Bitcoin jackpot 728x90

Author Topic: Question for AP Players  (Read 522 times)

0 Members and 1 Guest are viewing this topic.

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1188
  • Thanked: 674 times
Re: Question for AP Players
« Reply #15 on: October 16, 2017, 11:18:32 PM »
Reyth, something do not add up. In 16M spins ... these "spins over expectation" would mean absolutely nothing. In fact even these "hits over expectation " would mean nothing in 16M spins.   Looks like these values are accounted only for a fraction of the sample ( 37 spins as you do or so) and this way can not be statistically significant or show anything meaningfull. Probably you should include some other coefficients that reflect stability of performance to get more clear picture.
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3824
  • Thanked: 1206 times
Re: Question for AP Players
« Reply #16 on: October 16, 2017, 11:27:51 PM »
I have done plenty of that too.  Of course I am not saying that those numbers persisted as they are for the enitre sample (they may have, in some fashion) but that the phenomenon persisted even after 16M spins; I get it every time I run it, even though this is one of the "lowest" samples I have seen so far.  To me, that is quite valuable.

Of course, digging deeper to see the path of the numbers is high on my priority list and I am now very close to being in a position to start doing just that.

The sample was deliberately parsed to show only numbers with 100+ hits above expectation.  All other numbers were either negative or less than 100.

EDIT: I apologize of course that is HITS above expectation, which represents just over 6000 SPINS (less than expectation) in the top peformer.
« Last Edit: October 16, 2017, 11:33:59 PM by Reyth »
 

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1188
  • Thanked: 674 times
Re: Question for AP Players
« Reply #17 on: October 17, 2017, 12:55:51 PM »
Reyth, lm confused beyond my capabilities to understand. 
   There are levels of abstraction that are conventional: average , std, chi- square,  level of confidence  ( chance of randomness)... 
   There are not " conventional " ones like coefficient of stability, skips/ repetitive hits likehood...ets.
   We need these to determine how likely phenomenon that we observe is to repeat itself in play. It all boils down to the correlation and stability of correlation, if you wish " index of correlation coefficient "... all of these are valid math tools that show how likely and stable is phenomenon that we observe.
       In the nutshell,  we have to have a " trigger" that says " bet there when..." and it should be statistically significant.  If not, what we observe are just flick of fluctuations that we cannot predict,  be it rng or  real wheel.
      Naturally we are interested only on phenomenon that is consistent and predictable. ....  if idea is to make money of cause....
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3824
  • Thanked: 1206 times
Re: Question for AP Players
« Reply #18 on: October 17, 2017, 02:57:48 PM »
To me when I see 164 hits above expectation, I see money.
 

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1188
  • Thanked: 674 times
Re: Question for AP Players
« Reply #19 on: October 17, 2017, 03:05:47 PM »
Is it in 6k spins or in 16M? In 6k it's abusive money... no one have seen such in sentures,  if it's in 16M , then it's a loss.
    I gess there is misunderstanding in your explanation. 
   What would help is to specify sample size for wich counting is done and how many times each number hit. That will display  everything clearly.
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3824
  • Thanked: 1206 times
Re: Question for AP Players
« Reply #20 on: October 17, 2017, 04:04:30 PM »
So here is some output for the top numbers:

.#.................Hits (> expectation) <=== Hits greater than expectation
 7.................163
 10...............135
 19...............164 <=== This number hit 164 more times than expected (over 6000 less spins than expected)
 20...............140
 22...............121
 25...............137

These are the TOP PERFORMING numbers after 16M spins, all other numbers hit below expectation or less than 100 hits more than expected. 

This is a particularly LOW sample output; usually we see 8-9 such numbers and at least one selection that has hit over 200 times greater than expected.

So ya, this is money.

I mean obviously your point about the exact flow of these numbers within the output is well taken and understood but my goal was to show that the phenomenon would still be present regardless of the number of spins; that it is a consistent and long-term phenomenon.

I would actually expect that several of these numbers have switched with others that are not shown -- that is important, but not as important as the overall phenomenon persisting.

This phenomenon is just randomness, nothing to worry about. ;)
« Last Edit: October 17, 2017, 04:23:52 PM by Reyth »
 

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1188
  • Thanked: 674 times
Re: Question for AP Players
« Reply #21 on: October 17, 2017, 09:28:24 PM »
Reyth, sry to be annoying,  l just wanna understand how many times exactly each number hit and over how many spins. If not, it becomes equation with 2 unknown variables ( how many hits and how many spins) . What shold l do to understand it, like at school equal it to zero and resolve equation graphic way?
 Even so l will get too many multiple answers ( every point included in the direct line) .
      I'm not worried,  just confused, probably even more then you are.
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3824
  • Thanked: 1206 times
Re: Question for AP Players
« Reply #22 on: October 17, 2017, 10:01:42 PM »
I am so used to tracking hits vs. spins but in this case its TOTAL ACTUAL NUMBER OF HITS ABOVE EXPECTATION FOR 16M SPINS.

I don't care how useless it is, I know its useless.  What itsn't useless is the fact that the phenomenon is still present.

Later, sometime soon, I can start digging in and studying these massive sequences and make certain determinations about the behavior of the phenomenon.
« Last Edit: October 17, 2017, 10:03:28 PM by Reyth »
 

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1188
  • Thanked: 674 times
Re: Question for AP Players
« Reply #23 on: October 17, 2017, 10:54:58 PM »
 Why you say it's useless?  It actually permitted me to advance a bit on creation of " intelligent bot" . I was freaking my brains few months already how to make an excel take logical decisions while l sleep without halting RAM ;)
    Your coefficient  ( few like that actually) are very friendly in terms of simplicity of required math. logic. They permit to control skips and hits with much less hardware power nessesary.
    What can l tell to summarise. ... computers are really stupid sometimes. 
     Reyth, do not give up, you got interesting idea and lm really curious in this phenomenon you are observing.
 
The following users thanked this post: Reyth

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3824
  • Thanked: 1206 times
Re: Question for AP Players
« Reply #24 on: October 17, 2017, 11:31:56 PM »
Thanks!  I stumbled across it accidentally (thank God).  Real knows about it but as you both know, physical bias is far more reliable and that's how you guys make the big bucks. :)

Ya, its a very simple formula that does all the tracking in an instant with small RAM usage!  Its practially UNLIMITED! :D

There are other things I track, like cumulative gapping and coups:total gaps that help in my analysis too. :)
« Last Edit: October 17, 2017, 11:33:52 PM by Reyth »
 

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1188
  • Thanked: 674 times
Re: Question for AP Players
« Reply #25 on: October 18, 2017, 02:02:08 AM »
 I belive l make much less money then Real... he has a man power and l'm one soldier army.
    I know that probably it is a stupid dream to simulate artificial intelligence with Excel and visual tracker programme is as far from being perfect as it gets .... but theoretically it should be possible to create a bot that looks physical parameters as well and not only numbers. I call it " my stupid hobby ".
 
The following users thanked this post: Reyth

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3824
  • Thanked: 1206 times
Re: Question for AP Players
« Reply #26 on: October 18, 2017, 02:52:52 AM »
I would never use a bot to play any game, whether it be for money or not.  Programming the comp to bet is an important skill, if it is possible.  For you it is much more difficult than for me because I don't have any physical parameters to worry about. :)
 

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1188
  • Thanked: 674 times
Re: Question for AP Players
« Reply #27 on: October 18, 2017, 04:14:05 AM »
 Algorithm of my play is very easy. It can be described as " if(( A=x & B=y & C=z...),bet, collect data). Where A,B,C are variables and x,y,z are desired values.
    Most difficult part of the thing is to make bot collect the data. Require comp vision usage ( open CV libraries....
    Betting is an easiest part, can be done with task automaton and key/mouse moviment logger( already done). Best part of it all that logic engine is done also and it's completely in excel.
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3824
  • Thanked: 1206 times
Re: Question for AP Players
« Reply #28 on: October 18, 2017, 05:12:23 AM »
Wow very nice work!  Me and bots are as far as the east is from the west, even though I can make them. :)

Here is my formula for gap speed:

Speed = (total number of gaps * # coups expected for gap) - (total number of coups taken) * -1

Positive numbers show a slower gapping speed and negative numbers show a faster gapping speed.  You define what "gap" means in number of spins without a hit and calculate what expectation in number of COUPS is for a gap to occur, and plug those values into the formula. :) 
« Last Edit: October 18, 2017, 05:14:22 AM by Reyth »
 

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1188
  • Thanked: 674 times
Re: Question for AP Players
« Reply #29 on: October 18, 2017, 11:11:45 AM »
 Gaps are interesting as conditions qualifiers.  If you plot these side by side in excel,  for when your bet is performing well with when it not, you will see huge differences in gaps as well... 
    It's a way to create proper systems and betting plans specifically for a wheel you gonna play. Example would be my "sequential method" post. Choice of the trigger there was based on the fact of unusual concentration of gaps with zero value ( consequtive hitts).
   When it was bad, huge values of gaps, almost no consequtive hitts.... so no bets. When it's good, huge concentration of hitts, almost no gaps( small values).
    There are logical limits for a gap values when it good, but how to calculate it properly... lm still working on it, for now it's still based on empirical observation and is  manualy specified variable. 
   
 
The following users thanked this post: Reyth