Are Some Computer RNGs More Equal Than Others?

Quote: Sonny44

Craps math discussions are fun for, well, math people. They love their calculations out to the 9th decimal, and so be it. They love all those numbers. But, they work out their calcs on computers with flawed RNGs. The dice are the only reliable RNG, and that can be found only on a live table.

Sonny, I like you for your cordiality.
But, damn (!), you are one conclusion-jumper par excellence.
"work out their calcs on computers with flawed RNGs" Come on, gimme a break.

Quote: Wizard

May I recommend you visit John Patrick's forum.

Isn't that at least the 2nd time you've said that in a week? :D And someone on the other forum I constantly post in started a thread asking if John was even still alive today...lol

Quote: indignant99

Sonny, I like you for your cordiality.
But, damn (!), you are one conclusion-jumper par excellence.
"work out their calcs on computers with flawed RNGs" Come on, gimme a break.

Yeah, math people like to avoid flaws usually.

Quote: Sonny44

Do you disagree re: computer sims vs. experience at a live table? Are you suggesting machine/computer games are as valid as live tables? Do you affirm that computer RNGs are reliable as predictors of live table experience where the dice are the RNG? I will examine "John Patrick's forum," which I know nothing about, except that Mr. Patrick has some craps ideas of his own, discounted by others.

Sorry if I'm a thorn in your side. You have performed a valuable service to the craps community, and I don't discount what you have done. Your sites are required reading for any craps novice. Despite your wish I leave this site (as another poster has suggested), I will not, as there are many good discussions on it. I filter out the involved math discussions. I am free to express my opinion as much as anyone else on this board, whether you like it or not.

You are, but the board owners have the ability to say anything they like in response, suggest or ask you to leave, or anything else they desire.

I'd challenge you to distinguish between a series generate by dice, and a series generated by a good RNG (there are several ways to screw up a RNG). I suspect you can't over the lengths of sequences we are talking about.

Quote: Sonny44

Are you suggesting machine/computer games are as valid as live tables? Do you affirm that computer RNGs are reliable as predictors of live table experience where the dice are the RNG?

Yes and yes, for any decent RNG. I'm not saying every RNG is perfect, but it is not difficult to get one that will pass any test of statistical randomness you throw at it, including the one I use.

Now, please visit Patrick's forum.

Quote: Sonny44

Do you disagree re: computer sims vs. experience at a live table? Are you suggesting machine/computer games are as valid as live tables? Do you affirm that computer RNGs are reliable as predictors of live table experience where the dice are the RNG? I will examine "John Patrick's forum," which I know nothing about, except that Mr. Patrick has some craps ideas of his own, discounted by others.

Sorry if I'm a thorn in your side. You have performed a valuable service to the craps community, and I don't discount what you have done. Your sites are required reading for any craps novice. Despite your wish I leave this site (as another poster has suggested), I will not, as there are many good discussions on it. I filter out the involved math discussions. I am free to express my opinion as much as anyone else on this board, whether you like it or not.

I would appreciate an answer to the above questions.

Mea culpa. After reading responses to this post, I apologize for sounding so adamant. I have learned re: RNGs from this & other threads and thank those who posted. I really have no right to criticize others who are able to do what I have difficulty with. Obviously, math discussions re: craps have meaning for a lot of people. That they have little meaning for me beyond the basics is my problem.

I will continue reading and posting on this board and try not to jump to conclusions as 99 pointed out. As to other boards, I've checked Patrick's & it has some good discussions, but I spend a lot of time on the Craps Forum board. There are good bits & pieces on these boards; it's enjoyable reading experiences and thoughts of other players, which helps me decide what I want to do when I walk up to my next table.

Quote: Wizard

Yes and yes, for any decent RNG. I'm not saying every RNG is perfect, but it is not difficult to get one that will pass any test of statistical randomness you throw at it, including the one I use.

Now, please visit Patrick's forum.

ANY test? aprapparently you haven't been paying attention to any of Bob's RNG post's.

Quote: AxelWolf

Quote: Wizard

Yes and yes, for any decent RNG. I'm not saying every RNG is perfect, but it is not difficult to get one that will pass any test of statistical randomness you throw at it, including the one I use.

Now, please visit Patrick's forum.

ANY test? aprapparently you haven't been paying attention to any of Bob's RNG post's.

What IS hard is to test for randomness on a short sequence. That's not to say it's not possible. There's various papers on this very topic that are worth reading, but most assume a binary source of data, not a series of numbers (1-6,1-6, for instance, or 1-38). What I've never worked out is a correct translation of a datastream into a binary to use these techniques (vice versa is easy).

But that's just me.

Quote: Sonny44

Craps math discussions are fun for, well, math people. They love their calculations out to the 9th decimal, and so be it. They love all those numbers. But, they work out their calcs on computers with flawed RNGs. The dice are the only reliable RNG, and that can be found only on a live table.

This comment is still aggravating the livin' %$&! out of me.
Is pi (the constant) better as 3.14 or as 3.141592653 ?
Is e (Euler's number) better as 2.718 or as 2.718281828 ?
We could give you the formulas, but whoosh they'd fly way over your head.

I'm going to give you a link, to an "odds" or probability calculation I did. Here. I did not use any RNG. Hell, I didn't even use a computer. Just good old pen and paper. Throughout, I only used integers (counting numbers) and fractions. Only at the very end, when I had an exact answer (an exact fraction), did I convert it to a decimal equivalent. Because most people have trouble wrapping their heads around weird fractions like 384,156 / 60,466,176.

Now I'm wanting my user name to be "9th-decimal." Oh crap, maybe my calculator is infected with a rogue RNG.

Quote: indignant99

This comment is still aggravating the livin' %$&! out of me.
Is pi (the constant) better as 3.14 or as 3.141592653 ?
Is e (Euler's number) better as 2.718 or as 2.718281828 ?
We could give you the formulas, but whoosh they'd fly way over your head.

I'm going to give you a link, to an "odds" or probability calculation I did. Here. I did not use any RNG. Hell, I didn't even use a computer. Just good old pen and paper. Throughout, I only used integers (counting numbers) and fractions. Only at the very end, when I had an exact answer (an exact fraction), did I convert it to a decimal equivalent. Because most people have trouble wrapping their heads around weird fractions like 384,156 / 60,466,176.

Now I'm wanting my user name to be "9th-decimal." Oh crap, maybe my calculator is infected with a rogue RNG.

Didn't mean to get you so indignant, indignant99. ; ) And, I can see where 9 decimals is better than the fraction you cited. Where I've failed in my remarks is simply to say, craps math calculations carried to 9 decimals has no value at a live table. I mean, what can the difference between, say, 5.123456788 & 5.123456789, make at a live table? A live table deals only in integers (or, whole numbers, whatever).

I've apologized for discounting what, for many of you, you enjoy: Probability Theory applied to craps. (I'm conflating this thread w/ another one I started.) And, there may be some apologies due for those who develop craps games w/ RNGs. I can only say what I've said before: Success or failure of a particular strategy, method, etc. on a computer sim cannot be trusted to hold true on a live table. I know: I've experienced that.

I want to leave with, you math folks continue to enjoy what you enjoy & I'll sit down & shut up.

Quote: Sonny44

Didn't mean to get you so indignant, indignant99. ; )

In pursuit of clearing the air, and achieving a friendlier atmosphere, I make two concessions: (1) I don't like gambling simulators, either. Including the Wiz's. My live-play, on live tables, gets better results than the sim-play. That includes WinCraps, too. (2) Compared to the membership as a whole, here, your Spelling and Grammar score an "A." I can read you easily; not like some broken-word salad postings that I just have to click away in disgust.

The casinos seem to like math and simulations. They also seem to like a few decimal places.

Quote: Sonny44

And, I can see where 9 decimals is better than the fraction you cited. Where I've failed in my remarks is simply to say, craps math calculations carried to 9 decimals has no value at a live table. I mean, what can the difference between, say, 5.123456788 & 5.123456789, make at a live table? A live table deals only in integers (or, whole numbers, whatever).

Okay, suppose everyone who comes to Vegas decides to play their favorite game in the biggest bet allowed. If possible they will play their entire gambling allowance in a single bet. Several place thousands on a single spin on roulette, 1k or 5k bets on slots, whales bet 5 million on a single BJ hand. Whatever, just go with the idea as if it happened.

The results are some won, and some loss. They barely moved the dial on a scale of a gambling session as far as time involved. Some only lasted seconds and they won or were broke.

Question for you.

Do you believe odds played no factor in their wins or losses? That odds of the game had no effect whatsoever on the various outcomes.

My theory is RNG's are more close to true
random than anything mechanical in a
casino. Dice, wheel, cards, whatever.
That's why the results look different
on an RNG and you get different outcomes
in your systems.

Quote: EvenBob

My theory is RNG's are more close to true
random than anything mechanical in a
casino. Dice, wheel, cards, whatever.
That's why the results look different
on an RNG and you get different outcomes
in your systems.

I believe that Humans including dealer's shuffling card's are less random than any RNG.

The house edge eats up any non randomness. Either way it wouldn't matter because they might be not as random as a RNG it they might be shuffling in your favor or the opposite. In the end it all washes out.

Quote: AxelWolf

I believe that Humans including dealer's shuffling card's are less random than any RNG.
.

You realize that's what I just said. RNG's
are more random than mechanical random
from human sources. All casino games will
be RNG random one day and unbeatable.

Quote: EvenBob

You realize that's what I just said. RNG's
are more random than mechanical random
from human sources. All casino games will
be RNG random one day and unbeatable.

Not true they have lots of RNG games that are beatable. From Slots, Roulette, VP, and BJ.

Example of a RNG of a BJ game that was very lucrative.

https://lh6.googleusercontent.com/-aYDp9itfDy8/TycytVupWhI/AAAAAAAAAHo/83kO0QY8UGc/s640/IMG_20110930_143249.jpg.

Some had good rules fair penetration, cash and comps. Imagine spreading $2 to $500.