1. Welcome to the #1 Gambling Community with the best minds across the entire gambling spectrum. REGISTER NOW!
  2. Have a gambling question?

    Post it here and our gambling experts will answer it!
    Dismiss Notice
  3. Discussions in this section are assumed to be EV- as they are outside of the Advantage Play section. For EV+ discussions, please visit the Advantage Play section.
    Dismiss Notice

Roulette RNG vs. Real Wheels

Discussion in 'Roulette Forum' started by Shank, Nov 24, 2021.

  1. TwoUp

    TwoUp Well-Known Member

    Joined:
    Nov 30, 2020
    Likes:
    328
    Occupation:
    Unknown
    Location:
    Nowhere
    @gizmotron You're not reading what I wrote. This is not based on observations, this is based on how PRNGs actually work internally. There is no "confirmation bias".

    All RNGs use a sparse internal state and when that internal state becomes full of zeros there is not enough binary 1's left to mix to generate a quality output. I'm not talking about zeros being output, I'm talking about the outputs have coherence that takes a long time to recover and become decoherent.

    Read up on Linear Congruential Generators and stop being a moron.
     
  2. gizmotron

    gizmotron Well-Known Member Founding Member

    Joined:
    Jan 1, 2015
    Likes:
    3,040
    Occupation:
    Retired
    Location:
    The West Coast of USA, RV'ing
    The retarded scientist decided there were "too many zeros." There's that human misconception of what randomness looks like. By your own description you stated that there as an internal function to correct for that feature when none was ever needed. The zero is the 37th slot in Roulette. Some RNGs go from 0 to (number) and others go from 1 to (number), depends on platform and language being used. If your example of an RNG produces too may zeros then it's a pile of shit RNG to begin with and does not need correction do to human based observation bias. You must at some point come to the realization that there is a thing about computer algorithms, an axiom, "garbage in garbage out." Just because the slave to the human race computer makes mistakes does not mean the that person writing the software understands the real issues. You can look up why the new modern RNG algorithms change the random seed. It sounds just like shit taught in school. You read it in a book and believe it. The RNG should not be able to produce extra zeros in the first place. In this command: "put random(37) into holdResult" This is fourth generation X-Talk language. The zero is the 37th option and there is no way that last digit of any set comes out more times in millions of tests, or any derivative interval tested along the way. It sounds like the math used to predict global warming.
     
    mr j likes this.
  3. gizmotron

    gizmotron Well-Known Member Founding Member

    Joined:
    Jan 1, 2015
    Likes:
    3,040
    Occupation:
    Retired
    Location:
    The West Coast of USA, RV'ing
    I did. "The linear congruential method has the advantage of being very fast, requiring only a few operations per call. It has the disadvantage that it is not free of sequential correlation on successive calls."

    All this went away a long time ago when people started exploiting slot machines. Every time the computer starts it selects a new operating random seed. And there are millions of starting points with at least a half million results before reoccurrence takes place for every random seed available.
     
    mr j likes this.
  4. gizmotron

    gizmotron Well-Known Member Founding Member

    Joined:
    Jan 1, 2015
    Likes:
    3,040
    Occupation:
    Retired
    Location:
    The West Coast of USA, RV'ing
    Sorry about the science lesson:

    ...and they went beyond that by starting with a unique random seed to improve the results discovered by weakness found within m
     
    Last edited: Nov 25, 2021
    mr j likes this.
  5. Nathan Detroit

    Nathan Detroit Well-Known Member Founding Member

    Joined:
    Dec 25, 2014
    Likes:
    2,080
    Like the man said " Never wise up a chump "


    This cat was not wrong .
     
    mr j likes this.
  6. gizmotron

    gizmotron Well-Known Member Founding Member

    Joined:
    Jan 1, 2015
    Likes:
    3,040
    Occupation:
    Retired
    Location:
    The West Coast of USA, RV'ing
    Whereas "too many zeros" probably refers to machine language where it's binary 1's and 0's. That could be the case for your argument. I don't really care. Fare RNG is when you need 300 results in a row and it's basically straight forward with out manipulation or subversive targeting.
     
    mr j and Nathan Detroit like this.
  7. SPIKE

    SPIKE Well-Known Member

    Joined:
    Jan 1, 2020
    Likes:
    865
    Location:
    midwest
    'Most RNGs also only have fairness on first degree (single outcomes) they fail statistical tests on pairs, they fail binary-rank test and linear complexity tests."

    All I know is the outcomes are screwy compared to a real wheel and I can't beat them for very long. They do things I never see on real wheels. Either they are too random are not random enough, either way I avoid them.
     

  8. Rheti

    Rheti Active Member

    Joined:
    Apr 5, 2018
    Likes:
    70
    Location:
    rotterdam
  9. Nathan Detroit

    Nathan Detroit Well-Known Member Founding Member

    Joined:
    Dec 25, 2014
    Likes:
    2,080
    Any one going for air ball roulette ? LOL
     
  10. TwoUp

    TwoUp Well-Known Member

    Joined:
    Nov 30, 2020
    Likes:
    328
    Occupation:
    Unknown
    Location:
    Nowhere
    I don't trust any PRNG for my testing and I don't play in online casinos, but my reasons for that go far beyond RNG defects. I much prefer real entropy to eliminate issues in degrees of freedom and statistical and quality flaws in RNGs.

    My background is in high assurance cryptographic equipment where strength of randomness is essential so I have a different perspective on what is good.

    If you do use a RNG for testing then the Fortuna RNG is very good as it manages entropy pools and uses cryptographic strength number generation. It's not a PRNG as it has entropy as its input and also manages bias in the entropy as well.
     
    Last edited: Nov 25, 2021
  11. gizmotron

    gizmotron Well-Known Member Founding Member

    Joined:
    Jan 1, 2015
    Likes:
    3,040
    Occupation:
    Retired
    Location:
    The West Coast of USA, RV'ing
    I know what is happening because I know things. You know what happens after an ice age? It melts.
     
  12. SPIKE

    SPIKE Well-Known Member

    Joined:
    Jan 1, 2020
    Likes:
    865
    Location:
    midwest
    That is a good way of putting it, 'strength of randomness.' I have been saying for years and years that RNG's do not produce the same quality of random that a roulette wheel does. I get laughed at, mocked, told I'm insane. It's nice to have somebody with a science background verify it for me. To me it's obvious this the outcomes are different. I mostly cannot play my game on an RNG and I mostly can play my game on a real roulette wheel. We know for a fact that the wheel is producing real random results. Logical deduction means the RNG is not. The random it produces is good enough for what it's used for, but it's still not true random.
     
    TwoUp likes this.
  13. SPIKE

    SPIKE Well-Known Member

    Joined:
    Jan 1, 2020
    Likes:
    865
    Location:
    midwest
    Yet the largest glacier in Greenland is growing at a record pace. The glaciers along the China/India/Pakistan border are also growing. For every Glacier that's melting there is almost a glacier somewhere that is growing. So much for global warming. And the South Pole just had its coldest winter in recorded history producing God knows how many metric tons of new ice. They told us twenty years ago that by now the South Pole would be totally melted and New York City would be underwater. Oops, the exact opposite happened
     
    TurboGenius and TwoUp like this.
  14. SPIKE

    SPIKE Well-Known Member

    Joined:
    Jan 1, 2020
    Likes:
    865
    Location:
    midwest
    True random it's very hard to produce. We can't do it. There used to be a game on the internet where you try to produce random outcomes. You choose between A & B in any sequence you want and it tells you how close to random you are. Impossible to do and it's not even completely accurate because again it's an RNG telling you if you're accurate.
     

  15. Shank

    Shank Active Member

    Joined:
    Oct 29, 2021
    Likes:
    84
    Location:
    Earth
    Where can I find a huge database of real spins?
     
  16. Rheti

    Rheti Active Member

    Joined:
    Apr 5, 2018
    Likes:
    70
    Location:
    rotterdam
    Wiesbaden
     
    Nathan Detroit likes this.
  17. Dr. Sir Anyone Anyone

    Dr. Sir Anyone Anyone Well-Known Member Lineage to Founders

    Joined:
    Feb 18, 2015
    Likes:
    931
    Occupation:
    Shoe Cobbler
    Location:
    Merica
    The average system player isn't able to differentiate between an RNG wheel and a live wheel. Especially someone that bets on the outside ECs.

    There are some differences that an expert can spot.

    1. Live wheels are all biased to some degree. (Weak wheels with alternating spin direction aren't going to show up unless spin directions are separated.)
    2. As the spin samples increase, the live wheel chi will tend to continue growing, but the chi on the RNG wheel will remain somewhat flat.
    3. How the hits are distributed around the wheel also tends to look different as the spin samples grow larger. (After 20k spins they are often obvious.)
     
    Last edited: Nov 26, 2021
    fathead and TwoUp like this.
  18. SPIKE

    SPIKE Well-Known Member

    Joined:
    Jan 1, 2020
    Likes:
    865
    Location:
    midwest
    I can spot the difference in the EC's, easy peasy
     
  19. Dr. Sir Anyone Anyone

    Dr. Sir Anyone Anyone Well-Known Member Lineage to Founders

    Joined:
    Feb 18, 2015
    Likes:
    931
    Occupation:
    Shoe Cobbler
    Location:
    Merica
    Let's see some proof. How many spins do you need from each?
     
  20. SPIKE

    SPIKE Well-Known Member

    Joined:
    Jan 1, 2020
    Likes:
    865
    Location:
    midwest
    I don't do tests. Why would I give a rip if somebody believes me or not. When have I ever cared. I know I can do it, that's all that matters.
     

Share This Page