to see if you can repeat your results.

Best of luck,

Real

 
« Last Edit: February 26, 2015, 05:14:44 PM by Real »


 

kav

  • www.Roulette30.com
  • Administrator
  • Hero Member
  • *****
  • Posts: 1531
  • Thanked: 580 times
  • Gender: Male
Hi Real,

Good advice.
Trying a roulette system on a small sample of spins and especially a sample of unfavorable spins can help you understand the system, its advantages and disadvantages. Definitely it cannot tell you how the system will perform in the casino. That's why I'd avoid the term "test" for small samples of spins. 
 

palestis

  • Veteran Member
  • ****
  • Posts: 547
  • Thanked: 273 times
Hi Real,

Good advice.
Trying a roulette system on a small sample of spins and especially a sample of unfavorable spins can help you understand the system, its advantages and disadvantages. Definitely it cannot tell you how the system will perform in the casino. That's why I'd avoid the term "test" for small samples of spins.
That may be true. But how about this?
 If a system depends on small number sequences, or random rows, and most of them are, then testing smaller samples  thousands of times in many different number collections, should yield to a definite conclusion. If a system depends on the last 20 numbers spun, then testing a different group of 20 numbers thousands of times, will tell you with certainty what to expect next time you are facing 20 numbers.
Unfortunately I have seen  many players looking at 150 number score cards, and coming up with a winning system in 15 minutes.   
Then when they lose next time they are at a table, they are wondering what went wrong.
 

Real

  • Fighting the war on absurdity one foolish idea at a time.
  • Hero Member
  • ******
  • Posts: 1201
  • Thanked: 120 times
  • Gender: Female
Quote
they are wondering what went wrong.

The answer is that nothing went wrong.  They were simply facing "randomness" and their systems were worthless, as expected against the random game.

In time, testing should indicate that every system is equally worthless when played against the "game".

Players should try to develop methods that exploit the "gaming device", since it's not possible to beat the "random game" in the long run.

« Last Edit: February 26, 2015, 11:41:54 PM by Real »
 

palestis

  • Veteran Member
  • ****
  • Posts: 547
  • Thanked: 273 times
Quote
they are wondering what went wrong.

The answer is that nothing went wrong.  They were simply facing "randomness" and their systems were worthless, as expected against the random game.

In time, testing should indicate that every system is equally worthless when played against the "game".

Players should try to develop methods that exploit the "gaming device", since it's not possible to beat the "random game" in the long run.
Players should try to develop methods to exploit "the gaming device?" That may sound appropriate for an ad hoc device in someone's basement. How can you use such simple language when you are talking about MEGA BIILION $ companies? Like the casinos are. You are not talking about picking a lock in an abandon house.
 Device defects and biases are being talked in just about every forum in the world. And the casinos never heard of it? Or if they did they are helpless in doing something about it?  Then banks should  leave their doors open and their money vaults unlocked. Are you saying that that's what the casinos let happen? And any players with high enough  IQ can develop tools and methods  to get into their money vaults? I seriously doubt it.
As a final thought device exploitation when it comes to roulettes  may only exist in someone's dreams.
At least with systems you win with your mind. And there is nothing a casino can do to block someone's mind.
 

Real

  • Fighting the war on absurdity one foolish idea at a time.
  • Hero Member
  • ******
  • Posts: 1201
  • Thanked: 120 times
  • Gender: Female
Quote
Players should try to develop methods to exploit "the gaming device?" That may sound appropriate for an ad hoc device in someone's basement. How can you use such simple language when you are talking about MEGA BIILION $ companies? Like the casinos are. You are not talking about picking a lock in an abandon house. Device defects and biases are being talked in just about every forum in the world. And the casinos never heard of it? Or if they did they are helpless in doing something about it?  Then banks should  leave their doors open and their money vaults unlocked. Are you saying that that's what the casinos let happen? And any players with high enough  IQ can develop tools and methods  to get into their. money vaults? I seriously doubt it. As a final thought device exploitation when it comes to roulettes  may only exist in someone's dreams. At least with systems you win with your mind. And there is nothing a casino can do to block someone's mind-Palestis


Palestis,
I'm not sure as to what it is that you're trying to say, but then again, I doubt you do either so I'll leave it at that.  ;)
Best of luck,
Real

« Last Edit: February 27, 2015, 07:32:15 AM by kav »
 

kav

  • www.Roulette30.com
  • Administrator
  • Hero Member
  • *****
  • Posts: 1531
  • Thanked: 580 times
  • Gender: Male
That may be true. But how about this?
 If a system depends on small number sequences, or random rows, and most of them are, then testing smaller samples  thousands of times in many different number collections, should yield to a definite conclusion. If a system depends on the last 20 numbers spun, then testing a different group of 20 numbers thousands of times, will tell you with certainty what to expect next time you are facing 20 numbers.
Unfortunately I have seen  many players looking at 150 number score cards, and coming up with a winning system in 15 minutes.   
Then when they lose next time they are at a table, they are wondering what went wrong.

Hi Palaistis,
It really depends on the system and what exactly one is trying to find out.
I do try and check ideas and concepts on a more or less limited number of spins in order to identify the characteristics of the system.
Your "tests" may be useful to you for other reasons.

However in the "long run" most system tests reveal the same result, the average expectation, which is -2,7%.
That's why I even doubt the value of millions or billions of spins tests, especially when they are done with RNGs. If one must test a system he at least should use the real spins available here.
In this article we explain the theories of evaluating roulette systems.
« Last Edit: February 27, 2015, 08:03:58 AM by kav »
 

palestis

  • Veteran Member
  • ****
  • Posts: 547
  • Thanked: 273 times
Quote
Players should try to develop methods to exploit "the gaming device?" That may sound appropriate for an ad hoc device in someone's basement. How can you use such simple language when you are talking about MEGA BIILION $ companies? Like the casinos are. You are not talking about picking a lock in an abandon house. Device defects and biases are being talked in just about every forum in the world. And the casinos never heard of it? Or if they did they are helpless in doing something about it?  Then banks should  leave their doors open and their money vaults unlocked. Are you saying that that's what the casinos let happen? And any players with high enough  IQ can develop tools and methods  to get into their. money vaults? I seriously doubt it. As a final thought device exploitation when it comes to roulettes  may only exist in someone's dreams. At least with systems you win with your mind. And there is nothing a casino can do to block someone's mind-Palestis


Palestis,
I'm not sure as to what it is that you're trying to say, but then again, I doubt you do either so I'll leave it at that.  ;)
Best of luck,
Real

What I'm trying to say is very simple. The whole business of VB and AP and bias begins to sound more like a science fiction scenario made for Hollywood. I doubt if there is such a thing in real life and its practical applications. The more you praise it, the more the idea is losing its strength. Especially statements like "exploiting the device".
Yea I would say let's leave it at that too, and if you think that no system can beat roulette, then I guess roulette can never be  beat by nobody. Unfortunately for you and the likes, I hate to disappoint you, but there are players who beat roulette with systems each and every day. If they abolished the HE tomorrow, and they paid 37:1, no player would notice the difference, in winning and losing results. Things would be exactly the same as they are today under the HE.
« Last Edit: February 27, 2015, 08:02:24 AM by palestis »
 

palestis

  • Veteran Member
  • ****
  • Posts: 547
  • Thanked: 273 times
That may be true. But how about this?
 If a system depends on small number sequences, or random rows, and most of them are, then testing smaller samples  thousands of times in many different number collections, should yield to a definite conclusion. If a system depends on the last 20 numbers spun, then testing a different group of 20 numbers thousands of times, will tell you with certainty what to expect next time you are facing 20 numbers.
Unfortunately I have seen  many players looking at 150 number score cards, and coming up with a winning system in 15 minutes.   
Then when they lose next time they are at a table, they are wondering what went wrong.

Hi Palaistis,
It really depends on the system and what exactly one is trying to find out.
I do try and check ideas and concepts on a more or less limited number of spins in order to identify the characteristics of the system.
Your "tests" may be useful to you for other reasons.

However in the "long run" most systems reveal the same result, the average expectation, which is -2,7%.
That's why I even doubt such millions of spins tests especially when they are done with RNGs. If one must test a system he at least should use the real spins available here.
In this article we explain the theories of evaluating roulette systems.
I agree that short run tests are very useful in determining if a conceived system is worth putting more time to it, or discarded as a waste of time. Short run tests act like a filter in determining a system's viability. Since "long run" has never been defined you don't need long run tests (measured in millions of spins). All you need is enough tests, and it won't be very long when things will become very clear. Then  one can easily decide if a system is worth playing or abandoned.
 

Sputnik

  • Veteran Member
  • ****
  • Posts: 418
  • Thanked: 338 times

 Hope this is ok by Kav, this is one coder that i use at www.norcosoft.com
 He makes Roulette Extreme codes for around $40 depending on complexity, he code roulette softwares and bots.
 You reach him by sending him one email at hes site.

 Any serios player should have at least one or several coders.

 Cheers
 

Real

  • Fighting the war on absurdity one foolish idea at a time.
  • Hero Member
  • ******
  • Posts: 1201
  • Thanked: 120 times
  • Gender: Female

Quote
What I'm trying to say is very simple. The whole business of VB and AP and bias begins to sound more like a science fiction scenario made for Hollywood. I doubt if there is such a thing in real life and its practical applications. The more you praise it, the more the idea is losing its strength. Especially statements like "exploiting the device". Yea I would say let's leave it at that too, and if you think that no system can beat roulette, then I guess roulette can never be  beat by nobody. Unfortunately for you and the likes, I hate to disappoint you, but there are players who beat roulette with systems each and every day. If they abolished the HE tomorrow, and they paid 37:1, no player would notice the difference, in winning and losing results. Things would be exactly the same as they are today under the HE. -palestis

If you'd like to start your own thread to discuss it further, then that's fine.  However, that discussion is off topic for this thread.

Quote
I agree that short run tests are very useful in determining if a conceived system is worth putting more time to it, or discarded as a waste of time. Short run tests act like a filter in determining a system's viability.Palestis

Show me even one system that's viable and worth further testing.  Your short term results are simply the result of a random fluctuation, and a poor understanding of basic probability by the observer.
 
 
 
Quote
Since "long run" has never been defined you don't need long run tests (measured in millions of spins). All you need is enough tests, and it won't be very long when things will become very clear. Then  one can easily decide if a system is worth playing or abandoned.-Palestis

 
Rather than "the long run" what you should be more interested in is the point at which a system is guarenteed to loose 99.99% of the time.  For example, after how many spins does your system fail to break even after a three or four standard deviation upward swing?  In the 00 roulette game, most of the outside systems crash completely after only 5k placed bets when flat betting. (Betting every spin).  The more numbers you bet, the sooner it will crash  By the way, if you're one of the gambler's fallacy guys that likes to use "triggers" and hop in and out, or wait on the sidelines before betting, you may stand around, wasting time, watching for over 100k spins, but once you've actually placed 5k flat bets, you too will likely be a loser as well.  Standing around does nothing to help your cause.   The player that stands around waiting for his "trigger" before betting will theoretically lose, after the same number of placed bets, as the player that bets every spin.
 
You should focus on how many standard deviations the system wins over a carefully defined test, and then see if you can replicate the results on future tests.  For example, on another forum there's a guy testing a "warm number system".  He regularly curve fits the next bet making by hopping back and forth between methods, looking back and forth in time, and then posts his win for the 40 or 50 spins played, while attempting to justify his results via BS.  It's absurd. When testing, you must define the test parameters, test a statistically relevant number of spins, and calculate the statistical significance of the test.  You can't look backwards and forward in time to find the best way to bet, and you can't derive any useful information looking at entry and exit points from small groupings of spin data.   When testing, you must remain true to yourself and not cheat the test.  And please...don't make up BS terms or use words out of context to try and make your system look like something that it is not.



 
Best of luck, and keep it real.
 
Real

« Last Edit: February 27, 2015, 06:01:39 PM by Real »
 

rouletteMann

  • New
  • **
  • Posts: 1
What if there is a system which operates on a small number of spins, and especially on a sample of 10000 spin, or more ...

 

Real

  • Fighting the war on absurdity one foolish idea at a time.
  • Hero Member
  • ******
  • Posts: 1201
  • Thanked: 120 times
  • Gender: Female
The number of spins on which you bet is what counts, not the number of spins that you wait for or watch.

This is why the trigger player will have the same theoretical results as the person that doesn't wait to bet.

« Last Edit: February 27, 2015, 09:22:20 PM by Real »
 

palestis

  • Veteran Member
  • ****
  • Posts: 547
  • Thanked: 273 times

Quote
 
Rather than "the long run" what you should be more interested in is the point at which a system is guarenteed to loose 99.99% of the time.  For example, after how many spins does your system fail to break even after a three or four standard deviation upward swing?  In the 00 roulette game, most of the outside systems crash completely after only 5k placed bets when flat betting. (Betting every spin).  The more numbers you bet, the sooner it will crash  By the way, if you're one of the gambler's fallacy guys that likes to use "triggers" and hop in and out, or wait on the sidelines before betting, you may stand around, wasting time, watching for over 100k spins, but once you've actually placed 5k flat bets, you too will likely be a loser as well.  Standing around does nothing to help your cause.   The player that stands around waiting for his "trigger" before betting will theoretically lose, after the same number of placed bets, as the player that bets every spin.
 
You should focus on how many standard deviations the system wins over a carefully defined test, and then see if you can replicate the results on future tests.  For example, on another forum there's a guy testing a "warm number system".  He regularly curve fits the next bet making by hopping back and forth between methods, looking back and forth in time, and then posts his win for the 40 or 50 spins played, while attempting to justify his results via BS.  It's absurd. When testing, you must define the test parameters, test a statistically relevant number of spins, and calculate the statistical significance of the test.  You can't look backwards and forward in time to find the best way to bet, and you can't derive any useful information looking at entry and exit points from small groupings of spin data.   When testing, you must remain true to yourself and not cheat the test.  And please...don't make up BS terms or use words out of context to try and make your system look like something that it is not. [/font]


 
Best of luck, and keep it real.
 
Real


When I sit down to play roulette I'm only concerned with what I see in front of me. If the last 20 numbers is my concern then I concentrate on those 20 numbers. Then when it comes to testing all I have to do is test segments of 20 numbers, 1000 times. More if I need to.
In those 1000 tests you will encounter every conceivable combination of standard deviations positive and negative. And many other factors will be encountered many times. Each time showing what the end result will be under a specific betting style . It won't take long until you start seeing patterns. There is no way that you will find 1000 unique situations in 1000 system specific tests. You will certainly find repetitions of the same conditions many times. Then red flags can be raised on combinations that are detrimental to the system and therefore avoided when encountered. 
If you are saying that every segment testing is unique in 10,000 tests then your assumption is wrong. Very wrong. On top of that you assume continuous play. Another wrong assumption. Another variable, however important, is that you fail to take into account the freedom to stop after  winning a certain amount and also stop after a certain amount of loss. When this variable is entered in the equation exact computation is impossible under your assumptions. Conveniently you assume non stop play with no ending in sight. You purposely are forgetting  the freedom of choice to stop. Therefore your scientific calculations are always wrong. Because you exclude factors that are vital during a playing session. There is more to roulette than standard deviation and the HE. A player who aims at $200 to win and he's willing to risk $1000, is very different than a player who aims at $200 but willing to risk only $300. Huge difference there. So how do you dare generalize your calculations so that one size fits all?
On top of all that, every system player here and everywhere already knows where he stands. You are talking as if members here are only contemplating to start playing roulette. Most have been playing for a long time. Therefore they all know by now where exactly they stand. So criticizing the system method is redundant. Your posts would have meaning, if nobody has already started plying roulette. But that's not the case. The football game has already started and has finished. What's the point? Help prospective new players from losing their hard earned money?

 

Real

  • Fighting the war on absurdity one foolish idea at a time.
  • Hero Member
  • ******
  • Posts: 1201
  • Thanked: 120 times
  • Gender: Female
Quote
On top of that you assume continuous play. Another wrong assumption. Another variable, however important, is that you fail to take into account the freedom to stop after  winning a certain amount and also stop after a certain amount of loss. When this variable is entered in the equation exact computation is impossible under your assumptions. Conveniently you assume non stop play with no ending in sight. You purposely are forgetting  the freedom of choice to stop. Palestis

 
It doesn't matter whether you play for 10k spins continuously, or play them off and on over the course of a year.  In the end, the results will be the same.  Your choosing when to stop isn't affecting the long term results.  What matters is the number of bets you place over the term, not the duration.
 
Just what magic is it that you believe takes place that enables you to win between when you quit, and when you return?  There is no mystical place to which you can run to after you quit that enables you to win the next time you return.
   
Quote
On top of all that, every system player here and everywhere already knows where he stands. You are talking as if members here are only contemplating to start playing roulette. Most have been playing for a long time.-Palestis

 
You haven't....I can tell by your misinterpretation of basic probability, and the lack of knowledge of the game history.  As a matter of fact, there's a series of steps that players typcially progress through.  It's called the "Evolution of The System Player".   Based on the system being used, and it's application you can basically "guesstimate" how long someone has been playing the game and or how often they actually play.

Evolution of The System Player.
1. The player discovers roulette
2. The player discovers the Martingale.  At this time the player is so astonished by it’s limited success, that he’s willing to completely ignore all of the experts and mathematicians.
3. The player discovers that the Martingale doesn’t quite work and begins looking for just the right progression to make it work.
4. The player can’t find just the right progression to make it work and begins looking for triggers, and revisits step 2.
5. The player may realize after some period of time that triggers are of no value.
6. The player turns toward money management in an effort to make the triggers perhaps work and to make the progression work better.  The believe in the “hit and run” is alive and well and he revisits steps 2 through 4.
7. The player begins to believe that greed is the problem, “if they just focus on winning only one unit per day that they can perhaps make it work.”  The believe in the “hit and run” is still alive, but not as much so.
8. The player begins to realize that “hit and run” and money management doesn’t work, because there is no time frame that enables it to work, and there is no magic place in which to run that enables it all to work again when he returns to play again.
9. The player faces gambling ruin.
10. The player begins to realize that perhaps the experts weren’t so wrong after all, and that “Denial ain’t a river in Egypt”.


If you want to win in the long term, you must design your methods around beating the wheel, not the random game. 

But look, if you don't trust me, then trust history.  The history of the game is documented.  You can read about what fails, and why,... as well as what has really worked.  You just have to take the time to actually read it!  A good article that you can read is called the Evolution of The Wheel and you can find it at rouletteresearch.com.   You should also read more on Wikipedia as well.  Also, it's important to look up the meaning of a very important word.  The word is "hubris".

Palestis,

Before you reply to this post, all that I ask is that you read an article on the history of the game, written by any encyclopedia.  That's all.  Just provide a link to what you've read. 
The more we come to know, the more we realize how little we know.

Read, research, and learn. 

-Real
« Last Edit: February 28, 2015, 12:43:27 AM by kav »