I have an issue that I really want to get it out for a debate. I really see so many "roulette players" that are testing different systems for, I don't know, 10k spins, 100k spins, millions of spins. Ok, it's good that all the systems that come up are being tested, but - and here comes my issue - where do you test these? On RX? On random.org? What's that have to do with a real roulette?
IMO RX or random.org are nothing else but some number generators when it comes to testing systems. I mean.. the only difference between these softwares and you writting 10000 numbers on paper is that the softwares are faster.
Don't missunderstand me. RX is awesome for coding, tracking, showing you what to bet and when. Nobody can argue with that. But for testing? Let's get real here...
I really think that a true test of a system has to be done with live dealer or MAYBE, just MAYBE on an airball. Nothing more! You'll say that it will take forever to test a system like that. You're right! But I'ld rather test a system for a week or more to come up with a true valid oppinion rather than test it on some I don't know how many thousands of numbers a software can generate in 10-20mins.
I know that I started a subject that might explode, but... this is just my oppinion. If I'm wrong... feel free to argue!
According to the last FLAT_INO, a
"successful lifetime bet" will be in profit after 400 bets.
(It works for me)
Quote from: Steeefan2014 on Feb 11, 05:07 PM 2020here comes my issue - where do you test these? On RX? On random.org? What's that have to do with a real roulette?
You can load real spins into RX or any simulator, but it makes no difference unless you're using a physics based method of play. Providing the RNG is fair, you won't be able to tell the difference between real spins or an RNG. The roulette wheel is after all just a random number generator.
Listen to Joe, he's right.
For the purposes of testing systems or mechanical methods, all types of fair random number generators are equal.
Let's get real here, if you need to argue this point then you are a novice.
Don't feel too bad though, as this argument has been going around in circles for years.
As for the amount of trials adequate for a comprehensive system test, 50k is relatively dependable.
I really don't need to argue! There are just my thoughts. As I said, maybe I am wrong... but this is how I see it!
If there is a logical explanation, I'm open to that, ofc.
ok, sorry my bad.
If you need to debate this point then you are a novice.
Don't feel too bad though, as this debate has been going around in circles for years.
50k spins should do it.... wherever they came from.
Quote from: Joe on Feb 12, 04:07 AM 2020The roulette wheel is after all just a random number generator.
Only if your approach is completely ineffective.
Quote from: Taotie on Feb 12, 05:55 AM 2020all types of fair random number generators are equal.
I'll be publishing an article about beating rng tomorrow. But it won't help flat earthers who refuse to open eyes.
Ikeepmateethinajarbesidemabed
Quote from: Steve on Feb 12, 07:05 AM 2020
Only if your approach is completely ineffective.
I'll publishing an article about beating rng tomorrow. But it won't help flat earthers who refuse to open eyes.
Ikeepmateethinajarbesidemabed
For rng roulette you need insider infos to beat it, its like try to beat a slot machine.You need to know the RNG "algorithm"
Yes, you do. The article covers that. Ie ways to detect and exploit anomalies.
Ok Steve, we'll be waiting for your article
Quote from: Steve on Feb 12, 07:05 AM 2020Quote from: Joe on Today at 04:07 AM
The roulette wheel is after all just a random number generator.
Only if your approach is completely ineffective.
Steve, ok let me qualify that. The 'essence' of a roulette wheel is to be an RNG, although of course particular wheels may be biased. Even if your approach is effective on a biased wheel, it will be ineffective on a fair wheel if the only thing you have to go on are the numbers themselves and no other data. I'll be interested to see your post on beating RNGs though.
Quote from: Steeefan2014 on Feb 12, 06:10 AM 2020If there is a logical explanation, I'm open to that, ofc.
I'd like to hear your logical explanation of why you think testing is invalid unless done with a live dealer? I'm talking about typical systems here and not anything based on ball movement and speed, etc.
Quote from: Clf7 on Feb 12, 07:30 AM 2020For rng roulette you need insider infos to beat it, its like try to beat a slot machine.You need to know the RNG "algorithm"
Many, if not most modern RNGs are hardware based like random.org. They use atmospheric noise as the source of randomness, so there's no algorithm to crack. The old excel RNG was reputed to be pretty bad in that you could predict the output to some extent, but even casinos which don't use a hardware RNG would never use an algorithm which could be exploited for possible weaknesses.
A roulette RNG isn't like a slot machine algorithm which can be programmed to adjust the payouts; it works just like a real wheel except that the numbers are simulated.
What about airball? Can we consider it a good way to test? I mean... Is there in any way different from a normal wheel? Is it a good chance to be rigged? Should we avoid it?
Quote from: Joe on Feb 12, 08:58 AM 2020
I'd like to hear your logical explanation of why you think testing is invalid unless done with a live dealer? I'm talking about typical systems here and not anything based on ball movement and speed, etc.
When we are talking about live dealer, you can not exclude the ball movement, the speed the dealer is implementing on the ball/wheel. It's impossible to do that. Airball... is the next closest thing, being a physical roulette.
RNG... is a software based number generator. When we're talking about RNG, all rules are off. There you will be able to see things that at this moment you might consider impossible based on any theory/system you have in mind (probabilities, strategies, pattern breakers, whatever else is out there)
Quote from: Steeefan2014 on Feb 12, 10:37 AM 2020When we are talking about live dealer, you can not exclude the ball movement, the speed the dealer is implementing on the ball/wheel. It's impossible to do that.
Not sure what you mean when you say it's impossible to exclude ball movement etc. Most roulette players at B &M casinos don't take any notice of the ball movements etc and certainly don't factor it into their play.
QuoteRNG... is a software based number generator. When we're talking about RNG, all rules are off. There you will be able to see things that at this moment you might consider impossible based on any theory/system you have in mind (probabilities, strategies, pattern breakers, whatever else is out there)
What sort of things do you mean? like seeing 100 reds in a row?
You never see 'impossible' things on an RNG that you wouldn't also see on a real wheel because they are both random number generators.
The fact that you can't exclude ball movement etc from a real wheel/dealer doesn't mean that those things change the pattern of results when looking at the spins. Both an electronic RNG and a real wheel are designed to generate random results which means that every number has an equal chance of hitting over time. If you look at the actual recorded results from a real wheel you'll see that this is the case. The ball movment, speed of wheel etc are irrelevant.
For someone using a typical mechanical system, and as long as the RNG is not crooked, they will get the same results using an RNG or real wheel because every number is equally likely to hit in both cases and so they will generate produce the same probabilities and patterns.
If you don't believe this, try to identify some spins (a) from an unbiased wheel and (b) downloaded from random.org. I guarantee you won't be able to tell the difference. ;) That's why it's ok to use RNG to test systems.
Quotetry to identify some spins (a) from an unbiased wheel and (b) downloaded from random.org. I guarantee you won't be able to tell the difference. ;) That's why it's ok to use RNG to test systems.
I can at a much higher rate than probability would dictate as being possible. Just provide statistically relevant sample sizes.
And I'm willing to prove it.
General, what do you call a 'statistically relevant' sample size?
Guys forget it, it is very simple.Normal Roulette(physical wheel) can be beaten with Steves Computers,Kimo Li's method
(ofc not as effective like steves) or several other methods that maybe we dont know or we will find out.Until now nobody has beaten RNG roulette with any method(except Insider infos from algorithm designers),simple as that.Thats the reason Casino owners love rng roulette because there they can make extra profit(not forget greed is endless).
Quote from: Steeefan2014 on Feb 11, 05:07 PM 2020IMO RX or random.org are nothing else but some number generators when it comes to testing systems. I mean.. the only difference between these softwares and you writting 10000 numbers on paper is that the softwares are faster.
Not that it matters, but I think there would be a difference between a wheel vs human or rng vs human. I think it would be very difficult for a human to write down 10,000 roulette numbers that are not biased in some way.
Btw I agree with Joe. There is no difference between rng a real wheel results. Test them yourselves if you think there is.
A file that contains 3981 continuous real wheel spins is shared in this thread. (link:s://:.rouletteforum.cc/index.php?topic=10392.0)
A file that contains 30K casino rng spins is shared in this thread. (link:s://:.vlsroulette.com/index.php?topic=10046.msg63041#msg63041)
Just keep the first 3981 spins from the rng spins and compare the two. I did a few tests, the stats are almost identical. The funniest thing is that there are exactly 550 number cycles in both set of 3981 spins. ;)
Quote from: Clf7 on Feb 12, 02:45 PM 2020
Guys forget it, it is very simple.Normal Roulette(physical wheel) can be beaten with Steves Computers,Kimo Li's method
(ofc not as effective like steves) or several other methods that maybe we dont know or we will find out.Until now nobody has beaten RNG roulette with any method(except Insider infos from algorithm designers),simple as that.Thats the reason Casino owners love rng roulette because there they can make extra profit(not forget greed is endless).
That's beside the point. I'm not denying that AP is the way to go, but if you're testing mechanical systems of the kind 'bet dozen 1 after it's gone missing for 10 spins' it doesn't make a whit of difference whether you use RNG or real spins. That is the subject and point at issue of the OP.
Quote from: ati on Feb 12, 03:01 PM 2020Just keep the first 3981 spins from the rng spins and compare the two. I did a few tests, the stats are almost identical. The funniest thing is that there are exactly 550 number cycles in both set of 3981 spins.
I did a similar test using chi-square. No significant difference between the sets. TBH, I don't know what the general is talking about. He's an expert on bias wheels but I'm talking about unbiased wheels, unless he's saying that there's no such thing as an unbiased wheel. Maybe so, but it can't be the case that every wheel is biased to the degree that you can profit from it.
Quote from: Joe on Feb 12, 03:20 PM 2020
That's beside the point. I'm not denying that AP is the way to go, but if you're testing mechanical systems of the kind 'bet dozen 1 after it's gone missing for 10 spins' it doesn't make a whit of difference whether you use RNG or real spins. That is the subject and point at issue of the OP.
Ah yes If that is the point then you are right, but those are not winning systems...I say that winning systems that Beat physical wheels cant beat RNG.
one question: airball is rng or real wheel?
thx
The general needs 50,000 spins from a wheel & 50,000 spins from a rng to make it a statistically relevant sample.
He needs to compare two sets like this for 50,000 times to make his results a statistically relevant sample.
Who wants to post the first 50,000 spin samples?
C'mon guys, get involved. I'll wait until everyone has posted a combined 49,999 samples of 50,000 for the general, then I'll post the final number 50,000, 50,000 spin sample.
Quote from: Taotie on Feb 12, 04:07 PM 2020The general needs 50,000 spins
On a 3 standard deviation level it could be possible to have positive results up to ~ 50,000 spins.
The gifted of you can do the calculation. O0
Quote from: Joe on Feb 12, 03:24 PM 2020
I did a similar test using chi-square. No significant difference between the sets. TBH, I don't know what the general is talking about. He's an expert on bias wheels but I'm talking about unbiased wheels, unless he's saying that there's no such thing as an unbiased wheel. Maybe so, but it can't be the case that every wheel is biased to the degree that you can profit from it.
Show me 25k or 50k from a live wheel and put the data next to an rng wheel. Even if you think that the live wheel is random.
Even on what you consider a random wheel there are differences in how the data is distributed and the chi squares drifts to far away from the norm. In short, every live wheel is biased to some degree.
Ok general, show us your magic. :D
I've attached 3 spin files just to make it a bit more interesting. There are 50k spins in each. One contains actual spins from a wheel and the other two are RNG : one from random.org and the other are generated from a good quality software RNG.
Of course there is a 33.3% you'll identify the actuals file just by chance, so you should show using stats why you have chosen it.
You get extra credit if you can differentiate the RNG spins. ;D
Quote from: Joe on Feb 13, 11:33 AM 2020
Ok general, show us your magic. :D
I've attached 3 spin files just to make it a bit more interesting. There are 50k spins in each. One contains actual spins from a wheel and the other two are RNG : one from random.org and the other are generated from a good quality software RNG.
Of course there is a 33.3% you'll identify the actuals file just by chance, so you should show using stats why you have chosen it.
You get extra credit if you can differentiate the RNG spins. ;D
Before I start, I must ask...did you collect these spins yourself or are you relying on one of the German casino streams?
If you're using the German online data banks then you're not just looking at one specific wheel, but several different wheels that have been rotated in and out of a specific location.
Very few people have access to significantly large spin samples segregated by wheels, and unfortunately there aren't any available online.
Quote from: The General on Feb 13, 02:16 PM 2020Before I start, I must ask...did you collect these spins yourself or are you relying on one of the German casino streams?
Neither, but they are from a reliable source and definitely from just one table.
Quote from: Joe on Feb 13, 03:47 PM 2020
Neither, but they are from a reliable source and definitely from just one table.
They all look random as hell. If one of them is a live wheel then it's definitely a bi directional wheel. I need single spin direction. I have the data if you need it.
Again, don't assume that the German wheels are always the same wheel and the same is true for online wheels. They are changed out regularly.
The file with the actuals is spin2.txt. I got them from a website called rouletteresearch.com (no longer there) some time ago. You probably know of the owner George Melas who is, or was, the chief wheel engineer at Huxley. The spins came from a London casino if I remember rightly.
QuoteI need single spin direction.
But if you need additional data you're crossing into AP territory and the point of the OP was that testing of
systems should be done using live wheel spins. Do you agree with him? I've just shown that it makes no difference, haven't I?
Of course if a wheel isn't maintained then over a period it's going to become somewhat biased to the extent that it's vulnerable, but this isn't the 'typical' case, is it?
Maybe you can answer a question regarding VB; if a wheel is level (no dominant diamonds) is it possible to use VB effectively on it? There are vendors who sell VB packages and claim that their methods can be used on level wheels, but I'm skeptical.
VB can sill beat perfectly level wheels, although only in strict conditions. It takes significantly more analysis, a better understanding of the physics, and a practical approach to make level-wheel play viable with VB.
Quote from: Joe on Feb 14, 04:19 AM 2020
The file with the actuals is spin2.txt. I got them from a website called rouletteresearch.com (no longer there) some time ago. You probably know of the owner George Melas who is, or was, the chief wheel engineer at Huxley. The spins came from a London casino if I remember rightly.
But if you need additional data you're crossing into AP territory and the point of the OP was that testing of systems should be done using live wheel spins. Do you agree with him? I've just shown that it makes no difference, haven't I?
Of course if a wheel isn't maintained then over a period it's going to become somewhat biased to the extent that it's vulnerable, but this isn't the 'typical' case, is it?
Maybe you can answer a question regarding VB; if a wheel is level (no dominant diamonds) is it possible to use VB effectively on it? There are vendors who sell VB packages and claim that their methods can be used on level wheels, but I'm skeptical.
99.9% of they system players aren't going to be able to tell the difference because most of the them don't play for longer than 10 or 15 spins and they view 100 spins as being the long term. So no, the lost souls like Notto aren't going to know the difference.
The live data from George's shifty site was bi directional data, and it will wash out a great deal of visible bias. I'm sure he also cherry picked the best data to fit his site. If it was single direction the chi squares and graphs would differ sharply from the RNG. I have single direction data from many different wheels and can demonstrate the differences. Oddly enough, most of that data was from newer wheels too.
By the way, it's not always the wood expansion and contraction that causes the problems with wheels, it's usually the assembly. Huxley will sometimes put some out there.
QuoteMaybe you can answer a question regarding VB; if a wheel is level (no dominant diamonds) is it possible to use VB effectively on it? There are vendors who sell VB packages and claim that their methods can be used on level wheels, but I'm skeptical.
I don't know much about vendor claims. I don't like answering some questions regarding VB for trade secret reasons, so I'll simply say, Santa Claus is real.
QuoteI've attached 3 spin files just to make it a bit more interesting.
@ Joe...can you please tell me where is spin file number 3 from....thanks
maestro, I generated the spins from the program language I use. It uses an algorithm called mersenne twister, which isn't the best, but it passes most statistical tests of randomness and is popular.
link:s://en.wikipedia.org/wiki/Mersenne_Twister
thanks.. :thumbsup: