Author Topic: why would someone test a method using simulation software?  (Read 726 times)

0 Members and 2 Guests are viewing this topic.

mr j

  • Hero Member
  • ******
  • Posts: 1049
  • Thanked: 330 times
  • Gender: Male
Re: why would someone test a method using simulation software?
« Reply #30 on: June 10, 2018, 10:27:29 PM »
LOL  :D
 

MickyP

  • Hero Member
  • ******
  • Posts: 1328
  • Thanked: 567 times
  • Gender: Male
Re: why would someone test a method using simulation software?
« Reply #31 on: June 11, 2018, 04:07:36 AM »
I get it. Do you?  ::)
 

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1924
  • Thanked: 1026 times
Re: why would someone test a method using simulation software?
« Reply #32 on: June 11, 2018, 06:56:14 AM »
Coming back to the topic. ..
   So why one would bother?
 Generally speaking, one number per spin determines final result.  That number is payed if player betted on it. All other numbers loose. There are 37-38 numbers to choose from..  lots of degrees of freedom for a ball final position.
    Game of roulette is probabilistic in nature, even genuine physics based prediction have both: chance to be correct and chance to be wrong on every individual spin.
    Data itself is used to determine optimum playing conditions, where numbers/ positions  ( distances) of interest are on the graph( wheel), optimal bet coverage, optimum money management strategy and " how forgiving is playing situation for possible players mistakes".
    Target itself ( bet selection) may suffer permutations during the game. ...move,  become more/ less wide..ets. To account for changes in the game player has to model them using past spins data collected previously.  Obviously more data collected- better for a player.
   There is no need to use millions of spins to determine anything. It's simply not proactive.  However longer data setts do help player to create playing model that is more flexible and more reliable.
   
 

MickyP

  • Hero Member
  • ******
  • Posts: 1328
  • Thanked: 567 times
  • Gender: Male
Re: why would someone test a method using simulation software?
« Reply #33 on: June 11, 2018, 08:15:09 AM »
Well stated, MrPerfect.
 

mr j

  • Hero Member
  • ******
  • Posts: 1049
  • Thanked: 330 times
  • Gender: Male
Re: why would someone test a method using simulation software?
« Reply #34 on: June 11, 2018, 12:47:29 PM »
"There is no need to use millions of spins to determine anything" >> I agree.

Ken
 

MickyP

  • Hero Member
  • ******
  • Posts: 1328
  • Thanked: 567 times
  • Gender: Male
Re: why would someone test a method using simulation software?
« Reply #35 on: June 11, 2018, 01:11:18 PM »
Roulette is one long game...lol. Millions of spins equals one game...lol. What percentage of the spins of this single long game is adequate to test a method?  Haha, just messing with the flat earth people.
 

mr j

  • Hero Member
  • ******
  • Posts: 1049
  • Thanked: 330 times
  • Gender: Male
Re: why would someone test a method using simulation software?
« Reply #36 on: June 11, 2018, 01:27:14 PM »
Yes correct........ 7 sessions, 9 sessions, 12 sessions, 2 days, 6 days, skipping 27 days, 5 different casinos, 3 different cities. Its *ALL* together as ONE. Sorry, them the facts.
 

MickyP

  • Hero Member
  • ******
  • Posts: 1328
  • Thanked: 567 times
  • Gender: Male
Re: why would someone test a method using simulation software?
« Reply #37 on: June 11, 2018, 02:07:46 PM »
Yup, the Easter bunny is also a fact
 

Mike

  • Great Contributor
  • *****
  • Posts: 912
  • Thanked: 188 times
Re: why would someone test a method using simulation software?
« Reply #38 on: June 11, 2018, 02:57:05 PM »
"There is no need to use millions of spins to determine anything" >> I agree.
I also agree, but if you don't know WHY a method works (or seems to work), you need strong statistical evidence that it does, and too many are satisfied with weak or no evidence. This is because they are basically amateurs and have no knowledge of statistical inference or understanding of what constitutes strong/weak evidence.
Question : you have tested a system which bets on 3 numbers and after 3k bets you have made 6k flat betting (@ $25 units). Is this good evidence that the system "works"? Whatever your answer, can you explain WHY it is or isn't?
 

MickyP

  • Hero Member
  • ******
  • Posts: 1328
  • Thanked: 567 times
  • Gender: Male
Re: why would someone test a method using simulation software?
« Reply #39 on: June 11, 2018, 03:25:00 PM »
Mike, the idea of testing methods is a controversial one to say the least. Players get put into boxes according to many things including how long or short their tests are. I think each player should test a method to a point that they are comfortable with. I don't see the need to test for millions of spins either.

I think people grow confidence in a system that starts off showing a profit and continues showing a profit throughout the test, as long as the winnings do not go below the original starting bankroll. The idea of long tests is to try and test through "the run from hell" and to see where the strengths and weaknesses of the method lie.
 

Sputnik

  • Veteran Member
  • ****
  • Posts: 703
  • Thanked: 579 times
Re: why would someone test a method using simulation software?
« Reply #40 on: June 11, 2018, 03:28:17 PM »


I read variance like a open book and f*** lady luck from behind.

Cheers
 

Mike

  • Great Contributor
  • *****
  • Posts: 912
  • Thanked: 188 times
Re: why would someone test a method using simulation software?
« Reply #41 on: June 11, 2018, 03:31:31 PM »
@ Sputnik,
In that case my question should be easy for you to answer.  ;)
Micky,

Care to have a shot at it?

 
 

MickyP

  • Hero Member
  • ******
  • Posts: 1328
  • Thanked: 567 times
  • Gender: Male
Re: why would someone test a method using simulation software?
« Reply #42 on: June 11, 2018, 03:37:42 PM »
Regarding your question; how do you determine what three numbers to play? Can you give some guidelines to this method? It will be difficult to have a meaningful discussion without additional information. Maybe Ken can jump in as he only plays 2-4 numbers but I don't think he will.
 

Sputnik

  • Veteran Member
  • ****
  • Posts: 703
  • Thanked: 579 times
Re: why would someone test a method using simulation software?
« Reply #43 on: June 11, 2018, 04:07:55 PM »


Well, Laurance Scott teach me the difference between what is random and true bias with simulation software and spreadsheets and material to explain the Monte Carlo Simulations.
There is a complete playing model to verify and tell the difference without defect spotting.

Laurance Scott got inspired after some discussion with Thorp.

No interest or time to discuss this with you and Caleb can verify is all true as I send him the whole package.
Any who has it got it from me, no one else has it.

Cheers
« Last Edit: June 11, 2018, 04:10:54 PM by Sputnik »
 

Mike

  • Great Contributor
  • *****
  • Posts: 912
  • Thanked: 188 times
Re: why would someone test a method using simulation software?
« Reply #44 on: June 11, 2018, 05:42:03 PM »
Regarding your question; how do you determine what three numbers to play? Can you give some guidelines to this method?
The details of the system don't really matter, and I didn't have a particular one in mind anyway. The point is that it's possible to evaluate the results of ANY system purely from a statistical point of view without knowing WHY it works. But as I mentioned, it's always better if you can give logical reasons (such as physics, bias etc, ie cause & effect), and then the statistical analysis only serves as confirmation of your "theory", or not.
« Last Edit: June 11, 2018, 05:47:59 PM by Mike »